Number Tu Biology

The Biology
of the Xenarthra
Edited by Sergio F. Vizcano and W. J. Loughry
university press of florida
The Biology of the Xenarthra
University Press of Florida

Florida A&M University, Tallahassee
Florida Atlantic University, Boca Raton
Florida Gulf Coast University, Ft. Myers
Florida International University, Miami
Florida State University, Tallahassee
New College of Florida, Sarasota
University of Central Florida, Orlando
University of Florida, Gainesville
University of North Florida, Jacksonville
University of South Florida, Tampa
University of West Florida, Pensacola
The Biology of the Xenarthra
Edited by Sergio F. Vizcano and W. J. Loughry

Gainesville/Tallahassee/Tampa/Boca Raton
Pensacola/Orlando/Miami/Jacksonville/Ft. Myers/Sarasota
Copyright 2008 by Sergio F. Vizcano and W. J. Loughry

All rights reserved
Library of Congress Cataloging-in-Publication Data
The biology of the Xenarthra / edited by Sergio F. Vizcano
and W. J. Loughry.
p. cm.
Includes bibliographical references and index.
ISBN 978-0-8130-3165-1 (cloth)
ISBN 978-0-8130-3718-9 (e-book)
1. Xenarthra. I. Vizcano, Sergio F. II. Loughry, W. J.
QL737.E2B56 2008
599.3'1dc22 2007042546
The University Press of Florida is the scholarly publishing
agency for the State University System of Florida, comprising
Florida A&M University, Florida Atlantic University, Florida
Gulf Coast University, Florida International University,
Florida State University, New College of Florida, University
of Central Florida, University of Florida, University of North
Florida, University of South Florida, and University of West
Florida.
15 Northwest 15th Street
Gainesville, FL 32611-2079
http://www.upf.com
To Bill and Janet Loughry for then,

Colleen McDonough for now,
and Claire Loughry forever.
WJL
To Eric and Negra, my hardy roots,

Miriam, my unwavering guidepost,
and Josefina and Julieta, my flowering branches.
SFV
Contents
List of Figures ix
List of Tables xiii
Preface xv
1. Xenarthran biology: Past, present, and future 1
Sergio F. Vizcano and W. J. Loughry
Part 1. Phylogeny
2. Recent advances and future prospects in
xenarthran molecular phylogenetics 11
Frdric Delsuc and Emmanuel J. P. Douzery
3. Morphology-based investigations of the

phylogenetic relationships among extant and
fossil xenarthrans 24
11. The spleen of the armadillo: Lessons of organ

adaptation 120
Emma B. Casanave and Elena J. Galndez
12. The use of armadillo clones from the genus

Dasypus as experimental models to investigate the
source of physiological variability 126
Patrice Boily
13. The physiology of two- and three-toed sloths 130

Desmond Gilmore, Denia Fittipaldi Duarte, and Carlos
Peres da Costa
14. Sperm evolution in dasypodids 143

Pablo D. Cetica and Mara Susana Merani
Timothy J. Gaudin and H. Gregory McDonald
15. Reproductive biology of the nine-banded

armadillo 151
Part 2. Fossil Xenarthra
Richard D. Peppler
4. Fossil history of sloths 39

H. Gregory McDonald and Gerardo De Iuliis
5. Paleogeographic distribution and anatomical

adaptations in Peruvian megatheriine ground
sloths (Xenarthra: Megatherioidea) 56
Franois Pujos
6. Skeletal anatomy and the fossil history of the

Vermilingua 64
H. Gregory McDonald, Sergio F. Vizcano, and
M. Susana Bargo
7. The evolution of armored xenarthrans and a

phylogeny of the glyptodonts 79
Juan Carlos Fernicola, Sergio F. Vizcano, and
Richard A. Faria
8. Form, function, and paleobiology in

xenarthrans 86
Sergio F. Vizcano, M. Susana Bargo, and Richard A. Faria
Part 3. Living Xenarthra: Physiology and Genetics

9. Armadillos and dimorphic pathogenic fungi:
Ecological and evolutionary aspects 103
Eduardo Bagagli and Sandra de Moraes Gimenes Bosco
10. Leprosy 111

Richard W. Truman
16. Reproductive parameters and placentation in

anteaters and sloths 160
Kurt Benirschke
17. Placentation in armadillos, with emphasis on

development of the placenta in polyembryonic
species 172
Allen C. Enders
18. Sequencing the armadillo genome 181

Jean L. Chang and James E. Adams
19. Chromosomal studies in the Xenarthra 196

Wilham Jorge and Helio Rubens Jacintho Pereira Jr.
20. Genes and demes: Population genetic analyses of

the Xenarthra 205
Paulo A. Prodhl, W. J. Loughry, and
Colleen M. McDonough
Part 4. Living Xenarthra: Conservation

21. Conservation status of the Xenarthra 215
John M. Aguiar and Gustavo A. B. da Fonseca
22. Maintenance of Xenarthra in captivity 232

Mariella Superina, Flvia Miranda, and Tinka Plese
23. Exploitation of xenarthrans by the GuaranIsoseo indigenous people of the Bolivian Chaco:
Comparisons with hunting by other indigenous
groups in Latin America, and implications for

conservation 244
28. Ecology and conservation of three species of

armadillos in the Pampas region, Argentina 300
Andrew J. Noss, Rosa Leny Cullar, and Erika Cullar
Agustn M. Abba and Marcelo H. Cassini
Part 5. Living Xenarthra: Ecology and Behavior
29. Biology and ecology of armadillos in the Bolivian

Chaco 306
24. Anteater behavior and ecology 257

Flvio H. G. Rodrigues, sis M. Medri, Guilherme H. B. de
Miranda, Constana Camilo-Alves, and Guilherme Mouro
25. Sloth ecology: An overview of field studies 269

Adriano Garcia Chiarello
26. Behavioral ecology of armadillos 281

Colleen M. McDonough and W. J. Loughry
27. Xenarthrans of the Paraguayan Chaco 294

Dennis A. Meritt Jr.
Erika Cullar
30. The natural history of the pichi, Zaedyus pichiy, in

western Argentina 313
Mariella Superina
Bibliography 319
List of Contributors 357
Taxonomic Index 361
Subject Index 365
Figures
11.3. The transitional zone of Chaetophractus

vellerosus 123
11.4. The red pulp of Zaedyus pichiy 124
11.5. The sinus wall of the spleen of Chaetophractus
vellerosus 124
14.1. Frontal plane of a Dasypus septemcinctus
spermatozoon 145
14.2. Dasypus septemcinctus sperm head 145
14.3. Sperm dimensions of the four dasypodid
sperm groups 146
14.4. Frontal plane of Chaetophractus vellerosus
spermatozoa 147
14.5. Concave side of a Euphractus sexcinctus sperm
head 147
14.6. Frontal plane of a Priodontes maximus
spermatozoon 147
14.7. Frontal plane of a Tolypeutes matacus
spermatozoon 147
14.8. Convex side of a Tolypeutes matacus sperm
head 147
14.9. Sagittal section of a Chaetophractus villosus
sperm head and transverse section of a middle
piece of the sperm tail 148
14.10. Proposed evolutionary trend for spermatozoa in
dasypodids 149
15.1. Placenta from female armadillo with four
young 152
15.2. Ontogeny of plasma testosterone concentration
in male armadillos 153
15.3. Annual plasma progesterone levels in
nonpregnant adult female armadillos 154
15.4. Annual plasma progesterone levels in a
second group of nonpregnant adult female
armadillos 155
15.5. Plasma progesterone levels in pregnant adult
female armadillos during delayed implantation
and gestation 156
15.6. Plasma progesterone and estradiol levels in
nonpregnant adult female armadillos 156
15.7. Daily changes in plasma progesterone
concentration for female armadillos 157
2.1. Remaining uncertainties in armadillo molecular

phylogeny 17
2.2. Phylogenetic relationships and molecular
timescale of extant xenarthran genera 19
3.1. Cladogram depicting the relationships among
the three suborders of Xenarthra 25
3.2. Cladogram summarizing relationships within
the Cingulata 29
3.3. Cladogram summarizing relationships within
the Phyllophaga 31
4.1. Oligocene sloths 45
4.2. Early and middle Miocene sloths 47
4.3. Late Miocene sloths 49
4.4. Pleistocene sloths from South America 50
4.5. Pleistocene sloths from North America 51
4.6. Late Pleistocene sloths from the Caribbean 52
5.1. Paleogeographic distribution of megatheriine
ground sloths in Peru 58
5.2. Skull and dentary, humerus, and femur of
Megatherium (Megatherium) americanum
and Megatherium (Pseudomegatherium)
tarijense 60
5.3. Hypothetical reconstruction of Megatherium
(Pseudomegatherium) 61
6.1. Skulls of extant and fossil Vermilingua 67
6.2. Hypothesized phylogeny of the Vermilingua 72
7.1. Drawing of cingulates 82
7.2. Phylogenetic relationships of glyptodonts 84
8.1. Early drawing of Megatherium americanum 88
8.2. Skeletons of fossil and extant tardigrades 90
8.3. Skeletons of fossil and extant cingulates 93
8.4. Skulls and mandibles of various cingulates 95
8.5. Skulls and mandibles of various tardigrades 96
9.1. P. brasiliensis dimorphism 106
9.2. Necropsy of D. novemcinctus viscera 108
10.1. Map of the southeastern United States by
county indicating areas of incidence of and
sampling for leprosy 116
11.1. The splenic parenchyma of Zaedyus pichiy 121
11.2. An interdigitant-like cell from the marginal
zone of Chaetophractus vellerosus 122
ix
Figures
15.8. Plasma progesterone levels of adult female

armadillos in response to injection with
clomiphene citrate 158
15.9. Daily changes in plasma testoterone
concentration for male armadillos 159
16.1. Macroscopic appearance of a giant anteater
placenta 162
16.2. Gross appearance of another term placenta
from a giant anteater 162
16.3. Giant anteater placenta 163
16.4. Section from near the margin of a giant anteater
placenta 164
16.5. Site of attachment in the giant anteater 165
16.6. Choloepus didactylus gestation 166
16.7. Placenta of Choloepus hoffmanni 166
16.8. Site of implantation in C. hoffmanni 166
16.9. Cross section through placental labyrinth of C.
hoffmanni 167
16.10. Electron micrograph of barrier in C.
didactylus 167
16.11. So-called pigmented epithelioid (Hofbauer)
cells in C. didactylus 168
16.12. Uterus with fetus attached to placenta by short
umbilical cord of Bradypus tridactylus 169
16.13. Opened uterus of three-toed sloth (B.
tridactylus) 169
16.14. Portion of the attached placenta from a threetoed sloth (B. tridactylus) 170
16.15. The relation of a villous/lamellar fetal
structure in B. tridactylus 170
16.16. Higher magnification of maternal-fetal relation
in B. tridactylus 171
16.17. Edge of the placenta of the three-toed sloth, B.
tridactylus 171
17.1. Sagittal section of an armadillo uterus prior to
implantation 173
17.2. An early implantation stage in the fundus of the
uterus 174
17.3. Diagram of an early implantation site 175
17.4. A more advanced implantation site after loss of
the abembryonic trophoblast 175
17.5. Higher magnification showing numerous
trophoblast cells around the dark gland
cells 175
17.6. Two embryonic shields 176
17.7. Portions of four embryonic shields 176
17.8. Depiction of a developing placenta within the
uterus 177
17.9. Placental villi 177
17.10. Area of attachment 178

17.11. Placenta in late pregnancy 178
17.12. A mature villus 179
18.1. Phylogeny of placental mammals selected for
low-coverage sequencing 183
18.2. Sequencing and scaffold assembly strategy 186
18.3. Enhancement of de novo assembly by alignment
of WGS reads to a reference genome 188
18.4. Human genes found by nucleotide BLAST to
armadillo scaffolds 190
19.1. Karyotypes of the three-toed sloths (Bradypus) 199
19.2. Karyotypes of two-toed sloths (Choloepus) 201
19.3. Karyotype of Cabassous sp. 203
19.4. Karyotype of Tamandua sp. 204
20.1. Microsatellite gel of nine-banded armadillos 207
20.2. Spatial dispersion of adult and juvenile
cloneships in a population of nine-banded
armadillos 208
21.1. Current distributions of sloth species 218
21.2. Current distributions of anteater species 221
21.3. Current distribution of the giant armadillo,
Priodontes maximus 222
21.4. Current distributions of armadillos in the genus
Chaetophractus 223
21.5. Current distribution of Euphractus sexcinctus 224
Tolypeutes 225
Cabassous 226
Chlamyphorus 227
21.9. Current distribution of the pichi, Zaedyus
pichiy 228
Dasypus 229
22.1. Captive Priodontes maximus feeding on a
termite mound 233
22.2. Chaetophractus villosus kept in an inadequate
enclosure 234
22.3. Chaetophractus vellerosus with vitamin A
deficiency and pyodermia 235
22.4. An appropriate enclosure for Tamandua 237
22.5. Thoracic radiographs of a healthy giant anteater
and a giant anteater with cardiomyopathy 238
22.6. Anteater enclosure enriched with various
items 240
22.7. Juvenile Bradypus variegatus with fungal
infection 240
22.8. Enriched captive environment for sloths 241
Figures
22.9. Juvenile Bradypus variegatus feeding on fresh,

hydrated leaves 242
23.1. Kaa-Iya National Park and Isoso indigenous
territory, Bolivia 246
23.2. Isoseo hunting of armadillos (19962004) 248
24.1. Home ranges of giant anteaters, Myrmecophaga
tridactyla 262
25.1. Average percent of time five radio-collared
sloths spent resting, feeding, moving, and
grooming 280
25.2. Average distances five radio-collared sloths
traveled during the day, night, and 24 h 280
28.1. Distribution of three species of armadillos in
the Pampas region 302
xi
28.2. Diagrams of the burrows of the three species of

armadillos 303
29.1. Activity patterns for four armadillo species 310
29.2. Armadillos captured by camera traps in
different areas of the Bolivian Chaco 311
30.1. Sampling areas in Mendoza province,
Argentina 314
30.2. Head morphology of Zaedyus pichiy 315
30.3. Patagonian steppe in southern Mendoza
province, Argentina 316
30.4. Chacoan monte in northern Mendoza
province, Argentina 316
30.5. Captive-born pichis, Zaedyus pichiy 318

Tables
23.2. Hunting of Xenarthra and Other Principal

Mammal Taxa by Indigenous Peoples in
Bolivia 249
23.3. Uses of Xenarthra by Indigenous Groups in
Bolivia 250
23.4. Hunting of Xenarthra and Other Principal
Mammal Taxa by Indigenous Peoples Outside of
Bolivia 250
23.5. Uses of Xenarthrans by Indigenous Peoples
Across the Neotropics 253
24.1. Proportions of Termites and Ants in the Diet of
Anteaters 259
24.2. Home Ranges of Giant Anteaters 261
25.1. Features of Sites Where Ecology and Behavior of
Sloths Have Been Studied 271
25.2. Composition of Diets of Three Sloth Species in
the Wild 272
25.3. Diet of Three-Toed Sloths at Four Study Sites 273
25.4. Home-Range Size and Day-Range Length of
Three Sloth Species 275
25.5. Density and Sex Ratio Estimates of Five Sloth
Species 277
25.6. Main Activities of Four Sloth Species 278
26.1. Field Studies of the Behavior and Ecology of
Armadillos, 19002006 283
26.2. Field Studies of Armadillo Diets 284
26.3. Home-Range Values and Reports of Various Social
Interactions in Field Studies of Armadillos 288
28.1. Habitat Use by Three Species of Armadillos 302
28.2. Summary of Activity Patterns of Three Species of
Armadillos 303
28.3. Characteristics of the Burrows of Three Species
of Armadillos 303
28.4. Diet Composition of Three Species of
Armadillos 304
29.1. Reproductive Status of Female Armadillos Killed
by Hunters 309
29.2. Principal Components of Armadillo Diets 309
29.3. Percentage of Animals Collected by Hunters in
Five Habitats 309
30.1. Morphometric Measurements of Adult WildCaught Pichis 315

2.1. Where Do Xenarthrans Fit among Placentals? 15

4.1. Geographic and Biochronologic Distribution of
Sloth Genera 43
6.1. Chronologic Distribution of Fossil
Vermilingua 73
8.1. Estimations of Body Mass of Some Fossil
Xenarthrans 91
9.1. Fungal Detection in Armadillos 107
10.1. Representative Clinical Values for Normal
Armadillos and Animals Experimentally
Infected with M. Leprae 114
10.2. Locations and Features of Wild Armadillos
Examined for Naturally Acquired Leprosy 117
11.1. Morphometric Data for the Spleens of Four
Species of Armadillos 121
11.2. Comparative Characteristics of the Red Pulp of
Armadillo Spleens 123
13.1. Summary of Cardiac Function and Arterial
Blood Pressure in Sloths 136
13.2. Respiratory Rates Measured under Different
Conditions in Sloths 138
13.3. Pulmonary Volumes and Capacities in
Anesthetized Sloths 139
13.4. Features of the Respiratory System of B.
Variegatus 139
15.1. Plasma Progesterone Levels in Female Armadillos
Following Unilateral Ovariectomy 157
18.1. Mammalian Species Selected for LowRedundancy Sequencing 183
18.2. Coverage of Selected Armadillo Genes Found in
DasNov1 Contigs 193
19.1. Characterization of Species in the Genus
Bradypus 198
19.2. Characterization of Species in the Genus
Choloepus 200
19.3. Chromosomal Features of Armadillo Species 202
19.4. Karyotypic Data for Anteater Species 203
21.1. Conservation Status and Main Threats to the
Extant Xenarthrans 217
23.1. Subsistence Hunting of Mammals in Isoso,
19962004 247
xiii
Preface
be followed by shorter chapters covering specific topics

or taxa. To some extent, this organization has been retained in the books final form, but there was considerable modification along the way. Ultimately, we found
it more useful to separate chapters by subject matter.
As the book is now formatted, each topical section begins with one or more longer review chapters. In most
cases, shorter, more targeted chapters then follow. We
have endeavored to cross-link all chapters as much as
possible to avoid needless repetition of the same material. We hope we have succeeded and the result is a
book that readers will find reasonably well integrated
and easy to read.
Unquestionably, the inspiration for this book is The
Evolution and Ecology of Armadillos, Sloths, and Vermilinguas (Montgomery 1985a). In many respects, our
X book is directly modeled on that volume. However,
there are some differences. For example, considerable
space was devoted to systematic descriptions of living
taxa in Evolution and Ecology. We view most of those
issues as having been resolved and so there is little mention of them here. However, there is one item we must
address. Gardner (2005) recently published a taxonomic scheme that differed in three major ways from
previous work. The first change elevated Xenarthra to
the level of superorder with Cingulata (the armadillos)
and Pilosa (the anteaters and sloths) as the two contained orders. Second, Cyclopedidae was recognized
as a separate family of anteaters. Third, a new genus of
armadillos, Calyptophractus, was created to contain one
of the species of fairy armadillos formerly in the genus
Chlamyphorus. We admit to no strong feelings on these
issues. For example, regarding the taxonomic rank of
Xenarthra, it seems to us the key thing is the evolutionary relationships within the group. Thus, whether Xenarthra is viewed as an order, superorder, or something
else seems less relevant than the fact that all available
evidence indicates it is a monophyletic group containing two distinct clades. Recent classification schemes
that extend beyond extant taxa to include fossil forms
(e.g., McKenna and Bell 1997) are not without their own
set of problems as well, some of which are detailed later
The beginnings of this book occurred during our brief

visit at the Florida Museum of Natural History, Gainesville, in August 2004. At some point the conversation
turned to the fact that the last authoritative work on
the Xenarthra, The Evolution and Ecology of Armadillos,
Sloths, and Vermilinguas, edited by Gene Montgomery
(1985a), was nearly 20 years old. On top of that, the
chapters in that book were based on a conference held
in 1979. Thus, much of the information in the book was
25 years old. Except for a symposium volume on the
morphology of living and fossil xenarthrans (Faria
et al. 2003), no comprehensive or synthetic reviews of
xenarthran biology had been published in the interim.
The fact that Evolution and Ecology had held up so well
was a testament to the books authors. Indeed, it was
still the obligate reference and, in our experience, it
would have been very difficult to think of a single article on xenarthrans that did not reference one or more
of its chapters. Nonetheless, 25 years is a long time and
we felt that so much new information had accumulated
that perhaps it was time for an update.
Our goal in developing this book, which we informally referred to as the X Book, was to provide the
best, most comprehensive review of current research on
the Xenarthra. Unfortunately for the study of the Xenarthra, but fortunately for us, the xenarthran research
community is relatively small. We invited most of those
currently active within this community to contribute
something to this book. Luckily, nearly everyone accepted. The result is a volume that we hope will hold
up as well as Montgomerys Evolution and Ecology (although we hope this new volume will inspire so much
new research that it will not take another 20 years to
produce the next update) and become the new standard
reference on the Xenarthra for students and researchers
alike.
Because the xenarthran community is so small, we
cannot claim that this volume represents a comprehensive treatment of all aspects of xenarthran biology.
Rather, it was our intent to highlight the work currently being done. As originally envisaged, a handful
of longer review chapters covering broad issues would
xv
xvi
Preface
in this book. Consequently, the reader should be alerted

that we have opted against strict adherence to any particular taxonomic framework, preferring instead to
highlight the ongoing dynamics and disagreements associated with xenarthran phylogeny. As one conspicuous example of this, the sloth clade is currently referred
to by multiple names: Phyllophaga, Tardigrada, and Folivora (cf. Faria and Vizcano 2003). It is not clear how
this nomenclatural issue will be resolvedor whenso
in this book we have allowed each author to use the
name he or she prefers. We hope the reader does not
find this too confusing. To conclude this discussion by
returning briefly to Gardners (2005) scheme, there was
little disagreement with recognition of Cyclopedidae as
a distinct family, so it is recognized as such here. However, most authors felt creation of Calyptophractus was
premature, as was elevation of Xenarthra to superorder.
Consequently, in this volume both fairy armadillos are
retained in the genus Chlamyphorus and Xenarthra is
still referred to as an order.
Another feature of Evolution and Ecology is that it
contained reports primarily from North American
scientists. The Xenarthra are a quintessentially South
American group and much of the work on them has
been done by individuals from Latin America. One of
the more obvious differences between our X book and
the Montgomery volume is the increased diversity of
the contributors. Indeed, one of our hopes for the book
is that it will serve to substantially raise the profile of
xenarthran biologists working in Latin America.
There are a number of people we would like to thank
for their contributions to this book. First, our thanks
to all the chapter authors, who managed to meet our
deadlines (usually) and worked diligently to ensure
that their papers (the X files) were accurate, up to
date, and correctly formatted. The staff at the University Press of Florida, in particular editor-in-chief John
Byram, was extremely helpful with all the logistical
and technical aspects of putting this book together. We
would also like to thank all the reviewers who commented on various chapters: R. Aguilar, M. Anderson,
P. Arenas, B. Bagatto, S. Bargo, K. Benirschke, P. Brennan, C. Cartelle, M. Cassini, P. Cetica, M. G. Chavez,
A. Chiarello, D. Croft, E. Cullar, G. De Iuliis, P. Deps,
K. Dewar, A. Enders, R. Faria, J. Garcia, T. Gaudin, J.
Gramieri, F. Knight, P. Lara-Ruiz, D. Lorton, C. McDonough, R. McPhee, D. Meritt, N. Moraes-Barros, W.
Murphy, G. Navone, A. Noss, D. Perea, F. Pujos, A. Restrepo, E. Roldan, A. Rylands, P. Samollow, R. Samudio,
G. Scillato-Yan, W. Sechrest, J. Shaw, E. Taube, and J.
Wible. In addition to these reviews of specific chapters,
Michael Mares and Don Wilson read and commented
on the entire volume. Finally, a special note of thanks
to Agustn Abba, Leonardo Avilla, Pablo Cetica, and
Cristina Bertoni Machado for their tireless aid in translating and proofing the abstracts. We are grateful for all
these contributions as they substantially improved the
quality of the book.
To conclude on a more personal note, WJL would
like to thank Valdosta State University for providing
him with reassigned time to work on this book, John
Pascarella for help with translations, and Colleen McDonough for help and support during all phases of the
project. SFV would like to thank his Xamigos: Susi
Bargo, Chochn Faria, and Dolape De Iuliis for
their friendship and sympathy during a difficult personal situation that occurred during the period when
most of the work on the book was being done. We believe all the work and effort has been worth it. We hope
others will agree.
1
Xenarthran biology
Past, present, and future
Sergio F. Vizcano and W. J. Loughry
Resumen
Evolution of the Xenarthra
El grupo de los Xenarthra constituye uno de los clados mayores de mamferos. Tanto las formas vivientes
como las fsiles comparten atributos que hacen de estos animales objetos de estudio altamente deseables.
Sin embargo, y sorprendentemente, a diferencia de lo
que ocurre con muchos otros grupos de mamferos, los
Xenarthra parecen haber sido bastante ignorados. En
esta seccin presentamos una sntesis de lo que hace a
estos animales tan especiales y cientficamente interesantes. Esta argumentacin nos permite delinear lo que
ya sabemos acerca de los Xenarthra y, como lo ejemplifica este libro, los progresos que se estn alcanzando
en diferentes aspectos de su conocimiento. Asimismo,
esperamos sealar aquellas reas que permanecen inexploradas y que requieren de estudios futuros.
There are, at this writing, 31 living species of xenarthrans:

6 sloths, 4 anteaters, and 21 species of armadillos. The
clade presumably originated in South America and all
extant species are found within specific regions of Latin
America (Aguiar and Fonseca this volume). In addition, one species, the nine-banded armadillo (Dasypus
novemcinctus), has successfully colonized much of the
southern United States in the last 200 years (Taulman
and Robbins 1996). These species represent but a small
fragment of a much more diverse fossil assemblage that
includes such well-known oddities as the giant ground
sloths and glyptodonts.
Current molecular evidence indicates that the Xenarthra represent one of the four major clades of placental mammals (Madsen et al. 2001; Murphy, Eizirik,
Johnson et al. 2001; Murphy, Eizirik, OBrien et al.
2001; Delsuc et al. 2002; Springer et al. 2005; see also
Delsuc and Douzery this volume). Although the most
widely accepted scenario envisions a southern origin
of the Xenarthra after the separation of the African and
South American continents about 6580 million years
ago (mya), the origin of the group is still something of
a mystery (Delsuc et al. 2003 and references therein).
This is partly because molecular data point toward an
origin about 105 mya (Delsuc et al. 2004), suggesting
the existence of a xenarthran ghost lineage that left
no traces for almost 50 million years.
Within Xenarthra, two distinct groups evolved (Engelmann 1985; Rose et al. 2005; see also the chapters
by Delsuc and Douzery, and Gaudin and McDonald
this volume): (1) the Cingulata, exemplified by mod-
Resumo
Os Xenarthra constituem um dos maiores clados entre
os mamferos placentrios. Tanto as formas viventes e
quanto as fsseis possuem inmeros atributos que tornam esses animais objetos de estudo altamente desejados. Porm, em contraste com muitos outros grupos
de mamferos, o dos Xenarthra inesperadamente negligenciado. Aqui provemos uma reviso dos aspectos
que fazem estes animais especiais e cientificamente interessantes. Essa discusso nos permite resumir o que
j sabemos sobre Xenarthra e, como exemplificado
por este livro, o que o que est sendo feito atualmente.
Adicionalmente, esperamos apontar as reas que permanecem inexploradas e merecem estudos futuros.
S. F. Vizcano and W. J. Loughry
ern-day armadillos and noted for the armor covering

their head, body, and tail, and their minimal, peglike
teeth, and (2) the Pilosa, comprising two groups: the
Vermilingua, or anteaters, characterized by an elongate,
tubular skull, the absence of any teeth, and a prehensile
tongue used in acquiring prey, and the Phyllophaga, or
sloths (also known as Folivora or Tardigrada, cf. Faria
and Vizcano 2003; Gardner 2005; Delsuc and Douzery
this volume). Modern sloths are noted as almost exclusively arboreal herbivores; however, many extinct
forms were much larger and terrestrial.
Fossil history
Xenarthrans belong to that group of mammals present
in South America at the dawn of the Tertiary. Indeed,
osteoderm fragments and postcranial bones assigned
to armadillos are found in Paleocene sediments, that
is, from about 55 mya. Because of the later diversification of the group, their fossils are among the most
prominent and abundant in South America (McKenna
and Bell 1997). Thus, together with the marsupials,
xenarthrans are the mammals with the longest history
in South America. Evolving in the splendid isolation
that the continent provided for much of the Tertiary
(Simpson 1980), xenarthrans underwent an explosive
radiation. Fossil forms representing more than 150
genera have been identified (McKenna and Bell 1997).
Establishment of land bridges allowed for faunal exchanges with North America, so that, although few species live there now, xenarthran fossils found in Central
and North America indicate their presence during the
latest Tertiary and the Pleistocene. Xenarthrans began
to disappear from Central and North America during
the last of these faunal exchanges, the Great American Interchange (Stehli and Webb 1985), although the
cause(s) of this decline remain speculative. In contrast,
xenarthrans remained very diverse in South America
until the end of the Pleistocene. Indeed, more than half
of the late Pleistocene megamammals belonged to this
group (Faria 1996). Nonetheless, just as in other parts
of the Americas, xenarthran diversity was ultimately reduced as all of these giant forms went extinct between
the end of the Pleistocene and the beginnings of the
Holocene, possibly because of their large size (Lessa
and Faria 1996; Lessa et al. 1997).
The remarkable diversity of fossil xenarthrans is the
product of many unique and, in some cases, bizarre anatomical features (McDonald 2003b; Rose et al. 2005).
Some of these include (1) the articulations of the verte-
brae, from which the name of the order is derived, (2)

well-developed articulations between the sternal ribs
and sternebrae, and (3) formation of the synsacrum.
The dentition of xenarthrans is very peculiar too: tooth
number is strongly reduced in many forms; enamel is
absent in the adult and deciduous teeth of all extant
and, possibly, all extinct xenarthrans; the cuspal patterns present in other mammals are absent; the teeth,
composed of osteodentine, are always hypselodont;
and, although they may be lobated, the teeth are usually simple and separated by short diastemata.
As mentioned above, one of the most prominent
features of fossil xenarthrans is the tremendous body
size of some forms, particularly those in the Pleistocene
megafauna (last 2 million years) of South America. Examples among the cingulates include armadillos (Dasypodidae), some of them reaching sizes of about 50 kg,
pampatheres (Pampatheriidae), giant armadillo-like
herbivores that must have reached 200 kg, and glyptodonts (Glyptodontidae), whose terminal Pleistocene
forms reached body masses varying between one and
two tons (see Fernicola et al. this volume). Within Pilosa, fossil tardigrades attained the greatest diversity,
with four families (Mylodontidae, Megalonychidae,
Nothrotheriidae, and Megatheriidae) currently recognized (Gaudin 2004a; see also McDonald and De
Iuliis this volume), and about 80 named genera. Some
of these ground sloths reached gigantic size, with body
masses up to 4 or 5 tons.
Extant xenarthrans
Modern xenarthrans have been collected and studied
scientifically since the first European expeditions to
South America. While most early work centered on
determining taxonomic position, subsequent studies
began to reveal a number of fascinating and unique
attributes of the group. For example, it was first proposed nearly 100 years ago (Newman and Patterson
1910) that armadillos in the genus Dasypus exhibited
obligate polyembryony, where a single fertilized egg
divides into multiple embryos upon implantation in
the uterus, resulting in litters of genetically identical
offspring. Species in this genus are also unique in having one of the lowest metabolic rates reported in any
placental mammal (McNab 1980) and in being one of
the few animals other than humans known to contract
leprosy (Storrs 1971).
Other xenarthrans are equally fascinating. Anteaters, with their extremely long snouts and forelimbs ex-
Xenarthran biology: Past, present, and future
tended somewhat in a gorilla-like fashion, are certainly

odd looking at first glance. However, these features are
just the external manifestations of a highly specialized anatomy and physiology adapted for obtaining
most nutritional requirements from ants and termites
(Naples 1985, 1999; Reiss 1997; see also Rodrigues et al.
this volume). For example, consider the anatomy of the
forelimb, which is used to open insect nests (Taylor
1985). Unlike in normal mammals, where the triceps
muscle is an extensor of the forearm, in anteaters the
triceps helps in the flexion of digits bearing powerful
claws. Further, the extensors of the digits act as arm
flexors, in addition to the biceps.
Living sloths are an extreme example of convergence
to a similar way of life (Chiarello this volume). Species
in both genera mostly remain suspended upside-down
during active periods. This sort of behavior is very unusual for mammals and they are so committed morphologically to this habit that they are incapable of supporting themselves quadrupedally (Mendel 1985a and
references therein). Sloths have also evolved chambered
stomachs that are used, with the aid of symbiotic microorganisms, to ferment the leaves that constitute the
major part of their diet (Chiarello this volume; Gilmore
et al. this volume).
Xenarthran research
Despite all the intriguing features outlined above, xenarthrans seem understudied relative to many other
mammalian groups. This may be less true for fossil
forms, where numerous specimens have been collected.
Yet, even so, fossil xenarthrans seem less understood
than other mammals. This is probably due in part to
the natural incompleteness of the fossil record, but it is
likely compounded by features of the fossils themselves.
For example, xenarthrans lack tribosphenic teeth. The
considerable variation in this character has been the
basis for most mammalian phylogenetic and systematic
studies for more than a century. Thus, the absence of
these teeth in fossil xenarthrans has severely hampered
efforts to incorporate them into such studies and consequently limited our understanding of their evolution.
Further, the diversity of the fossils has created problems.
For example, carapace scutes of fossil cingulates are so
abundant and diverse that a parataxonomy has been
generated that, so far, has been poorly checked against
other features such as the cranial and postcranial skeleton (but see Fernicola et al. this volume). Finally, to
illustrate a different kind of problem, the dramatically
reduced diversity of extant xenarthrans relative to fossil

forms has meant that few living analogs are available
for comparison with extinct forms. This has severely
constrained, and in some cases even misled, the interpretation of the natural history of fossil xenarthrans
(Vizcano et al. this volume).
There are probably several reasons for the neglect of
extant species. First, many of these species are extremely
difficult to observe in the wild, so field studies of the
Xenarthra have been relatively rare (see the reviews by
Chiarello, McDonough and Loughry, and Rodrigues et
al. this volume). This may be exacerbated by the fact
that many xenarthrans are hunted, for a variety of reasons, by indigenous peoples in Latin America (Noss
et al. this volume). Such hunting presumably reduces
populations and may make survivors more wary of humans, with either outcome making observation even
more difficult. Second, many xenarthrans do not seem
to adapt well to captivity (Superina et al. this volume),
so it is often difficult to maintain laboratory populations for study. For example, although D. novemcinctus
has been promoted as an important animal model for
the study of leprosy (Storrs 1971), it has so far proved
impossible to establish breeding colonies of the animals
(Truman this volume). Thus, for a variety of reasons,
captive xenarthrans are not always readily available
and, even when they are, the data obtained from them
may provide an incomplete or unrepresentative picture
of what occurs in the wild.
However, we do not mean to be too negative here.
Clearly there have always been some scientists who
have found innovative ways to further our understanding of the Xenarthra. The primary evidence for this is
the volume edited by Montgomery (1985a), which summarizes a considerable body of work by a sizable number of contributors. Indeed, the impression that the
Xenarthra have been neglected may be somewhat misleading. The book by Montgomery largely emphasized
work being done by scientists from North America. As
evidenced by the present volume, there is a rich history
of xenarthran research by Latin American scientists.
Thus, while it may be true that, since the publication of
Montgomerys book, only a handful of European and
North American scientists have devoted themselves
to studying the Xenarthra, this has not been the case
within Latin America. The appearance of neglect is
probably due, at least in part, to the sometimes limited
availability of publications by Latin American scientists
in other parts of the world, as well as to the political
and financial impediments these individuals have faced
in being able to pursue and publicize their work. It is

one of our most profound hopes that this book will do
much to raise the profile of xenarthran researchers in
Latin America, who have made significant contributions to our understanding of this odd group.
An overview of this book

While there are more people studying the Xenarthra
than may be widely appreciated, it is still true that the
community of xenarthran researchers is quite small
compared to that associated with many other animal
groups. This fact is reflected in this book. At first glance,
its contents and organization may seem somewhat arbitrary. This was a point that worried us for a time, until we realized that, just as was true for Montgomerys
book in 1985, it is just a reflection of the field. With
such a small group of active researchers, any compilation is likely to be somewhat idiosyncratic, reflecting
the particular interests of each individual. Nonetheless, we believe comparison of the present volume with
Montgomerys book will show that our knowledge of
xenarthrans is advancing on a number of fronts. So,
while perhaps not an exhaustive account, this book represents our current state of the art: this is what we know
and where we stand in the study of the Xenarthra, as
reported by the individuals who have made the major
contributions.
This book is organized into the following parts:
first, two chapters on xenarthran phylogeny. Unlike the
Montgomery volume, which spent considerable time
on the systematic description of living forms, the present work views issues in that area as largely resolved
(but see Gardner 2005). Instead, in this book, we concentrate on broader evolutionary patterns. Delsuc and
Douzery review evidence from molecular genetic analyses of extant species that shed light on evolutionary
relationships among lineages within Xenarthra as well
as the relationship of the group to other mammalian
clades. This latter issue has been intensively investigated because, given Xenarthras putative basal position, such analyses should be critical in increasing our
understanding of early mammalian evolution. Gaudin
and McDonald provide a paleontological perspective,
using morphological data from fossil and extant forms
to describe the major evolutionary patterns within
Xenarthra. While congruent in many respects, an important task for the future must be how to completely
reconcile the phylogenies generated from molecular
versus morphological data.
With the major groups of xenarthrans now identified, we turn next to more detailed analyses of various
fossil forms. McDonald and De Iuliis review the fossil
history of sloths, which is remarkable for the great diversity of forms that evolved. Pujos adds to this account
with a discussion of the origin of some very interesting
fossil sloths from the Peruvian Andes. McDonald et al.
review the more limited fossil history of anteaters, while
Fernicola et al. cover the evolution of armored xenarthrans such as armadillos and glyptodonts. Description and understanding of fossil specimens typically
requires extensive morphological analyses. Vizcano
et al. outline how an approach rooted in functional
morphology has provided fresh insights into the biology and ecology of fossil xenarthrans. Taken together,
these chapters indicate that a sizable collection of fossil
material is already available with which to reconstruct
the rich evolutionary history of the xenarthran clade.
Nonetheless, new discoveries will no doubt continue to
augment and revise our understanding of this important group. In particular, given that molecular estimates
suggest Xenathra first arose ca. 105 mya, yet no fossil
xenarthrans are known until almost 50 my later (see
above), it will be extremely important to locate specimens that allow fuller description of the early stages in
xenarthran evolution.
The remainder of the book deals with living xenarthrans. We begin coverage of these species by examining various aspects of xenarthran physiology. One
important influence on physiological functioning is disease. Bagagli and Bosco describe pathogenic fungi associated with armadillos and discuss the ecological and
evolutionary consequences of such associations. One
especially important bacterial pathogen of armadillos
is Mycobacterium leprae, which causes leprosy. Truman
reviews the study of leprosy in D. novemcinctus, arguing that the incidence of the disease may be limited to
certain biogeographical areas (e.g., low-lying humid areas with moist soils). The spleen is an important organ
in combating disease. Casanave and Galndez provide
comparative analyses of spleen morphology in four
species of armadillos, documenting impressive plasticity in the anatomy and functioning of this structure.
Other physiological issues are addressed in subsequent chapters. Boily exploits the occurrence of polyembryony in D. novemcinctus to identify the genetic
component of intraspecific variation in physiological
performance by making comparisons within and between litters of clonemates. However, there are a number of pitfalls in this type of analysis and Boily details
the pros and cons of using armadillos for such studies.

Building on the earlier work of Goffart (1971), Gilmore
et al. provide a comprehensive review of most aspects of
sloth physiology, relating these to the unusual ecology
and lifestyle of these species.
One of the better-studied aspects of xenarthran
physiology is reproduction. Cetica and Merani provide a comparative analysis of sperm morphology in
armadillos, arguing that sperm anatomy is a useful
phylogenetic trait that has evolved by selection from
various features of female reproductive tracts. Next,
Peppler describes the typical reproductive cycle of the
nine-banded armadillo. One notable feature here is the
ability of females (at least in the Northern Hemisphere)
to delay implantation of the fertilized egg for a variable
length of time. Assuming sperm do successfully fertilize the egg, embryonic development should ultimately
occur. Benirschke reviews the general features of placentation in sloths and anteaters, while Enders focuses
specifically on armadillos, as exemplified primarily by
D. novemcinctus. Of particular interest here is Enders
discussion of the evolution of polyembryony in the genus Dasypus. Galbreath (1985) proposed that the area
in the female reproductive tract where the egg was held
prior to implantation was only large enough to accommodate a single egg. Thus, the only way for females to
increase their reproductive success was to bypass this
constraint by having the egg divide into multiple copies
after implantation (see also Craig et al. 1997). Enders
provides a critical re-evaluation of this hypothesis, suggesting that the anatomical constraint does not exist. If
true, this will mean we need to completely rethink the
evolution of polyembryony in armadillos.
As with so many species, molecular genetic analyses
have become important tools for investigating xenarthran biology. Chang and Adams report on recent work
to sequence the genome of a representative xenarthran,
D. novemcinctus. Their work has both evolutionary
implications in further illuminating the phylogenetic
relationships of the Xenarthra to other mammals, and
a practical benefit of identifying potential markers for
genes that may be related to traits such as leprosy susceptibility. In addition to sequence data, various other
genetic markers have been used to analyze xenarthran
populations. Jorge and Pereira review cytogenetic studies of the Xenarthra and how they have been used in
phylogenetic analyses, while Prodhl et al. examine the
use of mitochondrial and nuclear markers in describing
patterns of genetic variation within and between populations of xenarthrans.
Practical, or applied, issues take center stage in the

next section with the review by Aguiar and Fonseca of
the serious conservation issues faced by many xenarthrans. This is underscored in both subsequent chapters. First, Superina et al. point out that, for captive
animals, many questions remain unanswered regarding
appropriate husbandry conditions and breeding success
is low for many species. Thus, any conservation plans
relying heavily on captive populations may be problematic. Second, Noss et al. describe the significant human
impacts on xenarthran populations through activities
such as hunting. It seems clear that, in many cases, time
is growing short and that we must act quickly if we are
to broaden our understanding of some of these littleknown species.
Detailed field studies of wild populations seem a
critical component in addressing some of the conservations issues just mentioned. These are described in the
final part of the book. General reviews of field studies
on anteaters, sloths, and armadillos are provided by Rodrigues et al., Chiarello, and McDonough and Loughry
respectively. The general message in all these chapters
is that detailed, long-term studies of xenarthrans have
been rare. Just a few species have received the lions
share of attention, but even for these species our knowledge is often limited to just a few aspects of behavior
or ecology and to data from just a handful of locales.
An important task for the future will be to broaden
our coverage, not only in terms of the species and the
populations sampled, but also with regard to the types
of behavioral and ecological studies being conducted.
More detailed reports on the ecology and behavior
of particular species are provided in the succeeding
chapters. Meritt, Abba and Cassini, and Cullar provide
accounts of comparative ecology and behavior of xenarthrans (primarily armadillos) inhabiting specific areas within South America, and Superina reports results
from one of the first detailed field studies of Zaedyus pichiy. These accounts highlight the impressive diversity
of xenarthran adaptations to the environment and we
hope they will inspire a new generation of researchers
to pursue similar studies of additional species.
The future of xenarthran research

It should be apparent from the foregoing that a vibrant
community of scientists is exploring a diverse array of
questions regarding the Xenarthra. Even so, there are
still obvious holes in what we know. We have already
mentioned some of these above, and most of the chap-
ters in this book also point out where additional data

are needed. Thus, it is not our intent here to identify
every lingering question that needs answering. Indeed,
that is impossible. What we would like to do is at least
highlight some of the broad areas particularly deserving of future attention.
Taking field studies first, even though such studies
of xenarthrans are increasing, there are still many species that are virtually unknown to science. Obviously,
this makes conservation assessment and implementation of management plans extremely difficult. Further,
problems with maintaining xenarthrans in captivity
are undoubtedly linked with our poor understanding
of their natural ecology and behavior in the wild. Two
brief examples illustrate this point: first, during a workshop of the Edentate Specialist Group of the International Union for the Conservation of Nature designed
to update the conservation status of all species (Aguiar
and Fonseca this volume), no one present could report
ever having seen a juvenile giant armadillo (Priodontes
maximus) in the wild. It is amazing to think that this
could be the case with an animal so blatantly conspicuous. Nonetheless, without data on reproduction, any
assessment of current population status is little more
than guesswork. Second, consider D. pilosus, arguably the most intriguing species in the Dasypus genus.
It is the only member of the genus (1) found west of
the Andes (in Peru), (2) living at high altitudes, and
(3) possessing thick fur that covers the carapace. Despite these fascinating qualities, there has never been
any field study of this species. In fact, we barely even
know where to find it. These are only two cases, but we
could go on with similar stories for many other species.
Clearly, then, there are many opportunities for future
studies to make significant contributions.
An important component of any future field study
should be collection of samples for genetic screening. Genetic analyses are likely to play a critical role
in developing conservation plans for extant species.
In particular, we urgently need phylogeographic and
population genetics studies to identify evolutionarily
significant units that should be preserved. In addition,
although it appears many populations are in trouble, we
often do not know the reasons for this. For conservation measures to be effective, we will need better data
on the precise causes of population declines (cf. McDonough and Loughry 2005).
There are still other areas where more needs to be
done. While great strides have been made recently in
describing xenarthran phylogeny, controversies still remain, both in identifying relationships within the group
and in understanding the role of Xenarthra in mammalian evolution generally. A conspicuous example concerns the fossil anteater Eurotamandua from Europe.
Because of its unique and unexpected location, the
taxonomic status of this specimen has been debated for
some time. Current opinions based on molecular data
put it outside Xenarthra (Delsuc et al. 2001), but there
is no consensus on this point (Storch 2003). One reason
for the lingering controversy is that, to date, most molecular analyses have relied on relatively limited data,
from either a handful of genome markers such as microsatellites, or complete nucleotide sequences from just
a few genes. Despite the obvious successes with this approach, different markers or different genes sometimes
give conflicting pictures of evolutionary relationships
(Delsuc et al. 2002; Delsuc and Douzery this volume),
so they can be difficult to evaluate. Now that such data
are becoming available, it will be interesting to see what
will happen when the tools of comparative genomics
are employed.
Many phylogenetic hypotheses hinge on morphological analyses, but even here, in a discipline where
the conventional wisdom is that we learned everything
we needed to know during the heyday of comparative
anatomy in the 19th and 20th centuries, we still have a
lot to learn. For example, uncertainties still exist about
the homology of many xenarthran muscles and bone
pieces, as well as their teeth, in comparison with other
mammals. In another vein, elucidation of the biomechanical properties of different classes of xenarthran
dentine may help us to discern much about the diets
of the many fossil forms. Such analyses should provide
insights into the evolution of xenarthran physiology as
well as morphology. Similarly, we need additional insight into the physiology of living xenarthrans, where
we still have a very limited understanding of how many
body systems work. For example, the fact that we still
have so many problems attempting to breed xenarthrans in captivity suggests we still have much to learn
about xenarthran reproductive behavior and physiology.
Again, we do not mean for this discussion to be exhaustive. Rather than drag out this topic endlessly, we
will conclude by pointing out that, while much progress
has been made in our understanding of the Xenarthra,
a multitude of questions still remain, providing a wealth
of opportunities for a motivated student.
Summary
The Xenarthra remain an enigmatic group. This book
highlights the advances we have made in the study of
these remarkable animals, but, equally, it underscores
the many areas where we lack basic information. While
study of the Xenarthra may at times seem daunting,
we hope we have made a convincing case that devoting ones time and energy to this group can be intellectually rewarding. In our opinion, the best of times
lie ahead for xenarthran research; we hope this book
will inspire others to push ahead and rapidly make this
volume obsolete.
1
Phylogeny
2
Recent advances and future prospects in xenarthran
molecular phylogenetics
Frdric Delsuc and Emmanuel J. P. Douzery
se em carateres morfolgicos, citolgicos, imunolgicos e de protenas, mas as tentativas mais recentes beneficiaram-se da seqenciao de genes mitocondriais
e nucleares. Neste captulo, ns revisamos os avanos
recentes feitos na sistemtica molecular dos xenarthros.
Estes avanos conduziram reconstruo de uma filogenia de Xenarthra em nvel genrico, com o reconhecimento dos clados Pilosa, Folivora, Vermilingua, Cingulata, e Tolypeutinae + Euphractinae, mas deixando
incerta a posio de Xenarthra dentro dos mamferos
placentrios. Esta estrutura filogentica foi usada subseqentemente para definir uma escala temporal molecular para a histria evolutiva dos xenarthros existentes,
sugerindo a influncia potencial da mudana do clima
no Tercirio nesta ordem endmica da Amrica do Sul.
O trabalho filogentico futuro em Xenarthra inclui resolver seu relacionamento com os outros grupos placentrios usando dados genmicos e a reconstruo de
uma filogenia detalhada em nvel de espcie. Acoplado
com estudos filogeogrficos em nvel de populao, isto
permitir uma caracterizao adicional da diversidade
gentica desta peculiar ordem placentria e fornecer a
orientao aos planos de conservao para suas espcies ameaadas.
Resumen
En los ltimos 20 aos, la reconstruccin de las relaciones filogenticas de los xenartros ha sido revolucionada por datos moleculares. Las filogenias anteriores se
basaban en caracteres morfolgicos, citolgicos, inmunolgicos y de protenas, pero ensayos ms recientes se
beneficiaron con la secuenciacin de genes mitocondriales y nucleares. En este captulo repasamos los avances
recientes en sistemtica molecular de los xenartros. Estos avances han conducido a la reconstruccin de una
filogenia de Xenarthra a nivel genrico, con el reconocimiento de los clados Pilosa, Folivora, Vermilingua,
Cingulata y Tolypeutinae + Euphractinae, pero dejando
incierta la posicin de Xenarthra entre los mamferos
placentarios. Este marco filogentico se utiliz posteriormente para definir una escala temporal molecular
para la historia evolutiva de los xenartros vivientes, sugiriendo la potencial influencia de los cambios climticos durante el Terciario sobre este orden endmico de
Amrica del Sur. El trabajo filogentico futuro sobre
Xenarthra incluye la resolucin de su relacin con
otros grupos placentarios usando datos genmicos y la
reconstruccin de una filogenia comprehensiva a nivel
especfico. Junto con estudios filogeogrficos a nivel
poblacional, esto permitir la caracterizacin adicional
de la diversidad gentica de este peculiar orden de placentarios y orientar planes para la conservacin de sus
especies en peligro.
Introduction
Extant xenarthrans are currently represented by 31 living species of armadillos (Cingulata: Dasypodidae),
anteaters (Vermilingua: Myrmecophagidae and Cyclopedidae) and sloths (Folivora [also known as Phyllophaga or Tardigrada, see Faria and Vizcano 2003]:
Bradypodidae and Megalonychidae), classified in 13
(possibly 14, see Gardner 2005) genera and distributed
across the Americas, with most of the diversity centered
Resumo
Nos ltimos 20 anos a reconstruo dos relacionamentos filogenticos dos xenarthros foi revolucionada por
dados moleculares. As filogenias anteriores baseavam11
12
F. Delsuc and E. J. P. Douzery
in South America (Wetzel 1985a; Vizcano 1995; Anderson and Handley 2001). This quite modest taxonomic
diversity is in sharp contrast to that found in the fossil
record (Patterson and Pascual 1972; McKenna and Bell
1997). Living species are relicts from an evolutionary
radiation that occurred during the Tertiary isolation of
South America (Patterson and Pascual 1972). In fact,
the order was still quite diverse until the last mass extinction event just 10,000 years ago (Patterson and Pascual 1972; Lessa et al. 1997).
The use of molecular phylogenetics to reconstruct
the evolutionary relationships among the three main
xenarthran lineages dates back to the mid 1980s. The
first attempts were based on evolutionary comparisons
of protein sequences of -crystallin A (de Jong et al.
1985) and immunological distances derived from serum
albumins (Sarich 1985). These early studies marked the
dawn of the molecular era in xenarthran systematics
which, until that time, had been restricted to the study
of morphological and anatomical characters (Engelmann 1985). Since then, phylogenetic studies in xenarthrans (Delsuc et al. 2001, 2002, 2003) and, more generally, placental mammals (Madsen et al. 2001; Murphy,
Eizirik, Johnson et al. 2001; Murphy, Eizirik, OBrien et
al. 2001; Amrine-Madsen et al. 2003), have benefited
immensely from the sequencing of both mitochondrial
and nuclear genes, giving access to a large number of
phylogenetically useful characters. These studies allowed the reconstruction of a reliable phylogenetic
framework for extant xenarthran genera (except Chlamyphorus) and provided the basis for the definition
of a timescale for xenarthran evolution (Delsuc et al.
2004).
As we will review, modern studies, based on DNA
sequence analyses using state-of-the-art probabilistic
methods of phylogenetic reconstruction, have yielded
numerous new insights into xenarthran systematics
(Delsuc et al. 2001, 2002, 2003). They have also confirmed predictions of earlier molecular analyses, for
example, by corroborating the independent evolution
of xenarthrans and pangolins (de Jong et al. 1985, 1993;
Sarich 1985). Additionally, we outline future prospects
in the molecular phylogenetics of xenarthrans, among
which are the resolution of their place in the placental
tree using data from comparative genomics, and the
construction of a comprehensive species-level phylogeny, which will be critical in elaborating conservation
programs for the most threatened species.
The central place of Xenarthra within placental

mammals
From edentates to xenarthrans
Xenarthrans have long been of special interest to researchers involved in understanding the evolutionary
origins and relationships of placental mammals. Despite their highly distinctive morphologies, armadillos,
anteaters, and sloths have been recognized to form a
monophyletic group on the basis of shared derived characters such as atypical articulations between the vertebrae (Engelmann 1985; Gaudin 1999b), from which the
order name was derived (xenos = strange, and arthros
= articulation in Greek). Xenarthrans were grouped
with pangolins and aardvarks in early classifications
into a group called edentates (Edentata, see Glass 1985).
Aardvarks and pangolins were subsequently placed into
their own orders, Tubulidentata and Pholidota, respectively. However, the name Edentata was retained for
the superordinal grouping of Xenarthra with Pholidota
(McKenna 1975; Novacek and Wyss 1986; Novacek
1992). Although abandoned by Simpson (1945), the hypothesis of a close relationship between Xenarthra and
Pholidota was so popular among morphologists that it
was resurrected by Novacek and Wyss (1986) despite
contradictory anatomical evidence (Bugge 1979).
The first molecular studies, using immunological
and protein-based characters, clearly separated xenarthrans from pholidotes (de Jong et al. 1985; Sarich 1985;
Shoshani 1986). These results forced morphologists to
reconsider the evidence for Edentata, and it was concluded that support for such a relationship was actually
very weak (Rose and Emry 1993). Further results from
the analysis of -crystallin A protein sequences led de
Jong et al. (1993) to first propose a possible sister-group
relationship between Pholidota and Carnivora, a hypothesis that has been adopted by some morphologists
(Shoshani and McKenna 1998). The lack of evolutionary affinities and the independent evolution of Xenarthra and Pholidota has since been confirmed in numerous phylogenetic studies (van Dijk et al. 1999; Delsuc
et al. 2001, 2002; Madsen et al. 2001; Murphy, Eizirik,
Johnson et al. 2001; Murphy, Eizirik, OBrien et al. 2001;
Lin et al. 2002; Amrine-Madsen et al. 2003; Hudelot et
al. 2003; Reyes et al. 2004); some of these also found
strong support for grouping Pholidota with Carnivora
(Murphy, Eizirik, OBrien et al. 2001; Delsuc et al. 2002;
Amrine-Madsen et al. 2003). As a consequence, the
Xenarthran molecular phylogenetics
morphological similarities between xenarthrans and

pangolins, including the trend toward dental reduction
(Ferigolo 1985), are now generally considered adaptive
convergences associated with their fossorial and myrmecophagous habits (Rose et al. 2005).
Xenarthra and Epitheria
The phylogenetic relationships of armadillos, anteaters, and sloths to other placental mammals have been
of special interest for quite some time because of their
seemingly archaic morphology. Indeed, retention of
anatomical and physiological characters thought to be
plesiomorphic for placental mammals, such as a low and
poorly regulated body temperature, a lack of clear differentiation between uterus and vagina, and the intraabdominal position of the testes, has led to their being
considered an early offshoot of the placental mammal
radiation (Gregory 1910). McKenna (1975) was the first
to propose that Xenarthra represents the sister group
to all other eutherians, which were collectively named
Epitheria. This view has been widespread among morphologists (Novacek and Wyss 1986; Novacek 1992) and
was adopted in the most recent morphologically based
classification of mammals (McKenna and Bell 1997).
However, morphological synapomorphies defining epitherians are actually weak and their phylogenetic distribution among placentals is equivocal (Gaudin et al.
1996). Shoshani and McKenna (1998) summarized the
morphological view of placental radiation with the recognition of 18 orders whose interrelationships remain
largely hypothetical, except the grouping of rodents
and lagomorphs into Glires and a monophyletic origin
of elephants, hyraxes, and sirenians (Paenungulata). In
this synthetic tree, xenarthrans constitute the earliest
diverging branch, in agreement with the Epitheria hypothesis, even though they are separated from Pholidota based on available molecular evidence (Shoshani
and McKenna 1998).
The molecular revolution
The morphological view of placental phylogeny was
shattered in 2001 with the simultaneous publication of
two independent studies based on phylogenetic analyses
of multiple genes (Madsen et al. 2001; Murphy, Eizirik,
Johnson et al. 2001). These studies identified four major placental clades: Afrotheria (aardvarks, elephantshrews, golden-moles, tenrecs, hyraxes, elephants,
and sirenians), Xenarthra (armadillos, anteaters, and
13
sloths), Euarchontoglires (tree-shrews, flying-lemurs,

primates, rodents, and lagomorphs), and Laurasiatheria
(eulipotyphlans, bats, pangolins, carnivores, perissodactyls, and cetartiodactyls), of which only Xenarthra
had been previously recognized by morphological studies. The subsequent combination of these two datasets
(Murphy, Eizirik, OBrien et al. 2001) also supported
the grouping of Euarchontoglires and Laurasiatheria.
This superclade was named Boreoeutheria to reflect its
Northern Hemisphere origin, as opposed to Afrotheria
and Xenarthra, which both originated in the Southern
Hemisphere (Springer and de Jong 2001). The recognition of such biogeographical clades suggests that plate
tectonics may have played an important role in shaping
the early stages of placental diversification (Murphy,
Eizirik, OBrien et al. 2001). Additional phylogenetic
analyses, mainly of nuclear genes (Delsuc et al. 2002;
Amrine-Madsen et al. 2003; Waddell and Shelley 2003),
mitochondrial RNAs (Hudelot et al. 2003), and recent
studies of complete mitochondrial genomes (Lin et al.
2002; Reyes et al. 2004), all support recognition of the
four major placental clades. These studies have also
revealed the extent of morphological convergence, a
problem that has blurred any phylogenetic signal based
on morphology (see Springer et al. 2004 for a recent
review).
The question of the placental root
That xenarthrans represent one of the four major placental lineages underlines the evolutionary significance of this often neglected group. However, none
of the multigene studies (Madsen et al. 2001; Murphy,
Eizirik, Johnson et al. 2001; Murphy, Eizirik, OBrien et
al. 2001; Delsuc et al. 2002; Amrine-Madsen et al. 2003)
found statistically significant support for locating Xenarthra within placentals with confidence. Indeed, using
maximum likelihood (ML) and Bayesian analyses, all
the studies favored a basal position of Afrotheria, with
Xenarthra as a sister group to Boreoeutheria. However,
these statistical tests have been influenced by both character and taxon sampling (Delsuc et al. 2002; Holland
et al. 2005). Thus, it has proved difficult to distinguish
between three competing topologies for the position of
the placental root: (1) basal Afrotheria, (2) basal Xenarthra (= Epitheria), or (3) Afrotheria + Xenarthra.
The difficulty of the question is illustrated by the
results reported in table 2.1. This information was obtained after new analyses of the multigene dataset of
14
Amrine-Madsen et al. (2003), which consists of 17,736

unambiguously aligned nucleotide sites for 42 placental
taxa and two marsupial outgroups. The results of likelihood-based SH tests (Shimodaira and Hasegawa 1999),
performed for different partitions of the complete dataset, show that the Epitheria and the Afrotheria + Xenarthra hypotheses cannot be rejected, although each
partition gives the highest likelihood score for the basal
Afrotheria hypothesis (table 2.1). Moreover, as shown
previously (Delsuc et al. 2002), character sampling influences the results, with the three alternatives becoming
almost indistinguishable in terms of likelihood scores
when only the first two codon positions of nuclear
genes are used. In this case, the Afrotheria + Xenarthra
hypothesis, which makes more biogeographical sense
because it divides placentals into Northern and Southern Hemisphere clades (Waddell et al. 1999), differs
from the highest likelihood topology only at the decimal level, whereas the Epitheria hypothesis is slightly
less likely (table 2.1). While the nuclear first and second
codon position partition can be considered to contain
more reliable sites in terms of mutational saturation
(Delsuc et al. 2002), the question of the position of the
root and thereby the place of xenarthrans within placentals is left unresolved by the data currently available.
The future application of refined methods dedicated to
the identification of the most reliable sites for phylogenetic inference (Brinkmann and Philippe 1999; Burleigh and Mathews 2004; Pisani 2004) might help to
further evaluate the three competing alternatives.
The molecular phylogeny of living xenarthrans

Phylogenetic relationships among xenarthrans have
been investigated using both mitochondrial and nuclear genes. The first study considered a combination
of mitochondrial 12S and 16S ribosomal RNA (rRNA)
genes and nuclear exon 28 of the von Willebrand Factor
(VWF) for eight of the 13 living xenarthran genera (Delsuc et al. 2001). Subsequently, taxon sampling was increased to 12 genera, with only Chlamyphorus missing,
in analyses of three genetically independent proteincoding nuclear genes: the intronless -2B Adrenergic
receptor gene (ADRA2B), exon 11 of the Breast Cancer
Susceptibility gene (BRCA1), and, once again, the VWF
gene (Delsuc et al. 2002). Gene sampling in this latter
study was later expanded by the addition of two mitochondrial genes: 12S rRNA and NADH dehydrogenase
1 (ND1), representing a total of 6,968 nucleotide sites
(Delsuc et al. 2003). These three studies allowed the re-
construction of a comprehensive phylogeny of extant

xenarthrans at the genus level with only two remaining
uncertainties, both within armadillos.
The monophyly of the order
The monophyly of Xenarthra is well defined morphologically by characters generally thought to reflect
adaptation toward fossoriality and myrmecophagy
(Gaudin 1999b; see also Gaudin and McDonald this
volume). The common ancestry of extant xenarthrans
was suggested by early molecular studies (de Jong et al.
1985; Sarich 1985) and has been retrieved with strong
statistical support in all subsequent sequence-based
phylogenetic studies (Delsuc et al. 2001, 2002; Madsen
et al. 2001; Murphy, Eizirik, Johnson et al. 2001). Furthermore, these results are also supported by the occurrence of a rare three amino acid deletion in the -crystallin A chain, offering a diagnostic molecular signature
for Xenarthra (van Dijk et al. 1999).
Xenarthra intraordinal relationships
The respective monophyly of armadillos (Cingulata),
anteaters (Vermilingua), and sloths (Folivora) is also
well defined morphologically (Engelmann 1985; Gaudin
and McDonald this volume). In addition, unambiguous
support for a monophyletic origin of each xenarthran
lineage has been found in our molecular studies (Delsuc et al. 2002, 2003).
The interrelationships of the three xenarthran lineages were a matter of conflict among morphologists
(compare Guth 1961; Bugge 1979 with Engelmann 1985;
Patterson et al. 1992). The most recent classification
(McKenna and Bell 1997) groups anteaters and sloths
into a clade called Pilosa, which refers to their coat.
Early molecular studies did not contribute much to
this debate (de Jong et al. 1985; Sarich 1985). Our molecular data (Delsuc et al. 2001, 2002, 2003), which included samples from all anteater and sloth genera, provided strong support for Pilosa, extending the results
of Madsen et al. (2001) and Murphy, Eizirik, Johnson
et al. (2001), which included fewer taxa. These results
contradicted studies of the ear region (Guth 1961) and
cephalic arterial patterns (Bugge 1979) that favored the
early emergence of anteaters within Xenarthra. It seems
likely that these studies (Guth 1961; Bugge 1979) were
misled by the extreme specialization of the skull toward
myrmecophagy in anteaters. Indeed, subsequent cladistic studies of characters from the ear region (Patterson
et al. 1992) and other morphological and anatomical
features (Engelmann 1985) have provided synapomor-
23,0541.92
0.13
= 8.42
{image 3.eps}

19,2906.97
0.18
= 5.64
19,2906.53
0.21
= 5.20
Best
9,4239.90
= 1.55
9,4238.56
= 0.21
0.44
0.62
Best
Notes: Results are from SH statistical tests (Shimodaira and Hasegawa 1999) for the position of the placental root using different partitions of the multiple gene dataset of Amrine-Madsen et al. (2003). These
tests were computed with PAUP* 4.0b10 (Swofford 2002) under a concatenated GTR+8+I model with parameters estimated for each alternative topology. The highest log-likelihood value is shown in brackets for each dataset and the log-likelihood difference () relative to the best topology is given. The last hypothesis corresponds to Epitheria. pSH = probability of the SH test. Triangles in trees are drawn proportional to taxon diversity and branch length of the corresponding clade. Abbreviations: OUT = Marsupial outgroup, AFR = Afrotheria, XEN = Xenarthra, EUA = Euarchontoglires, LAU = Laurasiatheria.
21,1797.82
0.18
= 6.78
21,1797.86
0.16
= 6.82
[19,2901.33]
[9,4238.35]
23,0540.43
0.20
= 6.93
Best
[21,1791.04]
{image 2.eps}

Best
pSH
[23,0533.50]
pSH
{image 1.eps}
-lnL
-lnL
pSH
-lnL
-lnL
pSH
Positions 1+2 of
nuclear coding genes
(9,551 sites)

All genes
Nuclear genes
Nuclear coding genes
Competing topologies
(17,736 sites)
(16,089 sites)
(14,327 sites)
Table 2.1. Where Do Xenarthrans Fit among Placentals?
16
phies for Pilosa, such as the interruption of the zygomatic arch and the intra-pelvian location of the testes.
Phylogeny of anteaters (Vermilingua)
The classical arrangement of Vermilingua groups the
giant anteater (Myrmecophaga) with lesser anteaters
(Tamandua) to the exclusion of the pygmy anteater (Cyclopes), which is considered morphologically divergent
from the others (Engelmann 1985; Reiss 1997; Gaudin
and Branham 1998). Molecular results confirmed this
view by favoring the early emergence of the pygmy anteater (Delsuc et al. 2001, 2002, 2003; Barros et al. 2003),
a phylogenetic hypothesis also supported by cranial
muscle (Reiss 1997) and morphological (Gaudin and
Branham 1998) characters.
Phylogeny of sloths (Folivora)
The two living genera of three-toed (Bradypus) and
two-toed (Choloepus) sloths are unknown as fossils
and have been classified into distinct families (respectively Bradypodidae and Megalonychidae) on the basis of their numerous morphological differences and a
presumably diphyletic origin from two different fossil
lineages (Patterson and Pascual 1972; Webb 1985a). This
taxonomic distinction was supported by early immunological data demonstrating considerable evolutionary distance between the albumins of the two genera
(Sarich 1985). The diphyly hypothesis also found some
support from ancient DNA studies that sequenced mitochondrial 12S and 16S rRNA fragments from fossil
sloths (Hss et al. 1996; Greenwood et al. 2001). These
studies indicated that modern two-toed sloths (Megalonychidae) are closely related to the giant ground
sloth Mylodon darwini (Mylodontidae), whereas threetoed sloths (Bradypodidae) appear closer to the Shasta
ground sloth Nothrotheriops shastensis (Megatheriidae). However, these results have been contradicted
by the inclusion of additional modern sloth species in
phylogenetic analyses of partial 16S sequences (Barros
et al. 2003). Here, moderate support was obtained for
grouping Mylodon darwini with three-toed sloths instead of two-toed sloths. The sequencing of complete
mitochondrial genomes from fossil sloths might help
to resolve the controversy (H. Poinar personal communication).
The 16S rRNA study (Barros et al. 2003) was the
first to include three of the four living species from
the genus Bradypus. These authors found evidence of
a sister-group relationship between the pale-throated
(B. tridactylus) and brown-throated (B. variegatus)
sloths to the exclusion of the endangered maned sloth

(B. torquatus). The one species for which no molecular data are currently available is the newly described
dwarf three-toed sloth (B. pygmaeus), endemic to the
small island of Bocas del Toro in Panama (Anderson
and Handley 2001). Acquisition of molecular data from
this species will be important in deciding whether it
deserves species status or represents a morphologically
divergent population of B. variegatus.
There are only two living species of two-toed sloths
currently recognized: the southern two-toed sloth
(Choloepus didactylus) and Hoffmmans two-toed sloth
(Choloepus hoffmanni). However, given the large differences in chromosome number reported between
specimens within the genus (Jorge et al. 1985a; Jorge
and Pereira this volume), future molecular data may be
important in revealing cryptic species.
Phylogeny of armadillos (Cingulata)
With 21 living species classified into 8 genera (Wetzel
1985a; Vizcano 1995), armadillos (Cingulata, Dasypodidae) are the most speciose xenarthran lineage.
Taxonomically, the family Dasypodidae is generally divided into five tribes: Dasypodini, Euphractini, Tolypeutini, Priodontini, and Chlamyphorini. These appear
well defined morphologically (Wetzel 1985a; McKenna
and Bell 1997) and are consistent with the cytological
structure of their spermatozoa (Cetica et al. 1998; Cetica and Merani this volume). The tribe Dasypodini
is composed of seven species of long-nosed armadillos classified in the single genus Dasypus (Wetzel and
Mondolfi 1979; Vizcano 1995). The three genera of
hairy armadillosChaetophractus, Euphractus, and Zaedyus (Euphractini)are very similar morphologically
and ecologically, and their interrelationships have been
difficult to decipher on solely morphological grounds
(Engelmann 1985; Gaudin and Wible 2006). The two
species of three-banded armadillos (Tolypeutes) are the
only members of the tribe Tolypeutini and are famous
for being able to roll entirely into a ball thanks to the
articulation of their carapace (Wetzel 1985a). The giant armadillo (Priodontes maximus) and naked-tailed
armadillos (genus Cabassous) are traditionally united
within the tribe Priodontini (Engelman 1985; Wetzel 1985a). One feature uniting them is their unusual
spoon-shaped spermatozoa, which are among the largest found in mammals (Cetica et al. 1998; Cetica and
Merani this volume). Finally, the tribe Chlamyphorini
contains two species of fairy armadillos (genus Chlamyphorus; Gardner [2005] has argued that one of these
17
Figure 2.1. Illustration of the two remaining uncertainties in armadillo molecular phylogeny. Eight armadillo species (Delsuc
et al. 2003) were analyzed using mitochondrial (A), nuclear (B), and their combination (C) data. A consensus network was
computed (Holland et al. 2005) from the 100 maximum likelihood bootstrap trees obtained under the GTR+8 model using
PAUP* 4.0b10 (Swofford 2002), with a threshold of 10% in SplitsTree 4.1 (Huson and Bryant 2006). The consensus networks
therefore represent all splits that appear in more than 10 of the 100 bootstrap trees, with edge lengths corresponding to
bootstrap percentages.
species, Chl. retusus, should be placed in its own genus,

Calyptophractus, which would increase the total number of armadillo genera to 9). The extensively subterranean lifestyle of these cryptic animals renders the study
of their biology particularly difficult (Meritt 1985c).
The phylogeny of living armadillos was poorly studied for many years, with only one cladistic analysis of
morphological characters in fossil and extant taxa (Engelmann 1985). However, a recent craniodental analysis
has allowed testing Engelmanns hypotheses in a cladistic
context and revealed numerous incongruities (Gaudin
and Wible 2006). Our molecular studies (Delsuc et al.
2001, 2002, 2003) clearly identified three main lineages
of armadillos corresponding to the subfamilies defined
in the classification of McKenna and Bell (1997): Dasypodinae (Dasypus), Euphractinae (Chaetophractus, Euphractus, and Zaedyus) and Tolypeutinae (Priodontes,
Cabassous, and Tolypeutes). The early emergence of

Dasypodinae was also strongly supported, with Tolypeutinae and Euphractinae unequivocally clustering together. Such a relationship is congruent with the study
of spermatozoa (Cetica et al. 1998) but contradicts the
morphological studies of both Engelmann (1985) and
Gaudin and Wible (2006).
Molecular data failed to resolve relationships within
the subfamilies Tolypeutinae and Euphractinae, with
contradictory results obtained from mitochondrial
versus nuclear genes, which in turn led to poorly supported relationships in ML analyses when the data were
combined (Delsuc et al. 2003). The presence of contradictory signals in molecular data is illustrated in figure
2.1, which depicts phylogenies generated by the use of
consensus networks (Holland and Moulton 2003). This
method allows graphical representation of the uncer-
18
tainty in phylogenetic relationships by displaying alternative hypotheses in the form of a network, where edge
lengths are proportional to bootstrap support (Holland
et al. 2005). The consensus network obtained from ML
bootstrap analysis of two combined mitochondrial
genes (12S rRNA and ND1) displays a three-dimensional box for the relationships within Tolypeutinae,
and a two-dimensional cycle for the relationships within
Euphractinae (figure 2.1A). The three-dimensional box
in Tolypeutinae indicates that the three possible alternatives have at least some support, with Cabassous +
Priodontes as the most likely hypothesis, followed by
Cabassous + Tolypeutes, with Tolypeutes + Priodontes
being only marginally supported. Within Euphractinae, the two-dimensional cycle shows that there are
only two competing alternatives, with Chaetophractus
+ Euphractus being slightly favored over Chaetophractus + Zaedyus. The same network structure is observed
for the combination of three nuclear genes (ADRA2B,
BRCA1, and VWF) except that the hierarchy of alternative arrangements is different (figure 2.1B). Within
Tolypeutinae, Cabassous + Tolypeutes becomes the
favorite hypothesis, then Cabassous + Priodontes, but
Tolypeutes + Priodontes still ranks as the least well-supported. For Euphractinae, nuclear data favor Chaetophractus + Euphractus over Chaetophractus + Zaedyus.
Combining all five genes results in a consensus network
with two rectangular cycles describing the relationships
within the two subfamilies (figure 2.1C). Two competing
hypotheses remain for each subfamily, with Cabassous
+ Priodontes being favored over Cabassous + Tolypeutes
within Tolypeutinae, and Chaetophractus + Euphractus
being preferred over Chaetophractus + Zaedyus within
Euphractinae (figure 2.1C).
Within tolypeutines, a close relationship between
Cabassous and Priodontes would be consistent with
their classification in the tribe Priodontini, if their very
similar morphologies, spermatozoa (Cetica et al. 1998;
Cetica and Merani this volume), and characters related
to their fossorial habits, such as their enlarged manus
claws (Engelmann 1985; Wetzel 1985a; McKenna and
Bell 1997), are interpreted as synapomorphies rather
than symplesiomorphies. Concerning euphractines, the
grouping of Euphractus + Chaetophractus to the exclusion of Zaedyus is congruent with study of craniodental
characters (Gaudin and Wible 2006). Given the independent support coming from morphological studies,
it is tempting to consider these two schemes as the best
current phylogenetic hypotheses for the relationships
within the subfamilies Tolypeutinae and Euphractinae.

It is hoped that sampling of additional genes with different selective constraints, such as nuclear introns, will
confirm our predictions.
From a more general viewpoint, even if the global
molecular picture of armadillo phylogenetic relationships (Delsuc et al. 2003) remains somewhat incongruent with the most recent cladistic analysis (Gaudin and
Wible 2006), it is interesting to note that the two phylogenetic hypotheses are actually quite close in terms
of parsimony scores (Gaudin and Wible 2006). In fact,
the molecular tree appears far more compatible with
the craniodental evidence than all other morphologically-based hypotheses (see Gaudin and Wible 2006).
Molecular and morphological phylogenies of Cingulata
are thus closer to agreement than ever.
A molecular timescale for xenarthran evolution

The newly established phylogenetic framework for
extant xenarthrans was subsequently used to derive
a molecular timescale for their evolutionary history.
Because pervasive among-lineages substitution rate
variations were recorded for the genes compared, a
Bayesian method that relaxed strict molecular clock assumptions (Thorne et al. 1998) and allowed for amonggene rate heterogeneity (Thorne and Kishino 2002) was
employed (Delsuc et al. 2004). This method also had
the advantage of considering time intervals defined by
the paleontological record instead of fixed calibration
points (Kishino et al. 2001). Based on the study of three
nuclear genes (ADRA2B, BRCA1, and VWF) the age
of the xenarthran crown group was estimated at 65
5 million years (myr), close to the Cretaceous/Tertiary
boundary (figure 2.2). This is fully compatible with the
age obtained by Springer et al. (2003) with a similar set
of nuclear exon characters and using the same calibration points, but is more recent than previous estimates
suggesting a date around 80 myr (Sarich 1985; Hss et
al. 1996). This younger estimate makes Xenarthra the
major placental clade with the most recently diverged
extant lineages, but also with the deepest stem lineage.
The gap observed between the first occurrence of fossil xenarthrans in the late Paleocene of Brazil (ca. 58
myr ago [mya]), in the form of the earliest armadillo
scutes (Scillato-Yan 1976b; Oliveira and Bergqvist
1998; Bergqvist et al. 2004), and the molecular estimate
of their purported origin around 105 mya (Springer et
al. 2003; Delsuc et al. 2004), suggests the existence of
19
Figure 2.2. Phylogenetic relationships and molecular timescale for extant xenarthran genera based on analyses of three
nuclear genes and a Bayesian relaxed molecular clock (modified from Delsuc et al. 2004). The time scale is given in million
years. The mean age estimate SD is given for all nodes. Horizontal rectangles depict the uncertainty of age estimates based
on 95% confidence intervals. Note that the relationships within both Tolypeutinae (Tolypeutes, Cabassous, and Priodontes) and
Euphractinae (Zaedyus, Euphractus, and Chaetophractus) are left as unresolved to reflect the current phylogenetic uncertainty
about these nodes. Vertical lines demarcate geological periods. Abbreviations: E = Early, M = Middle, L = Late, Pli. = Pliocene,
P. = Pleistocene, D. = Dasypus.
an ancestral ghost lineage for almost 50 myr. This inference concurs with the view that the origin of xenarthrans constitutes a paleontological and biogeographic
enigma (McKenna 1975; Engelmann 1985) that only the
discovery of new fossils might help to resolve.
Results about the timing of xenarthran diversification estimated the early split between anteaters and
sloths at the transition between the Paleocene and
Eocene some 55 5 mya (figure 2.2). Within anteaters, molecular dating emphasized the antiquity of the
pygmy anteater (Cyclopes) lineage, estimated as emerging in the middle Eocene around 40 4 mya, relative
to the Tamandua and Myrmecophaga lineages, which
diverged only 10 2 mya. Also, the considerable divergence between two-toed (Choloepus) and three-toed
(Bradypus) sloths was confirmed, with an estimate of
their separation at around 21 3 mya. This result gives
further credit to their taxonomic placement in two
distinct families, as described above. Among armadillos, the early emergence of Dasypus was estimated to
have occurred during the middle Eocene around 40
5 mya, followed by the split between Tolypeutinae and
Euphractinae at about 33 4 mya (figure 2.2). Diversification in the latter two subfamilies happened relatively
quickly, but in markedly different epochs: the diversification of tolypeutines appeared quite ancient (ca. 22
3 mya), whereas euphractines diversified much more
recently (ca. 6 1 mya). These results reveal that the
family Dasypodidae contains lineages of fairly ancient
origin, as might be expected given their distinctive
morphologies (Wetzel 1985a) and marked structural
differences in spermatozoa (Cetica et al. 1998; Cetica
and Merani this volume).
Finally, the molecular estimates of xenarthran divergence dates were correlated with the relatively well-documented paleoenvironmental changes that occurred
20
during the Tertiary of South America (Patterson and

Pascual 1972; Pascual and Ortiz Jaureguizar 1990; Marshall and Sempere 1993). This allowed us to unravel the
potentially important role played by paleoenvironmental changes in the diversification of living xenarthrans
(Delsuc et al. 2004). Indeed, molecular dating revealed
a striking synchronicity in some diversification events
among independent xenarthran lineages. For instance,
the separation of Cyclopes from other anteaters was correlated with the separation of the Dasypus lineage from
other armadillos in the middle Eocene around 40 mya
(figure 2.2). Similarly, the diversification of Tolypeutinae paralleled the separation between the two modern
sloth genera around 2122 mya in the early Miocene.
Finally, the recent diversification of Euphractinae
matched well with the separation between Dasypus
novemcinctus and Dasypus kappleri some 67 mya in
the late Miocene (figure 2.2). Although we cannot rule
out the possibility that such a correlated history between independent lineages occurred by chance, it is
more likely these biological events reflect the impact of
environmental changes. Actually, all three synchronous
diversification events appear to follow periods of important environmental changes, possibly triggered by
major phases of Andean uplift (Marshall and Sempere
1993). The evolutionary history of extant xenarthrans
therefore seems to have been influenced by the environmental changes that occurred during the Tertiary
of South America.
Future prospects in xenarthran molecular

phylogenetics
The promises of comparative genomics
As described above, molecular phylogenetic studies
based on multiple genes currently leave the position
of Xenarthra within placental mammals unresolved,
with three alternative positions for the root being almost equally likely (see table 2.1). The solution to this
difficult problem, as with many other uncertainties in
the tree of life, might come from future phylogenomic
studies taking advantage of the wealth of genomic data
generated by ongoing large-scale sequencing projects
(Delsuc et al. 2005). Indeed, genomes of 27 mammalian
species, including representatives from all the major
placental clades, are currently being sequenced at varying levels of coverage, an effort offering great promise
for mammalian phylogenomics (Murphy et al. 2004).
Among these species, of particular importance are the

platypus (Ornithorhynchus anatinus) and two marsupials (the Tammar wallaby, Macropus eugenii, and the
gray short-tailed opossum, Monodelphis domestica)
that can be used as outgroups in both comparative genomic and phylogenetic studies. Within Xenarthra, two
speciesthe nine-banded armadillo (Dasypus novemcinctus) and Hoffmans two-toed sloth (Choloepus hoffmanni)have been or will be sequenced (see Chang
and Adams this volume).
Genomic data have the potential to provide answers
to many remaining questions in placental mammal
phylogeny, most importantly the position of the root,
and thereby the relationships among Afrotheria, Xenarthra, and Boreoeutheria. Indeed, the wealth of data
accumulated by the phylogenomic approach offers the
luxury of selecting only the most reliable sites for classical sequence-based phylogenetic analyses (Delsuc et
al. 2005). In the context of mammalian phylogenomics, consideration of the platypus as an additional outgroup might prove particularly useful for rooting the
placental tree, because it might break the problematic
long branch of marsupials, as was previously demonstrated with complete mitochondrial genomes (Cao et
al. 1997; Philippe 1997). Moreover, access to complete
genomes affords the opportunity to mine them for signatures of common ancestry (Delsuc et al. 2005). Useful signatures include rare genomic changes such as
diagnostic insertion/deletion events, intron positions,
transposable element (SINEs and LINEs) integrations,
gene fission/fusion events, and evidence of shared gene
families or chemical pathways. These whole-genome
features can be used as an independent source for corroboration of sequence-based phylogenies (Delsuc et al.
2005). Of special interest for resolving the relationships
of Afrotheria and Xenarthra is the study of SINEs integration: specific families have been identified in both
groups (Nikaido et al. 2003; Churakov et al. 2005) and
successfully used at the intraordinal level in placental
phylogenetics (Nikaido et al. 1999; Schmitz and Zischler 2003; Nishihara et al. 2005; Schmitz et al. 2005).
Such an approach was recently applied by Kriegs et al.
(2006), who identified two retroposon markers found
to be absent in Xenarthra but present in all other placental Afrotheria and Boreoeutheria, providing some
support for the Epitheria hypothesis. However, conclusions are tentative because evidence that the markers
are also missing in marsupials is still uncertain. The
sequencing of marsupial and monotreme complete
genomes will soon provide the opportunity to check

whether these elements are really absent in placental
outgroups.
Toward a phylogeny of living xenarthran species
Ultimately, continued efforts on the molecular front
should result in a well-resolved phylogeny of extant
xenarthrans at the species level. This will provide answers to evolutionary questions, some of which we
highlight below.
The evolutionary affinities of fairy armadillos
At present, the only lineage of armadillos for which
there are no molecular data is the tribe Chlamyphorini
(fairy armadillos or pichiciegos), which includes the
single genus Chlamyphorus Harlan 1825 (but see Gardner 2005). Two species, Chl. truncatus (pink fairy armadillo) and Chl. retusus (greater fairy armadillo),
are traditionally recognized. They are rather similar
morphologically in being highly adapted to burrowing,
with enlarged digging claws and reduced eyes. However, marked differences between the two species exist, most notably in the structure of the carapace and
the shape of the cephalic shield. To reflect the degree of
morphological differentiation between these two species, the larger species, Chl. retusus, was assigned to its
own genus, Burmeisteria Gray, 1865 or Calyptophractus Fitzinger, 1871. Moeller (1968) proposed retaining
Burmeisteria (however, this name is preoccupied, so
the appropriate choice should have been Calyptophractus, see Gardner 2005). In contrast, cladistic analyses
of morphological characters indicate the two species
of fairy armadillos are monophyletic and phylogenetically related to members of the family Euphractinae
(Engelmann 1985; Gaudin and Wible 2006; Gaudin
and McDonald this volume). However, as it has never
been molecularly tested, the possibility of a diphyletic
origin of the two species cannot be rejected a priori.
Indeed, based on characters like the shape and color
pattern of the carapace, as well as the shape of the anterior claws, Chl. retusus resembles naked-tailed armadillos from the tribe Priodontini, whereas Chl. truncatus
seems to be closer to Euphractini. These observations
leave open the possibility that the morphological similarities between fairy armadillos might be the result of
convergence due to extreme selective pressures induced
by their subterranean lifestyle. Given their rarity and
the difficulties in obtaining biological material from
these species, molecular analyses of fairy armadillos
21
may have to exploit museum specimens, using ancient DNA techniques. This might limit the potential
molecular markers to mitochondrial genes, which are
much easier to amplify in this context. The ND1 and 12S
rRNA genes, for which data from all other armadillo
genera are already available (Delsuc et al. 2003), seem
therefore to represent the best candidates. Perhaps with
these data we will finally be able to resolve the phylogenetic affinities of fairy armadillos and thus better understand the evolution of morphological characters in
these enigmatic animals.
Tracking the origin of polyembryony in long-nosed armadillos
One of the most fascinating features observed in xenarthrans is the occurrence of monozygotic polyembryonythe splitting of one sexually produced embryo
into manyin long-nosed armadillos (genus Dasypus,
see the chapters by Enders, Prodhl et al., and McDonough and Loughry this volume). In these species,
the origin and evolution of polyembryony is generally
interpreted as a response to the phylogenetic constraint
represented by an unusual uterine shape with only one
implantation site (Galbreath 1985; but see Enders this
volume). Galbreath proposed an evolutionary scenario
in which the specialized uterus evolved first, thus preadapting for monozygotic polyembryony any species
that underwent selection for increasing litter size. Testing this evolutionary hypothesis requires, first, molecular confirmation of the occurrence of polyembryony
in each of the seven Dasypus species (cf. Prodhl et al.
this volume) and, second, reconstruction of their phylogenetic relationships. Such analyses are a prerequisite
for understanding the evolution of twin number (i.e.,
litter size) and the structure and organization of the reproductive tract (Enders this volume). The reconstruction of a species-level phylogeny for the genus Dasypus
would therefore provide new insights into the origin
and evolution of polyembryony.
Modern cytogenetics and molecular phylogenetics
Cytological and karyological studies in xenarthrans
have long been restricted to classical studies in which
karyotypes were described using techniques such as
G-banding (Jorge et al. 1985a; Jorge and Pereira this
volume). However, molecular phylogenetics is not the
only discipline where advances in technology have led
to renewed interest. Indeed, cytogenetics is currently
experiencing a rebirth thanks to the development of
22
new techniques such as chromosome painting (FISH:

fluorescence in situ hybridization) and radiation hybrid
mapping, both of which allow the fine-tuned study of
chromosomal evolution (Murphy, Stanyon, and OBrien
2001). These techniques were recently applied successfully to characterize the dynamics of chromosome
evolution in placental mammals as a first step toward
reconstructing their ancestral karyotype (Murphy et al.
2005).
Unfortunately, data from xenarthrans, which are
pivotal for inferring the ancestral placental karyotype,
were missing until recently. A first attempt at applying
these new techniques to xenarthran cytogenetics was
recently made using chromosome painting comparisons among the three major lineages. Data came from
a two-toed (Choloepus didactylus, 2n = 64) and threetoed sloth (Bradypus tridactylus, 2n = 52), the lesser
anteater (Tamandua tetradactyla, 2n = 54), and the
six-banded armadillo (Euphractus sexcinctus, 2n = 58,
Dobigny et al. 2005). By using the phylogenetic framework and timescale derived from previous molecular
studies (Delsuc et al. 2003, 2004) to map inferred chromosomal changes, these authors revealed a low rate of
genomic repatterning in Xenarthra relative to other
placentals (Dobigny et al. 2005). Moreover, by identifying homologous chromosomal segments that have been
conserved among members of the three lineages, this
study provides clues to the likely architecture of the ancestral karyotype for extant xenarthrans (Dobigny et al.
2005). Additional insight into this ancestral karyotype
was provided recently by Svartman et al. (2006), who
hybridized human chromosome probes to metaphases
of Dasypus novemcinctus, Tamandua tetradactyla, and
Choloepus hoffmani. They showed that the two-toed
sloth, C. hoffmani, (2n = 50), exhibited a chromosome
complement strikingly similar to the proposed 2n = 48
ancestral Eutherian karyotype. Future consideration
of additional species with less conserved genome architecture (see Jorge and Pereira this volume), such
as three-banded armadillos (Tolypeutes) and the silky
anteater (Cyclopes didactylus), will critically test the
conclusion of low genomic repatterning in Xenarthra.
Also, the inclusion of armadillo species belonging to
the two currently unsampled groups (Tolypeutinae and
Euphractinae) will allow a more precise delineation of
the ancestral xenarthran karyotype.
Phylogeny and conservation genetics
As reviewed by Prodhl et al. (this volume), population
genetic studies of living xenarthrans are just beginning,
with most efforts concentrated on the most common

and widespread xenarthran species: the nine-banded
armadillo (Dasypus novemcinctus). However, the ninebanded armadillo constitutes a case study illustrating
well the insights that can be gained from molecular
approaches, that is, combining fine-scale population
genetics with phylogeographic studies at a larger scale.
Such molecular studies have the potential to enhance
our understanding of proximal and historical factors
that influenced the evolutionary history of xenarthran
species and might help define conservation strategies.
Conservation genetic studies within Xenarthra so
far have been limited to flagship species such as the
endangered maned sloth (Bradypus torquatus) of the
remnant Brazilian Atlantic forest (Moraes-Barros et al.
2002), and the giant anteater (Myrmecophaga tridactyla), for which microsatellite loci are only beginning
to be developed (Garcia et al. 2005). Unfortunately,
Xenarthra contains a number of vulnerable and endangered species (Aguiar and Fonseca this volume) for
which virtually no data exist on their genetic diversity.
Consequently, species delineation still relies on morphological, geographical, and ecological observations.
Given the high rate at which biological diversity is currently being eroded, phylogeographic and population
genetic studies are urgently needed to characterize the
extent of genetic diversity in xenarthran populations
and thus identify biologically important conservation
entities. The definition of a comprehensive phylogenetic and taxonomic framework is a prerequisite for
designing efficient conservation plans for living xenarthran species (Purvis et al. 2005).
Conclusions
The use of molecular data to assess xenarthran phylogenetic relationships has provided several new insights
into the evolutionary history of this often neglected
major lineage of placental mammals. However, much
still needs to be done on the molecular side to answer
such fundamental questions as the position of Xenarthra within placentals. The new genomic era appears
full of promise to adequately locate the root of the placental tree, and modern cytogenetics has the potential
to reveal how the genomic architecture of xenarthrans
has evolved. The ultimate goal of obtaining a fully resolved molecular phylogeny for extant xenarthrans will
be achieved only through the collaborative and concerted efforts of the xenarthran research community.
Such a phylogenetic framework is urgently needed to
help design conservation strategies for the numerous

endangered xenarthran species (see Aguiar and Fonseca this volume).
Acknowledgments
We would like to thank the editors, Jim Loughry and
Sergio Vizcano, for inviting us to contribute to this
volume and for their efforts in bringing together the
xenarthran research community. Gerry de Iuliis and
William J. Murphy are also thanked for their helpful
suggestions on the manuscript. The following people
and institutions deserve some special thanks for kindly
giving access to xenarthran tissue samples over the
years: Tammie L. Bettinger (Cleveland Metroparks
Zoo, USA), Pablo Carmanchahi, Pablo Cetica, Jorge
23
Omar Garca, and Rodolfo Rearte (Complejo Ecolgico

Municipal de Presidencia Roque Senz Pea, Chaco,
Argentina), Danny Devillier, Eric Hansen, Guillermo
Anibal Lemus, Jesus Mavarez, Guillermo Perez Jimeno,
Mariella Superina, and Jean-Christophe Vi and his
team Faune Sauvage in French Guiana. Mark Springer
kindly provided the Amrine-Madsen et al. dataset.
Our work on xenarthran molecular phylogenetics
benefited from grants associated with the European
Community TMR Network Mammalian Phylogeny
(contract FMRX-CT98-022) and the European Marie
Curie Project HOTSPOTS Understanding and Conserving Earth Biodiversity (contract MEST-CT-2005020561). This is contribution ISEM 2006-005 of the Institut des Sciences de lEvolution de Montpellier (UMR
5554-CNRS).
3
Morphology-based investigations of the phylogenetic
relationships among extant and fossil xenarthrans
Timothy J. Gaudin and H. Gregory McDonald
de la fauna eocena de Messel en Alemania. Virtualmente todas las reconstrucciones recientes de la filogenia de los perezosos apoyan el origen difiltico de los
dos gneros arborcolas vivientes, pero difieren en las
hiptesis de relaciones con diferentes taxones fsiles.
Un comprehensivo anlisis reciente ubica a Bradypus
como el taxn hermano de los restantes perezosos y
rene a Choloepus con los megalonquidos extinguidos. Este estudio corrobora la monofilia de las familias
Nothrotheriidae, Megatheriidae, Megalonychidae y
Mylodontidae, apoya la alianza de notrotridos, megatridos y megalonquidos en un clado Megatheriodea
y, dentro de este ltimo grupo, rene a notrotridos
y megatridos en un clado llamado Megatheria. Las
relaciones entre estas familias de perezosos, particularmente Mylodontidae y Megalonychidae, as como
la taxonoma alfa de todos los xenartros, requieren de
nuevos estudios.
Resumen
En un grupo como el de los Xenarthra, en el que la diversidad conocida de formas extinguidas excede por
mucho la de las vivientes, aquellos anlisis morfolgicos
que incorporen taxones fsiles aumentan su importancia al mejorar nuestra comprensin de la filogenia. En
este estudio se revisan las investigaciones morfolgicas
recientes sobre filogenia de los xenartros. Se resaltan
las reas de amplio consenso, incluyendo la monofilia
de Xenarthra, de cada uno de sus tres subgrupos principales [Cingulata, Vermilingua, Phyllophaga (= Tardigrada o Folivora)] y de los Pilosa [Vermilingua + Phy
llophaga]. Tambin se revisan los estudios recientes de
las relaciones supraordinales del grupo, as como los
anlisis filogenticos dentro de los Cingulata, Vermilingua y Phyllophaga. Las afinidades de Xenarthra con
otros grupos de mamferos placentarios an generan
controversias y requieren de ms investigaciones. Dentro de los Cingulata an son poco comprendidas la ta
xonoma y la filogenia de los gliptodontes y se requieren ms estudios, aunque los anlisis recientes proveen
nueva informacin sobre la sistemtica de este grupo.
Un estudio cladstico confirma hiptesis filogenticas
previas sobre las relaciones entre armadillos extinguidos y vivientes, pampaterios y gliptodontes, incluyendo
la monofilia de los eufractinos vivientes y un clado que
rene a gliptodontes y pampaterios y la posicin basal
de los dasipodinos dentro de los Cingulata. Sin embargo, no sostiene la monofilia de los armadillos como
un todo o de los tolipeutinos, eutatinos y eufractinos.
Con respecto a los Vermilingua, las relaciones entre los
miembros indiscutidos del grupo no produce controversias, pero existe un gran desacuerdo sobre la ubicacin de Eurotamandua, el presunto oso hormiguero
Resumo
Em um grupo como o dos Xenarthra, em que a diversidade conhecida de formas extintas excede com muito a
vivente, os estudos morfolgicos que incorporam txons fsseis aumentam sua importncia, ao melhorar
nossa compreenso da filogenia. Nesse estudo, revisamse os estudos morfolgicos recentes sobre a filogenia
dos xenrtros. Resaltam-se as reas de amplo consenso,
incluindo o monofiletismo de Xenarthra, de um de seus
trs sub-grupos [Cingulata, Vermilingua, Phyllophaga
(= Tardigrada ou Folivora) e dos Pilosa [Vermilingua +
Phyllophaga]. Tambm, revisam-se os estudos recentes
das relaes supra-ordinrias do grupo, assim como as
anlises filogenticas dentro dos Cingulata, Vermilingua e Phyllophaga. As afinidades de Xenarthra como
24
Phylogenetic relationships among extant and fossil xenarthrans
25
Figure 3.1. Cladogram depicting the relationship among the three suborders of Xenarthra:
Cingulata, Vermilingua, and Phyllophaga (=
Tardigrada or Folivora). The alliance of anteaters and sloths in a monophyletic Pilosa to the
exclusion of armored xenarthrans has been
supported by a wide variety of subsequent morphological and molecular studies (see Introduction). Animals depicted across the top, from left
to right: Doedicurus (Pleistocene, SA), Chaetophractus (hairy armadillo), Myrmecophaga (giant
anteater), Choloepus (two-toed sloth), Megatherium (Pleistocene, SA). Modified from Gaudin
(2003). Abbreviations: SA = South America.
outros grupos de mamferos placentrios ainda geram

controvrsias e requerem mais estudos. Um estudo
cladstico confirma hipteses filogenticas prvias sobre
as relaes entre tatus extintos e viventes, pampatrios
e gliptodontes, incluindo o monofiletismo dos eufractinos viventes, e um clado que rene os gliptodontes e
pampatrios, e a posio basal dos dasipodinos dentro
dos Cingulata. Apesar disso, no sustenta o monofiletismo dos tatus como um todo, ou dos tolipeutinos, eutatinos e eufractinos. Com respeito aos Vermilingua, as
relaes entre os membros indiscutveis do grupo no
produz controvrsias, mas existe um grande desacordo
sobre a posio de Eurotamandua, o presumido tamandu da fauna eocnica de Messel na Alemanha. Virtualmente todas as reconstrues recentes da filogenia das
preguias apiam uma origem difiltica dos dois gneros de preguias arborcolas viventes, contudo diferem
sobre as hipotticas relaes com diferentes txons fsseis. Uma anlise recente posiciona Bradypus como o
txon-irmo das preguias restantes e rene Choloepus
com os megaloniqudeos extintos. Esse estudo corrobora o monofiletismo das famlias Nothrotheriidae,
Megatheriidae, Megalonychidae e Mylodontidae, apia
a aliana de notroterideos, megaterideos e megaloniqudeos em um clado Megatherioidea e, dentro deste
ltimo grupo, rene os notroterideos e megaterideos

em um clado chamado Megatheria. As relaes entre
estas famlias de preguias, particularmente Mylodontidae e Megalonychidae, assim como a taxonomia alfa
de todos os xenrtros, requer novos estudos.
Introduction
Historically, the roots of systematic analysis in biology lie in comparative studies of organismal morphology. Linnaeus (1758) based his great Systema Naturae
on similarities in overall form among the organisms
he surveyed. The systematists and taxonomists of the
eighteenth and nineteenth centuries who constructed
much of the basic taxonomy of living and extinct xenarthrans did so on the basis of morphology (much
of this early literature is reviewed in Hoffstetter 1958,
1982; Glass 1985; Wetzel 1985a). Individual species and
higher-level taxa have been recognized using external
morphological or skeletal characteristics, be they living
(Wetzel 1985a) or extinct (Hoffstetter 1958). The various
groupings and subgroupings within Xenarthra were
established largely on the basis of skeletal and dental
traits (appendix 3.1, figure 3.1): the Cingulata, the armored xenarthrans, which bear a carapace formed by
26
T. J. Gaudin and H. G. McDonald
a mosaic of dermal ossifications and epidermal scales,

can be divided into armadillos, with their peglike teeth,
and pampatheres, with enlarged bilobate teeth, both of
which have imbricating osteoderms that allow flexibility in the carapace, and the glyptodonts, with their very
peculiar trilobate teeth and immobile carapaces; the
Vermilingua or anteaters, which have tubular, edentulous skulls; and the Phyllophaga (= Tardigrada or Folivora), including the living tree sloths and extinct ground
sloths, which share a hypselodont herbivorous dentition and a characteristic reduced dental formula with
a maximum of five upper and four lower teeth in each
half of the jaw (Gaudin 1999a). Armadillos, anteaters,
and sloths were originally placed in different groups by
Linnaeus, with armadillos assigned to his Bestiae along
with pigs, some insectivores, and opossums, whereas
anteaters and sloths were allocated to Bruta along with
elephants, manatees, and pangolins. Nevertheless, these
three very disparate types of mammals were eventually
united into a single order because they share a suite
of unique and unusual morphological characteristics
(Gaudin 1999a; McDonald 2003b; Rose et al. 2005; appendix 3.1), among them the feature for which the order was named, the xenarthrous intervertebral articulations (Gaudin 1999b).
The advent of modern cladistic techniques has facilitated improvements in our detailed understanding
of xenarthran phylogeny, as has the development of
molecular systematic analyses. As discussed in OLeary
(1999) and OLeary and Geisler (1999), morphologybased phylogenetic analyses are able to incorporate a
much broader taxonomic sample than molecular studies because of the inclusion of fossil taxa, but at the cost
of limited character sampling, being restricted almost
entirely to osteological characters, particularly if the
fossil record is to be included. Xenarthra is unusual (but
not unique) among mammalian orders in the sense that
it includes a well-known and tremendously diverse radiation of extinct taxa, including more than 150 genera
(McKenna and Bell 1997), but only 14 extant genera and
31 extant species (Gaudin 2003). Two major groups, the
glyptodonts and pampatheres, have no modern representatives and sloths are only minimally represented (2
extant genera with 6 species). Thus, systematic studies that consider only extant taxa are likely to be particularly susceptible to taxonomic sampling problems
such as long-branch attraction (Hillis 1998). Therefore,
morphological studies of xenarthran phylogeny, especially those that examine a wide variety of extinct taxa,
should continue to play a vital role in improving our

knowledge of the phylogenetic history of Xenarthra.
Recent advances in the study of xenarthran phylogeny utilizing both morphological and molecular investigations have been reviewed by Garcia (2003) and
Gaudin (2003). In addition, a separate chapter (Delsuc
and Douzery this volume) treats recent developments
in the use of mitochondrial and nuclear gene sequences
to examine relationships of extant xenarthrans. Therefore, in this chapter we will focus on morphology-based
investigations of phylogeny that have been published
subsequent to Gaudin (2003), while also considering
pertinent earlier studies that have laid the groundwork
for our current understanding of the phylogeny of the
group. We will follow the basic outline of Gaudins
(2003) review. First, we will consider issues related to
the supraordinal relationships of Xenarthra. Several
aspects of xenarthran phylogeny are well established in
the published literature (Delsuc et al. 2002; Delsuc et al.
2003; Garcia 2003; Gaudin 2003, 2004a; Rose et al. 2005;
Gaudin and Wible 2006) and need not be treated in detail here. These include the monophyly of Xenarthra as
a whole and the monophyly of each of the three major
subgroups, Cingulata, Vermilingua and Phyllophaga, as
well as the monophyly of the Pilosa, a clade that unites
sloths and anteaters to the exclusion of cingulates (appendix 3.1, figure 3.1). However, we will review studies
that investigate phylogenetic relationships within each
of the three primary subgroups.
Supraordinal relationships
Morphology-based studies of mammalian phylogeny
over the past several decades have tended to place
Xenarthra in a remote position within Placentalia, for
example, as the sister taxon to other placentals (Epitheria), although perhaps allied with certain extant or
extinct groups like pangolins (Order Pholidota) and
Palaeanodonta (Novacek and Wyss 1986; Novacek
1992). Placement of the Xenarthra as the sister group
to the other placentals was first advocated by early
cladistic morphological studies (McKenna 1975). The
Xenarthra/Epitheria dichotomy has been criticized by
other morphologists (Rose and Emry 1993; Gaudin et
al. 1996), and both morphological and molecular work
has failed to support the xenarthran/pholidotan clade
(Rose and Emry 1993; Delsuc et al. 2002; Rose et al.
2005). However, recent molecular studies generally favor a remote position of Xenarthra among placentals.
Xenarthra has been identified as one of the four primary

placental clades, along with Afrotheria, Laurasiatheria,
and Euarchontoglires (Madsen et al. 2001; Delsuc et al.
2002; Springer et al. 2004). The relationships among
these four clades are not well resolved, although either
Xenarthra or Afrotheria, or a clade uniting the two, is
typically placed at the base of the placental tree (Madsen et al. 2001 Delsuc et al. 2002; Springer et al. 2004).
Recently, two studies of higher-level mammalian
relationships have been published that include representatives of Xenarthra and employ morphological
characters either exclusively (Luo and Wible 2005) or
in combination with molecular characters (Asher et
al. 2003). Interestingly, neither supports a basal position for Xenarthra within Placentalia. Luo and Wible
(2005) sampled only a small number of extant placental taxa, along with a variety of extinct basal eutherians. Nevertheless, xenarthrans along with carnivorans
formed a crown clade, exclusive of extant representatives from Glires and Lipotyphla (Luo and Wible 2005).
In Asher et al. (2003), when morphological characters
alone were considered, Xenarthra formed a clade with
aardvarks and pangolins that in turn was part of a large
basal placental multichotomy in the strict consensus
tree. However, when molecular characters were added,
Xenarthra was not only pulled apart from pholidotans
and tubulidentates in the strict consensus trees, it was
pulled into a crown clade including afrotherians and
laurasiatherians and excluding members of a paraphyletic stem group of Euarchontoglires (Asher et al.
2003).
It is worth reiterating that the morphological data
of Asher et al. (2003) supported a close relationship
between Pholidota and Xenarthrain fact, it resulted
in a sister-group relationship between the pangolin
genus Manis and the vermilinguan Tamandua, thus
resulting in a nonmonophyletic Xenarthra. However,
when molecular data were added, this relationship fell
apart. Pholidotans were not included in the Luo and
Wible (2005) study. Clearly, consensus concerning the
surpraordinal relationships of Xenarthra has not been
achieved, and more investigation of this question is
warranted.
Relationships within Cingulata

Cingulata encompasses the extant armadillos as well as
a large radiation of extinct forms, including not only
a wide variety of fossil armadillos, but also the her-
27
bivorous pampatheres and glyptodonts (figure 3.2; see

also Fernicola et al. this volume). Gaudin (2003) points
out that armadillos are the most diverse clade of living xenarthrans and have the oldest fossil record, yet
study of their relationships, and indeed the phylogeny
of all cingulates, has been largely neglected until recently. The work by Delsuc et al. (2002, 2003), using
nuclear and mitochondrial DNA sequences, has done
much to rectify the situation for extant armadillos, although data for the fairy armadillo Chlamyphorus are
as yet unavailable. Morphology-based investigations of
cingulate relationships, however, especially those that
consider extinct taxa, remain scarce, with one or two
noteworthy exceptions.
The systematics of the diverse and formerly abundant
assemblage of extinct glyptodonts are the least studied
and least understood among any of the major groups
of cingulates. Glyptodonts have been subdivided into
various subfamilies and tribes based primarily on characters relating to ornamentation of the carapace and the
structure of the caudal armor, along with selected features of the cranial and postcranial skeleton. However,
there is no consensus as to their taxonomy at any level
and certainly little phylogenetic analysis of the relationships among the various taxa. Paula Couto (1979) listed
four subfamilies: Propalaehoplophorinae; Hoplophorinae, with the tribes Palaeohoplophorini, Plohophorini,
Hoplophorini, Panochthini, Lomaphorini, Neothoracophorini, and Neuryurini; Doedicurinae; and Glyptodontinae. In their study of the ear region of cingulates,
Patterson et al. (1989) recognized the same four subfamilies: Propalaehoplophorinae, Sclerocalyptinae (=
Hoplophorinae), Glyptodontinae, and Doedicurinae,
but with fewer tribes, as the Sclerocalyptinae was not
subdivided into tribes and the Doedicurinae included
only the Panochthini and Doedicurini. It should be
noted that their work focused primarily on descriptive
anatomy of the ear region and that most of the study
concerned armadillos. Only two glyptodonts, Plohophorus (= Hoplophractus) and Eleutherocercus, were examined. Neither of the above taxonomies recognized
the enigmatic subfamily Glyptatelinae, which first appeared in the middle Eocene (Mustersan) and survived
into at least the middle Pleistocene (Irvingtonian) of
North America in the form of Pachyarmatherium
(Downing and White 1995, Vizcano, Rinderknecht,
and Czerwonogora 2003). In contrast, McKenna and
Bell (1997) divided the Glyptodontidae into five subfamilies: the Glyptatelinae and the four subfamilies
28
recognized by Paula Couto (1979) and Patterson et al.

(1989). The primary difference was that only two of
the subfamilies were subdivided into tribes, with the
Hoplophorinae having the largest number, including
the Hoplophorini, Lomaphorini, Palaehoplophorini,
Plohophorini, Panochthini, and Neuryurini. The
Glyptodontinae was subdivided into two tribes, the
Glyptodontini for South American forms and Glyptotheriini for the North American genus Glyptotherium.
The tribe Neothoracophorini was also recognized but
not placed in any subfamily.
These earlier classifications have been contradicted
in a number of important respects by the recent phylogenetic studies of Fernicola (2005). Fernicola (2005)
conducted a cladistic analysis based upon 84 craniodental characters in 12 glyptodont genera. His results
are summarized by Fernicola et al. (this volume), but
it is worth noting here that the monophyly of several
of the traditional subfamilies is not supported by his
work. Indeed, his results reinforce the observation that
glyptodonts are the one group within Xenarthra most
requiring intensive study and revision, with regard to
both taxonomy and phylogeny.
Among the cingulates the only broad, morphologybased cladistic study of the groups phylogeny is that
of Gaudin and Wible (2006, figure 3.2). Gaudin and
Wible (2006) examined the craniodental morphology
of a wide diversity of living and extinct armadillos,
including all 9 extant genera, 3 genera of eutatine armadillos, 4 genera of extinct euphractine armadillos,
the extinct genera Stegotherium and Peltephilus, and a
single representative each from the pampatheres (Vassallia) and glyptodonts (Propalaehoplophorus). Their
results are congruent with other, earlier investigations
of cingulate phylogeny in a number of respects. Like
Patterson et al. (1989) and Engelmann (1985), they recognize a close relationship between pampatheres and
glyptodonts (figure 3.2). Their alliance of Stegotherium
with the extant long-nosed armadillos (Dasypus) likewise echoes the conclusions of Patterson et al. (1989)
and Engelmann (1985), and the remote placement of
the horned armadillo Peltephilus as the sister group to
other cingulates (figure 3.2) is reminiscent of earlier
classifications (e.g., Hoffstetter 1958) where Peltephilus was placed in its own family. The conclusions of
Gaudin and Wible (2006) are also congruent with the
molecular analyses of Delsuc et al. (2002, 2003) in recognizing the monophyly of the extant euphractine armadillos and in placing Dasypus on a separate branch
from other armadillos at the base of the clade (figure
3.2). Nevertheless, many of the conclusions of the study

are startling when compared to previous work. In the
phylogeny of Gaudin and Wible (2006), the armadillos
(sensu lato) are paraphyletic, with the clade including
pampatheres and glyptodonts as a crown group nested
within the armadillo radiation (figure 3.2). The tolypeutine clade of extant armadillos recovered by Delsuc
et al. (2002, 2003) was not obtained in any of Gaudin
and Wibles (2006) analyses. Lastly, Gaudin and Wibles
(2006) study fails to unite extinct forms traditionally
placed in the Euphractinae with the extant euphractine
clade, arranging them instead in a paraphyletic assemblage that includes both extinct euphractines and eutatines, making both groups polyphyletic (figure 3.2).
Gaudin and Wible (2006) acknowledge that their
analysis is based on a restricted character base (appendix 3.1), and call for the addition of postcranial or
soft-tissue data to further refine our understanding of
cingulate phylogeny. Certainly a total-evidence analysis of cingulate phylogeny, combining morphological
and molecular data, would also be a valuable contribution to the systematic literature on this group. A more
complete understanding of cingulate phylogeny will
ultimately be an invaluable enhancement of ongoing
studies of anatomical and functional evolution among
armored xenarthrans (e.g., Hill 2004; Kalthoff 2004;
Vizcano, Faria et al. 2004; Fernicola 2005; Fernicola
et al. this volume; Vizcano et al. this volume).
Relationships within Vermilingua

Recent analyses have reached a consensus concerning the relationships among undoubted vermilinguans. Both morphological (Gaudin and Branham
1998; Gaudin 2003) and molecular studies (Delsuc et
al. 2002, 2003) support the alliance of the extant genera Tamandua and Myrmecophaga in a monophyletic
clade (Myrmecophagidae), with the third extant genus
Cyclopes placed in a separate group (Cyclopedidae).
Moreover, there is broad agreement on the affinities
of the South American fossil anteatersthe Miocene
Protamandua and Mio-Pliocene Neotamandua allied
with myrmecophagids, and the Pliocene Palaeomyrmidon allied with Cyclopes (Gaudin 2003; see also McDonald et al. this volume). The consensus breaks down,
however, when the affinities of the putative European
anteater Eurotamandua are considered.
Eurotamandua derives from the middle Eocene
Messel deposits of central Germany and was initially
assigned to the Vermilingua (Storch 1981), making it
Figure 3.2. Cladogram summarizing relationships within the Cingulata, as proposed by Gaudin and Wible (2006). Animals
depicted at the right of the figure, starting at the top: Glyptodon (Pleistocene, SA), Chaetophractus (hairy armadillo), Tolypeutes
(three-banded armadillo), Dasypus (long-nosed armadillo). Skulls depicted in left lateral view at the far right of the figure,
starting at the top: Propalaehoplophorus (Miocene, SA), Holmesina (Pleistocene, SA and NA), Proeutatus (Miocene, SA), Euphractus (yellow armadillo), Tolypeutes, Dasypus, Peltephilus (Miocene, SA). Animal figures of Glyptodon, Chaetophractus, and
Tolypeutes from Gaudin (2003). Abbreviations: NA = North America, SA = South America.
30
the only early Cenozoic xenarthran known from outside South America. Subsequent authors have argued
that Eurotamandua is not a true vermilinguan, but is
more properly placed within Xenarthra as the sister
taxon of Pilosa, outside Xenarthra as the sister taxon to
Palaeanodonta, or even in its own order, distinct from,
but distantly related to, Xenarthra and Palaeanodonta
(Rose et al. 2005). In the most recent review of the matter (Rose et al. 2005), the authors themselves could
not agree on the proper phylogenetic allocation of this
taxon, with one author favoring ties to the Vermilingua,
and the other three suggesting close relationship either
to Palaeanodonta or to the earliest pangolin Eomanis.
Indeed, several phylogenetic analyses carried out by one
of us (Gaudin 2004b, 2005) would tend to strengthen
this latter claim. Nevertheless, it seems unlikely that a
convincing case can be made for the affinities of this
enigmatic taxon in the absence of a more comprehensive phylogenetic analysis that includes representatives
from each of the major families within Pholidota, Palaeanodonta, and Xenarthra, and possibly other nonedentate placental mammals.
Relationships within Phyllophaga

By far the most active area for morphology-based investigations of xenarthran phylogeny has been among
the extinct and extant sloths. In the past decade numerous studies have examined relationships within and
among the various sloth families using morphological
(Perea 1992; Gaudin 1995, 2004a; De Iuliis 1996; White
and MacPhee 2001; McDonald and Muizon 2002; McDonald and Perea 2002; Pujos 2002, 2006; Muizon et
al. 2003, 2004a; see also McDonald and De Iuliis this
volume; Pujos this volume) and molecular characters
(Hss et al. 1996; Poinar et al. 1998, 2003; Greenwood
et al. 2001; Hofreiter et al. 2003). The latter group of
papers is remarkable for both its inclusion of ancient
DNA sequence data from extinct sloths, and, what is
equally important, the lack of agreement among the results of the various studies. This is no doubt attributable
to both the fragmentary nature of the recovered DNA
sequences and the extremely limited taxonomic sample
available for molecular study, which has included only
two extinct genera representing two separate families,
the nothrotheriid Nothrotheriops from North America
and the mylodontid Mylodon from South America,
along with a sample recovered from a coprolite (Hofreiter et al. 2003) that currently cannot be assigned to
a described genus or species. The conclusions of these
studies are detailed elsewhere (Garcia 2003; Gaudin

2003, 2004a; see also Delsuc and Douzery this volume),
and will not be discussed further here.
The most comprehensive of the recent morphologybased investigations is that of Gaudin (2004a; figure
3.3). Employing an extensive dataset of craniodental
characters examined in 33 extinct and extant sloth
genera, including representatives of each of the major
sloth families as well as a wide variety of xenarthran
outgroup taxa, Gaudin (2004a) was able to investigate
such questions as the monophyly or diphyly of the two
extant tree sloth genera and the relationships among
and within the various sloth families. Gaudins (2004a)
results strongly support the diphyly of the tree sloths
(figure 3.3), an idea previously proposed by Patterson
and Pascual (1972) and further developed by Webb
(1985a). Gaudin (2004a) positioned Bradypus as the
sister taxon to all other sloths (a clade termed Eutardigrada), in contrast to both Patterson and Pascuals and
Webbs studies, where Bradypus was considered closely
allied with the megatheres. Gaudins (2004a) results
are, however, congruent with those of Patterson and
Pascual (1972) and Webb (1985a) in the placement of
Choloepus within the family Megalonychidae, a clade
that includes the extinct Antillean sloths (figure 3.3).
Although the detailed relationships among tree sloths
and ground sloths advocated by Gaudin (2004a) differ from those proposed in other recent morphological
(White and MacPhee 2001; Pujos 2002) and molecular
analyses (Poinar et al. 1998, 2003; Greenwood et al. 2001;
Hofreiter et al. 2003), it is significant that all of them
support the diphyly of tree sloths. Gaudins (2004a)
phylogeny also corroborates the monophyly of the
three extinct families of ground sloths, the Mylodontidae, Megatheriidae, and Nothrotheriidae, as well as the
Megalonychidae (figure 3.3). Megatheriids, nothrotheriids, and megalonychids are joined in a monophyletic
clade Megatherioidea, within which Megatheriidae and
Nothrotheriidae form a monophyletic grouping termed
the Megatheria (figure 3.3). A variety of Santacrucian
(earlymiddle Miocene) taxa traditionally labeled
nothrotheres are removed to the base of the Megatherioidea, although one of these taxa, Eucholoeops, is
recognized as a basal megalonychid (Gaudin 2004a).
A number of Gaudins (2004a) phylogenetic conclusions are consistent with the findings of other recent
morphological studies. McDonald and Muizon (2002)
and Muizon et al. (2003, 2004a) support the recognition of a monophyletic family Nothrotheriidae for the
late Miocene-Pleistocene nothrotheres, and within that
Figure 3.3. Cladogram summarizing relationships within the Phyllophaga (= Tardigrada or Folivora), as proposed by Gaudin
(2004a). Sloths are divided into four monophyletic families: Megalonychidae, Megatheriidae, Nothrotheriidae, and Mylodontidae. The relationship of certain basal, mostly Santacrucian megatherioid sloths, to the three main megatherioid families
(Megalonychidae, Megatheriidae, and Nothrotheriidae) is not unambiguously resolved. The extant Bradypus is placed as the
sister taxon to all other sloths, whereas extant Choloepus is incorporated into the family Megalonychidae. Animals depicted
at the right of the figure from top to bottom: Choloepus (two-toed sloth), Megatherium (Pleistocene, SA), Bradypus (three-toed
sloth). Skulls depicted in left lateral view at the far right of the diagram, from top to bottom: Acratocnus (Pleistocene, WI),
Choloepus, Eremotherium (Pleistocene, SA and NA), Nothrotheriops (Pleistocene, NA), Hapalops (Miocene, SA), Paramylodon
(Pleistocene, NA), Bradypus. Skull and animal figures from Gaudin (2003, 2004a). Abbreviations: NA = North America, SA =
South America, WI = West Indies.
32
family, the close relationship between the North American Pleistocene genus Nothrotheriops and the South
American Pleistocene genus Nothrotherium. Indeed,
Muizon et al. (2004a) provide a formal diagnosis for the
newly recognized family. White and MacPhee (2001)
confirm the diphyly of the modern tree sloths and the
close relationship between extant Choloepus and smallbodied Antillean megalonychids. Their phylogenetic
conclusions were based on an analysis of cranial and
postcranial skeletal characters in a limited array of sloth
taxa, and their analysis did not include any megalonychid taxa from South America such as Pliomorphus.
This may explain why their results differ from those
of Gaudin (2004a) in a number of important respects.
For example, their study does not yield a monophyletic
Megalonychidae, but instead splits the Antillean radiation of megalonychids into distinct large-bodied and
small-bodied clades, suggesting two separate dispersal
events of this family into the Caribbean islands. Their
recognition of two distinct clades within the Antillean radiation of sloths is reminiscent of Kraglievichs
(1923) taxonomy, in which Megalocnus (the largest of
the Antillean sloths) was placed in a distinct subfamily (Megalocninae) whereas Microcnus (a small-bodied
form) and other Antillean sloths were assigned to the
Ortotherinae.
Gaudins (2004a) cladogram also closely parallels
some of the relationships of the mylodontine mylodonts derived by Perea (1992). The two studies cannot
be directly compared as Perea utilized a number of genera (e.g., Sphenotherus, Promylodon, Ranculcus, Megabradys and Prolestodon) not incorporated by Gaudin,
and likewise Gaudin utilized some genera (e.g., Thinobadistes and Paramylodon) not included by Perea. In
addition, Perea (1992) based his analysis solely on the
mandible because this is the only element available for
some taxa; hence he was limited to a smaller number
of characters. Despite all these differences, the two
cladograms are congruent at a number of places. In
both analyses Mylodon is the sister taxon to the other
genera, followed by Pleurolestodon, then Glossotherium,
with Lestodon as the most derived member of the Mylodontinae.
Several of the studies cited above deal in more detail with taxa given only superficial consideration by
Gaudin (2004a), including detailed, species-level phylogenies of megatheriid (De Iuliis 1996; Pujos 2002,
2006; see also Pujos this volume), scelidotheriine
(McDonald and Perea 2002), and mylodontine (Perea
1992) sloths. These studies demonstrate that there is

still much to learn about the phylogeny of sloths, a fact
readily admitted by Gaudin (2004a) in the conclusion
to his study. Questions still needing additional study
include the within-group phylogenetic relationships of
mylodontids and megalonychids and the detailed relationships of the extant tree sloths to the various groups
of extinct sloths. An analysis that incorporates postcranial skeletal characters, or soft tissue or molecular
characters, alongside Gaudins (2004a) extensive matrix
of craniodental features, should also prove valuable in
further elucidating the phylogenetic history of sloths.
Conclusions
Because of the disparities between the extinct and extant radiations in Xenarthra, with known fossil forms
greatly outnumbering the depauperate extant assemblage, morphology-based investigations of phylogeny that can account for this extinct diversity take on
particular importance. Morphological studies have so
far failed to achieve a consensus concerning the supraordinal relationships of Xenarthra, either among
themselves or with molecular studies. The latest morphological studies do not support a remote position
of Xenarthra within Placentalia, however, and offer at
best equivocal support for a relationship between Xenarthra and Pholidota and/or Palaeanodonta. Within
Xenarthra, support for the monophyly of three major
subgroups, Cingulata, Vermilingua, and Phyllophaga,
is consistently achieved, and there is robust support for
the alliance of the latter two groupings into a monophyletic clade Pilosa.
Among cingulates, the recent study by Gaudin and
Wible (2006) throws doubt on several traditional phylogenetic hypotheses, but is consistent with previous
studies, including molecular ones, in a number of respects, such as support of a monophyletic grouping of
glyptodonts and pampatheres and of extant euphractine armadillos and its recognition of dasypodine armadillos as a separate and basal cingulate group. The
systematics of glyptodonts is poorly known, although
recent work by Fernicola (2005; see also Fernicola et al.
this volume) holds promise for improving our phylogenetic understanding of this group.
Relationships within Vermilingua are broadly agreed
upon, with the exception of the position of the putative European anteater genus Eurotamandua. Strong
disagreements persist in the literature over the proper
allocation of this taxon, whether to Vermilingua, to

Palaeanodonta, to Pholidota, or to its own order. Only
a comprehensive phylogenetic analysis incorporating
representative taxa from all these groups is likely to
convincingly resolve this issue.
A comprehensive analysis of sloth phylogeny by
Gaudin (2004a) is consistent with other recent morphological and molecular studies in affirming the diphyletic
origin of the modern tree sloths. More controversial is
his allocation of Choloepus to the Megalonychidae and
Bradypus to a position as the sister taxon to all other
sloths. Gaudins (2004a) results also recognize a monophyletic grouping Megatherioidea including megatheriid, nothrotheriid, and meglonychid sloths to the exclusion of mylodontids, and within megatherioids a close
association between megatheriids and nothrotheriids.
Gaudins (2004a) recognition of a distinct family Nothrotheriidae for late Miocene-Pleistocene nothrotheres
is consistent with other recent studies. However, there
remains much to learn about the relationships within
each of the sloth families. Additional phylogenetic issues unaddressed or unresolved by Gaudins (2004a)
study include the relationship to other sloths of Pseudoglyptodon (Engelmann 1987; Wyss et al. 1990), Entelops (Pascual 1960), or the subfamily Schismotheriinae,
a grouping recognized by McKenna and Bell (1997) as
comprising eight genera of pre-Pliocene sloths, among
them the ever-popular root taxon Hapalops. All schismotheriines were formerly considered to be early nothrotheres. Do they represent a true monophyletic clade
or are they only an artificial paraphyletic taxon created
as a by-product of the more intensive study of the PlioPleistocene forms?
Lastly, it should be noted that there remains a critical
need for more alpha-level taxonomy in all xenarthran
groups, as the validity of many of the genera and species described in the past requires serious reexamination. There is no doubt that oversplitting of taxa at both
the genus and species level has resulted in the true diversity of fossil xenarthrans being hidden in a plethora
of names that lack any biological reality. If our understanding of the phylogenetic history of the Xenarthra is
to improve, any future analysis will require the inclusion of more taxonomically valid taxa.
Acknowledgments
We thank Jim Loughry and Sergio Vizcano for their
work in organizing and editing this volume, and for in-
33
viting us to contribute this review. We are grateful to

Julia Morgan Scott for producing almost all of the drawings used in this paper. We also wish to thank the editors and Susana Bargo and John Wible for their careful
critique of the manuscript. Tim Gaudins work on this
study was supported by NSF RUI Grant DEB 0107922.
Appendix 3.1. Synapomorphies of Xenarthra and

major subgroups.
Terminology for cranial anatomy follows Wible and
Gaudin (2004). Notations by the present authors are
set off in square brackets [ ].
Xenarthra
Engelmann (1985)(1) xenarthrous intervertebral articulations [see Gaudin 1999b below]; (2) fusion of ischium to
anterior caudal vertebrae; (3) dermal ossicles in skin; (4)
reduction of dentition [see McDonald 2003b below]; (5)
placement of infraorbital canal entirely lateral to body of
maxilla; (6) development of secondary scapular spine on
the posterior margin of the scapula; (7) presence of m.
rectus thoracis lateralis; (8) presence of m. pterygo-tympanicus; (9) presence of extensive retia mirabile in limbs;
(10) paired post-renal venae cavae.
Gaudin (1995)(11) dorsoventrally elongate ectotympanic
bone; (12) presence of the pterygoid bone in the bony
wall of the tympanic cavity; (13) reduction of the anteroposterior length of the postglenoid region of the skull;
(14) participation of the entotympanic in the rim of the
jugular foramen; (15) entotympanic/pterygoid contact;
(16) absence of ectotympanic/alisphenoid contact; (17)
fusion of distal tympanohyal to mastoid.
Gaudin et al. (1996)(18) stapedial artery lost in adults.
Reiss (1997)(19) m. mylohyoideus originates from dentaries, basicranium, and soft palate; (20) anterior digastric m. contributes to m. sternomandibularis; (21) transverse m. stylopharyngeus enters soft palate.
Gaudin (1999b)(22) wide zygapophyseal facets in posterior thoracic vertebrae; (23) enlarged metapophyses; (24)
enlarged anapophyses in posterior thoracic and lumbar
vertebrae; (25) medial and lateral zygapophyseal facets in
post-diaphragmatic vertebrae [modification of (1) from
Engelmann (1985)]; (26) xenarthrous articulations between metapophysis and anapophysis [modification of
(1) from Engelmann (1985)].
McDonald (2003b)(27) reduction or loss of premaxillary
teeth [modification of (4) from Engelmann (1985)]; (28)
loss of deciduous dentition [modification of (4) from Engelmann (1985)]; (29) reduction or loss of tooth enamel
[modification of (4) from Engelmann (1985)]; (30) pres-
34
ence of ventral articular processes on the sternebrae that

form complex synovial articulations with the sternal ribs;
(31) ossified sternal ribs that bear articular facets for one
another and one or two articular heads for the sternebrae; (32) loss of fibular/calcaneal contact; (33) proximal
phalanges of manus and pes proximodistally compressed;
(34) presence of a large palmar sesamoid or falciform in
the tendon of the m. flexor digitorum profundus.
Gaudin (2004a)(35) septomaxilla present; (36) facial
exposure of lacrimal larger than orbital exposure; (37)
posterior upper teeth slant labially, posterior lower teeth
inclined lingually; (38) ossified larynx.
Wible and Gaudin (2004)(39) multiple foramina for the
major, accessory, and minor palatine nerves and vessels;
(40) loss of alisphenoid canal; (41) a caudal palatine foramen that includes the minor palatine nerve and vessels; (42) absence of inferior petrosal sinus foramen; (43)
reduction or loss of foramen for the ramus superior of
the stapedial artery; (44) hiatus Fallopii absent; (45) lacrimal foramen on the face; (46) mandibular foramen at
the level of the alveolar plane; (47) absence of mastoid
foramen; (48) absence of stapedial artery groove; (49)
closed stylomastoid foramen; (50) rostral opening of the
pterygoid canal visible in lateral view below the sphenorbital fissure; (51) an extrabullar internal carotid artery.
Rose et al. (2005)(52) loss of interparietal; (53) presence of
a well-developed entotympanic; (54) anteroposteriorly
expanded ribs; (55) reduction in the number of lumbar
vertebrae; (56) tendency to incorporate caudal vertebrae
in the sacrum; (57) scapula with elevated spine and elongate acromion.
Cingulata
Engelmann (1985)(1) modification of dermal ossicles into
flattened, interlocking plates [= scutes]; (2) fusion of
axis to one or more cervical vertebrae; (3) fusion of tibia
and fibula; (4) development of a lateral keel on the radial
articulation of the humeral trochlea; (5) greater trochanter of femur extends proximal to the head; (6) loss of the
foramen rotundum [probably represents a fusion of the
foramen rotundum and sphenorbital fissure]; (7) presence of pronounced postglenoid fossa; (8) loss of pars
intermedia of hypophysis.
Gaudin (1995)(9) enlarged paroccipital process; (10) glenoid fossa of squamosal posterodorsally inclined.
Gaudin (1999b)(11) xenarthrous articulations between
anapophysis and ribs in thoracic vertebrae, between anapophysis and transverse processes in lumbar vertebrae.
McDonald (2003b)(12) homodont dentition; (13) development of a carapace that includes a cephalic shield and
in all but glyptodonts includes pectoral and pelvic buck-
lers connected by movable bands [there are also typically

movable bands around the tail].
Gaudin (2004a)(14) posterior teeth covered laterally by
the ascending ramus of the mandible; (15) nearly vertical
posterior edge of the condylar process of the mandible;
(16) anterior edge of nasals evenly convex; (17) occipital
condyles positioned immediately posterior to condyloid
foramina.
Wible and Gaudin (2004)(18) presence of a canal or foramen for the auricular ramus of the vagus nerve; (19)
carotid foramen between the basisphenoid and the petrosal; (20) cavum supracochleare is open ventrally; (21)
Glaserian fissure lies adjacent to the foramen ovale; (22)
hypoglossal foramen lies ventral to the jugular foramen;
(23) incisive foramen lies within, or nearly within, the
premaxilla; (24) presence of at least two mental foramina; (25) piriform fenestra confluent with the carotid
foramen; (26) presence of double posterior openings of
the posttemporal canal, the lower entirely within the petrosal; (27) sphenopalatine and caudal palatine foramina
confluent.
Gaudin and Wible (2006)(28) occipital height greater
than or equal to its width; (29) occipital artery traveling in a groove on the occiput extending dorsal to the
posttemporal foramen; (30) width of the external nares
much greater than their height; (31) external nasal aperture inclined anteroventrally in lateral view; (32) anterior portion of the nasoturbinal lying medial to the
nasal/maxillary or premaxillary suture; (33) rectangular
occipital condyles.
Pilosa
Engelmann (1985)(1) scapular fenestra [= coracoid foramen] immediately above the coracoid process of the
scapula; (2) concave articular surface for the navicular
on the astragalar head; (3) posterior displacement of
the kidneys into the pelvic cavity; (4) testes positioned
within the pelvic cavity; (5) expanded epitympanic sinus
within the squamosal; (6) reduction or loss of the postglenoid foramen; (7) jugular foramen recessed above and
behind the petrosal, entotympanic and/or tympanohyal;
(8) presence of a femoral head to the m. flexor cruris lateralis; (9) incomplete zygomatic arch; (10) loss of the pars
tuberalis of the hypophysis.
Gaudin (1995)(11) subarcuate fossa directly dorsal to internal auditory meatus; (12) enlarged jugular foramen;
(13) glenoid fossa of squamosal bounded by lateral shelf;
(14) entotympanic/mastoid contact.
Gaudin and Branham (1998)(15) lacrimal foramen with
prominent lateral walls; (16) absence of metacromion of
scapula; (17) fourth metacarpal longest metacarpal; (18)
sustentacular and navicular facets of astragalus confluent.

Gaudin (2004a)(19) angular process of mandible with
medially inflected tip; (20) lateral edges of mandibular
spout convergent anteriorly in dorsal view; (21) nasoturbinal and maxilloturbinal subequal in length; (22) jugal
loosely attached to skull; (23) separate foramen rotundum; (24) nuchal crest does not overhang occiput posteriorly.
Wible and Gaudin (2004)(25) Glaserian fissure well separated from the foramen ovale; (26) incisive foramen
elongate; (27) maxillary foramen visible in ventral view;
(28) optic canal positioned ventrally in the orbitosphenoid; (29) absence of an anterior opening of the orbitotemporal canal.
Vermilingua
Engelmann (1985)(1) loss of teeth; (2) elongate muzzle; (3)
posteriorly extended pterygoids [see Gaudin and Branham (1998), Gaudin (2004a)]; (4) ossified auditory bulla;
(5) flattened ribs.
Gaudin (1995)(6) entotympanic reduced; (7) pterygoid
enlarged to form medial bony wall of tympanic cavity;
(8) musculotubal canal exits tympanic cavity posteromedially; (9) musculotubal canal directed posteroventrally;
(10) arteria diploetica magna enters sidewall of braincase.
Reiss (1997)(11) jaw adductors weak; (12) loss of m. styloglossus; (13) loss of tongue insertion of m. hyloglossus; (14) loss of tongue insertion of m. palatoglossus; (15)
compound m. sternoglossus present.
Gaudin and Branham (1998)(16) dorsal process of premaxilla erect, compressed anteroposteriorly; (17) small
exposure of maxilla in orbit; (18) frontal/parietal suture
well anterior to glenoid; (19) temporal lines diverge posteriorly, widely separated from nuchal crest; (20) hard
palate extends posteriorly to the back of the tympanic
cavity [modification of (3) from Engelmann (1985)]; (21)
subarcuate fossa large, deep; (22) basicranial/basifacial
axis slightly curved [concave ventrally]; (23) mandibular
symphysis strongly downturned ventrally; (24) anterior
edge of spinous process of axis extends forward to level
of dens; (25) prehensile tail present; (26) entepicondylar
notch present; (27) intercondylar fossa of femur wider
than lateral condyle; (28) tibial sesamoid bone [= prehallux] present; (29) prominent lateral tuberosity present on proximal fifth metatarsal.
Gaudin (2004a)(30) mandibular condyle hooks laterally
in dorsal view; (31) mandibular symphysis very short,
<10% of maximum mandibular length; (32) mandibular
symphysis anteroventrally inclined; (33) palate elongate
and narrow, widened at base of zygomatic processes of
35
maxilla; (34) pterygoids exposed in hard palate [modification of (3) from Engelmann (1985)]; (35) pterygoid
without hamulus or free standing descending lamina;
(36) infraorbital foramina exposed in ventral view; (37)
zygomatic process of squamosal strongly reduced, 5%
basonasal length.
Rose et al. (2005)(38) elongate, tubular skull; (39) keratinized region of stomach for grinding up ingested insects.
Phyllophaga
Engelmann (1985)(1) presence of paired perforations
in the lumbar vertebral centra; (2) presence of a large,
asymmetrically developed intravertebral [= rhachidian] vein; (3) inclusion of the optic foramen within the
opening for the sphenorbital fissure; (4) reduction of the
medial trochlea of the tibial articular surface of the astragalus.
Gaudin (1995)(5) posterior crus of ectotympanic attached
to squamosal/mastoid bridge; (6) anteroposterior length
of entotympanic greater than ectotympanic; (7) mastoid
exposed in depression between exoccipital and nuchal
crests; (8) mastoid with weak lateral exposure; (9) stylomastoid canal directed posteroventrally; (10) internal
carotid artery sulcus saddle-shaped; (11) glenoid fossa of
squamosal bounded by medial shelf; (12) entotympanic
takes the form of a vertical plate attached dorsally to the
promontorium of the petrosal, with a horizontal medial
expansion lying dorsal to the internal carotid artery; (13)
entotympanic with lateral process that contacts tympanohyal; (14) tympanohyal directed ventrally and somewhat posteriorly; (15) presence of a groove connecting
the stylomastoid foramen and posttemporal foramen.
Gaudin (1999b)(16) weak development of anapophyses in
posterior thoracic and lumbar vertebrae.
McDonald (2003b)(17) presence of caniniform teeth in
which lower tooth occludes with the posterior surface
of the upper; (18) tubercular facets on ribs concave;
(19) presence of acromiocoracoid arch (absent in extant
sloths).
Gaudin (2004a)(20) the maximum ventral extent of the
entotympanic and ectotympanic roughly equivalent; (21)
entotympanic forms the lateral wall and roof of the sulcus for the internal carotid artery, and has a medial ridge
forming at least part of the medial wall of the sulcus; (22)
teeth characterized by a large core of well-vascularized
modified orthodentine; (23) presence of a posterior external opening of the mandibular canal near the junction
of the ascending and horizontal rami of the mandible;
(24) presence of a large posteriorly or posteroventrally
directed process on the proximal end of the stylohyal;
(25) presence of a rugose palate, marked by numerous
pits and grooves; (26) presence of a large pterygoid ex-
36
posure in the roof of the nasopharynx; (27) presence of a

broad, deep descending lamina of the pterygoid, typically
with a semicircular ventral margin. [Gaudin (2004a) lists
the above features as unique characters of sloths, but lists
at least 20 more unambiguous synapomorphies of sloths
in his appendix 4]
Rose et al. (2005)(28) five upper, four lower teeth; (29) jugal with large ascending and descending processes; (30)
proximal radius circular; (31) ilium flared laterally; (32)
thorax elongated and rigid.
2
Fossil Xenarthra
4
Fossil history of sloths
H. Gregory McDonald and Gerardo De Iuliis
Resumen
Resumo
A pesar de la limitada representacin moderna, restringida a los gneros Bradypus y Choloepus, ambos
arborcolas obligatorios con una distribucin limitada
a los estratos altos de los cerrados bosques tropicales,
la historia fsil de los perezosos es extraordinariamente
rica y diversa. Como grupo, es uno de los mayores y
ms caractersticos elementos de la fauna cenozoica
de Amrica del Sur y con una amplia distribucin
geogrfica, con restos encontrados en Amrica Central y del Norte, as como en el Caribe y la pennsula
Antrtica. La ecologa de los perezosos extinguidos fue
igualmente diversa con ramoneadores de niveles altos
y bajos, pastadores y, posiblemente, omnvoros y con
modos de locomocin terrestres, semiarborcolas y
acuticos. Los hallazgos ms antiguos del Eoceno y el
Oligoceno son limitados, pero en el Mioceno temprano
el registro es bastante bueno, tanto en nmero de taxo
nes como de especimenes. Nuestro conocimiento de
los varios taxones es espordico, con algunos de ellos
compuestos por ejemplares numerosos y de mltiples
localidades y otros conocidos solamente por el material
tipo. Los perezosos de Amrica del Sur y Amrica del
Norte se extinguieron junto con el resto de la megafauna del Nuevo Mundo unos 10.000 aos atrs, aunque
sobrevivieron en las islas del Caribe hasta ya entrado el
Holoceno. La causa de su extincin ha sido atribuida
tanto a agentes humanos como al cambio climtico y
ambiental. Nuestro conocimiento de la ecologa, la distribucin y el tiempo de la extincin de los numerosos
taxones del Pleistoceno tardo necesita ser mejorado
notablemente antes de que la contribucin relativa de
alguna de estas causas pueda ser dilucidada.
Apesar da limitada representaco moderna, restrita

aos gneros Bradypus e Choloepus, ambos arborcolas
obrigados com uma distribuio limitada s fechadas
canopas das florestas tropicais, a histria fssil das
preguias extraordinariamente rica e diversa. Como
grupo, um dos maiores e mais caractersticos elementos da fauna cenozica de Amrica do Sul, que tornouse de ampla distribuico geogrfica, com restos encontrados na Amrica Central e do Norte, como no Caribe
e a pennsula Antrtica. A ecologia das preguias extinguidas foi igualmente diversa com ramoneadores
baixos e pastadores, e modos de locomoco terrestres,
semiarborcolas e aquticos. Os achados mais antigos
do Eoceno e do Oligoceno so limitados, mas no Mioceno inferior o registro bastante bom em nmero de
txons ou de exemplares. Nosso conhecimento dos
vrios txons espordico, com alguns deles compostos por exemplares numerosos e de mltiplas localidades e outros conhecidos somente pelo material tipo.
As preguias de Sul- e Norteamrica extingiram-se
junto com o resto da megafauna do Novo Mundo uns
10.000 anos atrs, embora tenham sobrevivido nas ilhas
do Caribe at comeado o Holoceno. A causa da sua
extinco tem sido atribuda tanto a agentes humanos
quanto mudana climtica e ambiental. Nosso conhecimento da ecologia, a distribuio e o tempo da
extino dos numerosos taxa do Pleistoceno superior
precisa ser grandemente melhorado antes que possa ser
estabelecida a contribuico relativa destas causas.
39
40
H. G. McDonald and G. De Iuliis
Introduction
The sloths (Tardigrada, Phyllophaga, or Folivora) are
currently represented by only two genera, the tree sloths
Bradypus and Choloepus, which inhabit New World
tropical forests, are mainly folivorous, and almost completely arboreal (Chiarello this volume). While the
modern representation of this group is small, the tardigrade fossil record is extraordinarily rich, and the group
is one of the largest and most characteristic elements of
the Cenozoic fauna of South America. Sloths were also
well represented in North America during much of the
past three million years and appear to have dispersed
into the Caribbean islands by the Oligocene. Indeed,
sloths and glyptodonts comprise perhaps the most diverse groups of xenarthrans, with nearly 100 sloth genera named (McKenna and Bell 1997). Their generally
large body size and a reduced metabolism (implied
from their modern representatives) were main reasons
for the traditional view that the fossil forms were capable only of essentially slow, quadrupedal locomotion.
In recent years, however, various new discoveries and
morphofunctional analyses have suggested more diverse lifestyles for these extinct xenarthrans, including
arboreal, fossorial, and aquatic habits.
Despite the rich fossil history of sloths, the reality is
that many of the remains are known mainly from two
time periods, the Miocene Santacrucian South American Land Mammal Age (SALMA) and the Pleistocene
Lujanian SALMA and Irvingtonian and Rancholabrean
North American Land Mammal Ages (NALMA). As
might be expected, the Pleistocene remains are abundant and widely distributed, extending from southern
Argentina and Chile north through Central and North
America, reaching as far as Alaska and the Yukon. The
abundant remains of some taxa have led to a relatively
stable taxonomy and statistical analyses have allowed
an understanding of intraspecific variation. Thus, for
example, we now recognize a single, Panamerican
megatheriine species, Eremotherium laurillardi, from
among the numerous species described from the late
Pleistocene (Cartelle and De Iuliis 1995, 2006). This is
not the case for all taxa: many are known from only
fragmentary material, with consequently questionable
taxonomy.
The early Miocene remains are mainly restricted to
the Santacrucian SALMA from Patagonia. The record
of sloths from other time periods is relatively sparse,
which has limited our ability to understand their paleobiology and systematics. Although some important
gaps remain, recent finds have begun to improve our

knowledge of sloths and permitted inferences about
their paleobiogeography and paleoecology. Prominent
among these are (1) the earliest sloths, the mainly Oligocene Pseudoglyptodon and the paragravigrades,
(2) the Miocene, mainly Santacrucian, sloths such as
Hapalops, and (3) the early, mainly Miocene and Pliocene, megatheriines.
Major anatomical features of the different

groups of sloths
General characteristics of the Xenarthra have been
discussed by Rose and Emry (1993) and McDonald
(2003b; see also Gaudin and McDonald this volume)
and need not be repeated. Rather, we will focus on the
major skeletal characteristics that distinguish the four
major groups of sloths discussed here. These four major clades, traditionally recognized as families, include
megatheres, megalonychids, nothrotheres, and mylodonts. Each group differs from the others by a distinctive suite of skeletal characters presumably reflective of
both the phylogeny and the paleoecology of its members. This review is not meant to serve as either a systematic or phylogenetic overview of the sloths, as this
has been done recently in two separate studies (Pat
terson et al. 1992; Gaudin 2004a; see Gaudin and McDonald this volume). Although we may note that we
disagree with some, but not all, of the phylogenetic
conclusions obtained in these studies, they are the most
recent and comprehensive currently available.
Megatheres
These are the largest of the extinct sloths and comparable to proboscideans in body mass. The dentition is
bilophodont, with no modification of the anterior teeth
into caniniforms. The dental series forms a continuous
row with no diastema. All the teeth are quadrate in
cross section, with the two lophs parallel. The complementary lophs on the upper and lower teeth interlock
during occlusion. The condyle of the jaw is positioned
above the tooth row. The depth of the jaw is greatly exaggerated in megatheres and is deeper in proportion to
its size than in any other group of sloths (Bargo 2001b).
The humerus lacks an entepicondylar foramen. The
distal ulna has a styloid process, but there exists a gap
between it and the ulnare (= cuneiform), which has a
complementary process suggestive of only a ligamentous connection and that the ulnare was not weightbearing (Tito and De Iuliis 2003). In the femur, the
medial condyle shares a continuous articular surface

with the patellar surface but the lateral condyle is distinct, except in the basal Megathericulus, in which they
are contiguous (De Iuliis, Brandoni, and Scillato-Yan
forthcoming). The tibia and fibula tend to fuse in megatheres but not in the other sloths. The pes is rotated and
the fifth metatarsal is flattened. The calcaneum has a
secondary ossification on the tuber calcis that increases
the functional length of the calcaneum and increases
the lever arm of the gastrocnemius.
Nothrotheres
The nothrotheres are characterized by their relatively
narrow and elongated skull. While early nothrotheres
had the anterior-most upper and lower teeth modified
as caniniforms, in later forms including Nothrotherium,
Nothrotheriops, and Thalassocnus these teeth were lost,
resulting in four upper and three lower molariforms.
Unlike the megalonychids, the caniniform is positioned
midway between the cheek teeth and the anterior part
of the maxilla. The molariforms are quadrate and bilophodont, with a distinctive groove the length of the
tooth on the labial and lingual sides. The mandibular
spout is elongated and narrow and the condyle is positioned above the occlusal plane. The humerus has an
entepicondylar foramen. In the manus, the ungual of
digit two is flattened and semicircular in cross section,
while all of the other unguals are taller and triangular. The position of the third trochanter on the femur
is variable and may be distinct (Nothrotherium), form
a continuous ridge with the greater trochanter (Thalassocnus), or be continuous with the distal lateral epicondyle (Nothrotheriops). The pes is rotated into the pedolateral position but the development of the odontoid
process of the astragalus is not as highly modified as in
megatheres and mylodonts. Only the posterior portion
of the calcaneum contacts the ground, giving the pes an
arched appearance.
Megalonychids
The megalonychids are characterized by modification
of the anterior-most upper and lower teeth into either
caniniforms or incisiforms that are separated from the
molariforms by a diastema. The upper caniniform is
always positioned at the anterior end of the maxilla adjacent to the premaxillae contact, although the premaxilla does not contribute to its alveolus. The shape of the
caniniform varies among the different genera, although
the primitive morphology appears to have been a triangular cross section with a pointed tip and an elongated
41
occlusal surface on the posterior side of the upper, with

a complementary surface on the anterior surface of the
lower. In Megalocnus, the anterior teeth are rotated and
have incisiform morphology, reminiscent of rodents,
hence the species-specific epithet, rodens. The molariforms are bilophodont but differ from the megatheres
and nothrotheres in being triangular rather than quadrate in cross section, with the two lophs converging toward the labial or lingual side of the tooth depending
on whether it is an upper or lower tooth. The condyle of
the jaw is positioned above the tooth row. The mandibular spout may be large in some forms but is reduced in
others. The smallest spout is in Megalonyx, which has
the largest caniniforms of all megalonychids.
The humerus usually has an entepicondylar foramen.
The femur has a small third trochanter positioned at the
midpoint of the diaphysis in most forms. In the femur,
the articular surfaces of both distal condyles are separate from the patellar surface. The tibia and fibula are
separate. The pes is not rotated as in other sloths and
is more plantigrade, the astragalus lacks the enlarged
medial trochlea. The tuber calcis of the calcaneum is
greatly enlarged as a result of a mediolateral expansion.
Mylodonts
This appears to be the only group of sloths that became
adapted to grazing, or at least to consuming a significant percentage of grasses in their diet. Currently, two
major subdivisions within the family Mylodontidae are
recognized, the scelidotheres and the mylodontines,
although some earlier forms represented by Pseudoprepotherium and Acremylodon represent a distinct
subset whose phylogenetic relationships to the other
mylodonts have not yet been resolved. Traditionally,
this group of sloths was distinguished by having lobate teeth with flat occlusal surfaces, rather than the
bilophate surfaces seen in megatheres, nothrotheres,
and megalonychids. This definition has been modified
because in some earlier mylodonts, including Pseudoprepotherium, Acremylodon, and Glossotheriopsis, the
only lobate teeth are the last ones in the dental series.
In these genera the anterior molariforms are circular or
oval in cross section and their occlusal surface consists
of a single transverse loph; they are thus readily distinguished from megatheres, nothrotheres, and megalonychids. In primitive mylodonts with transverse lophs,
the condyle is positioned above the occlusal plane, but
in those forms with flat occlusal surfaces, the condyle is
lower. The mandibular symphysis forms a long and nar-
42
row spout in the scelidotheres and primitive mylodontines, but is generally short and broad in the advanced
mylodontines.
While the scelidotheres and primitive mylodontines
retain an entepicondylar foramen on the humerus, this
feature is absent in advanced mylodontines. Both the
radius and the tibia are considerably shorter than the
humerus and femur, respectively, much more so than in
the other sloths. In the scelidotheres, the radius is similar to that of megalonychids and nothrotheres in that
the distal end is offset anteriorly relative to the proximal
end. In contrast, in the mylodontines the two ends of
the radius are in the same axis. In the femur, the articular surfaces of the distal condyles are continuous with
the patellar articular surface. The pes is rotated in all
mylodonts, but in scelidotheres it more closely resembles nothrotheres in that the lateral trochlear surface of
the astragalus is curved and only the posterior edge of
the calcaneum contacts the ground. The mylodontines
are more similar to megatheres, with a flattened lateral
trochlear surface, and the entire ventral surface of the
calcaneum contacts the ground.
Biostratigraphy/chronology
While the overall fossil record of sloths is rich (table
4.1), their early history is uncertain. The oldest probable record is that of a caniniform tooth recovered from
the Eocene La Meseta Formation, composed of nearshore sandstones and siltstones, on Seymour Island, 100
km SE off the northern tip of the Antarctic Peninsula
(Vizcano and Scillato-Yan 1995; Vizcano et al. 1997).
Paleobotanical studies suggest that the terrestrial environment during the Eocene was heavily vegetated
Nothofagus-podocarp forest growing in a humid, temperate or cold-temperate climate, similar to presentday southern South America. This record suggests that
Antarctica played at least a minor role in tardigrade
history. Ameghino (1902, 1905) described several specimens based on fragmentary material from the early and
middle Eocene of Patagonia that may represent sloths,
such as the Casamayoran SALMA (early Eocene) Protobradys and the Mustersan SALMA (middle Eocene)
Proplatyarthrus (see Carlini and Scillato-Yan 2004;
Pujos and De Iuliis 2007).
Oligocene (Tinguirirican, Deseadan SALMA) sloths
The earliest unquestionable South American Tardigrada are from late Oligocene beds (Deseadan SALMA)
of the La Flecha locality, Patagonia (Hoffstetter 1982)
and the Salla-Luribay locality, Bolivia (Engelmann

1987). These early sloths were originally referred to as
the orophodontoids, and include Orophodon (figure
4.1A) and Octodontotherium (figure 4.1B). Somewhat
chronologically later remains have also been considered orophodontoids (see Santos et al. 1993; Villarroel
2000; Carlini and Scillato-Yan 2004). Traditionally,
orophodontoids were viewed as an early experiment
in sloth evolution, with markedly lobate teeth and
well-developed dermal ossicles, although the association of the dentition with the ossicles is questionable.
Indeed, Hoffstetter (1956) distinguished these sloths
as paragravigrades, in contrast to the remaining fossil ground sloths, and subsequently considered them
an early lateral offshoot of the main tardigrade clade
(Hoffstetter 1982). More recently, Gaudin (2004a) and
Carlini and Scillato-Yan (2004) accommodated them
within the Mylodontoidea, considering them probably
as basal members of this clade, in which lobation of the
teeth is a common feature.
Two features of orophodontid dentition suggest these
early tardigrades may constitute a particular group of
sloths. First is the central portion of harder dentine,
much as in cingulate glyptodontids and most pampatheriids, and second is the wear pattern of the occlusal
surfaces of the more distal teeth. In the latter, wear is
such that the distal part of the tooth is at a considerably
different level than the mesial part. Unfortunately, these
sloths are not particularly well represented. Indeed, except for Octodontotherium, there is no certainly associated cranial and postcranial material, and the cranial
material is typically sparse and fragmentary.
Recent morphofunctional analyses (review in Vizcano et al. this volume) have advanced our understanding of diet and masticatory functions for various
xenarthran groups. Such analyses, however, require
reasonably complete cranial remains. Thus, additional
material from the early sloths is required before such
studies can be applied to them. Nevertheless, the remains we do have hint at masticatory functions and
dietary modes that are distinct from those proposed
for other fossil xenarthrans. Further remains are also
required before the taxonomy and phylogeny of these
early sloths can be resolved. One promising region for
obtaining such material is the Salla-Luribay beds in
Bolivia. Further excavation is suggested there, in the
already fossiliferous levels as well as in those lower in
the sequence.
Pseudoglyptodon (figure 4.1C) is a singular pilosan.
This Oligocene form, known mainly from mandibular
Table 4.1. Geographic and Biochronologic Distribution of Sloth Genera

Time Period
North America
Antillean
Central America
Holocene
Choloepus

Rancholabrean/
Megalonyx
Megalocnus
Eremotherium
Lujanian
Nothrotheriops
Neomesocnus
Indet. mylodont

Eremotherium
Neocnus

Paramylodon
Parocnus

Miocnus

Acratocnus

Synocnus

Habanocnus

Paulocnus

Ensenadan

Irvingtonian/
Megalonyx
Uquian
Nothrotheriops

Eremotherium

Paramylodon
Blancan/
Megalonyx
Meizonyxa
Chapadmalalan
Eremotherium

Glossotherium

Montehermosan

Hemphillian/
Megalonyx
Huayquerian
Pliometanastes

Thinobadistes

South America
Choloepus
Bradypus
Xenocnus
Nothrotherium
Nothropus
Ocnopus
Megatherium
Eremotherium
Scelidotherium
Catonyx
Valgipes
Mylodon
Mylodontopsis
Glossotherium
Ocnotherium
Oreomylodon
Lestodon
Megalonychops
Nothrotherium
Nothropus
Megatherium
Eremotherium
Scelidotherium
Catonyx
Glossotherium
Lestodon
Synhapalops
Megatherium
Scelidotherium
Glossotherium
Diheterocnus
Pronothrotherium
Thalassocnus
Scelidotherium
Proscelidodon
Glossotherium
Pronothrotherium
Thalassocnus
Plesiomegatherium
Proscelidodon
Amphiocnus
Megalonychops
Neohapalops
Paranabradys
Pliomorphus
Pronothrotherium
Thalassocnus
Plesiomegatherium
Pyramiodontherium
Proscelidodon
Ranculcus
Sphenotherus
Megabradys
Acremylodon
Urumacotherium
continued
44
Table 4.1.Continued
Time Period
North America
Antillean
Central America
Chasicoan

Laventan

Colloncuran
Friasian

Santacrucian
Imagocnus

Colhuehuapian

Deseadan
Indet. megalonychidb

Tinguirirican
Divisaderan
Casamayoran
South America
Chasicobradys
Xyophorus
Octomylodon
Eucholoeops
Hapalops
Huilabradys
Indet. megathere
Pseudoprepotherium
Neonematherium
Glossotheriopsis
Glossotheriopsis
Amphibradys
Megathericulus
Planops
Sibyllotherium
Eucholoeops
Megalonychotherium
Hapalops
Hapaloides
Schmismotherium
Pelecyodon
Parapelecyodon
Analcimorphus
Hyperlleptus
Planops
Nematherium
Analcitherium
Holomegalonyx
Chubutherium
Deseadognathus
?Hapalops
Orophodon
Octodontotherium
Chubutherium
Pseudoglyptodon
Pseudoglyptodon
Indet. megalonychidc
Octodontobradys
Notes: Where two time periods are listed, the first is the North American Land Mammal Age (NALMA) and the second is the South
American Land Mammal Age (SALMA). Single listings are from the SALMA. Within each time period, genera are listed in order of their
appearance in the fossil record.
a. Specimen from El Salvador.
b. Specimen from Cuba.
c. Specimen from Antarctica.
material described by Engelmann (1987), is aptly described as a glypto-sloth because its dentition and
jaw morphology are a mix of glyptodontoid and sloth
features. The dental formula (C1, M4/c1, m3) is clearly
slothlike, with a large caniniform followed by three molariforms, as is the robust, ventrally bowed dentary. The
molariforms, however, are clearly glyptodontoid-like in
being trilobate. The phylogenetic position of Pseudoglyptodon is unclear and Engelmann (1987) considered
it a sloth of uncertain affinities. It is currently known
from the Deseadan SALMA of Bolivia and early Oligocene Tinguirirican SALMA of Chile (Wyss et al. 1993;
Flynn et al. 2003). Additional material, such as the cranial remains under review by B. Shockey and F. Anaya
Daza, might help clarify evolutionary relationships and
provide answers to questions such as whether the molariforms are convergent with those of glyptodontids,
or whether Pseudoglyptodon might represent an intermediate form between cingulates and pilosans (see Pujos and De Iuliis 2007).
45
B.
A.
C.
D.
Figure 4.1. Oligocene (Deseadan) sloths. A. Orophodon. B. Octodontotherium. C. Pseudoglyptodon. D. Deseadognathus. A and B
after Hoffstetter 1954, C after Engelmann 1987, D after Carlini and Scillato-Yan 2004.
The earliest documented megalonychid is Deseadognathus (figure 4.1D) from the late Oligocene Deseadan
(2924.5 mya) of Argentina (Carlini and Scillato Yan
2004). Currently known only from the mandible, the
specimen displays all the essential morphology seen in
later members of the family.
Another intriguing Oligocene sloth is an unnamed
genus and species of megalonychid found near the
town of Yauco in Puerto Rico (MacPhee and IturraldeVinent 1995). The record is based on the proximal end
of a femur, but retains sufficient morphology to allow
recognition as a sloth (assignment to the Megalonychidae is tentative but reasonable). The age of the Juana
Daz Formation from which the specimen was recovered is based on the presence of the foraminiferan,
Globigerina ampliapertura, which occurs from 3334
mya. While based on limited material, the discovery is
critical in two ways. First, it would represent the oldest
record of a megalonychid, predating Deseadognathus.
Its presence in the Caribbean indicates the group had
not only originated by the Oligocene, but must have
already been widespread in South America in order for
46
one member to disperse to this location. If true, this

clearly illustrates the major gaps that still exist in our
knowledge of the early history of most sloth clades. Second, it demonstrates that sloths had dispersed into the
Antillean islands and were established there long before
they entered North America in the late Miocene. Consequently, the entrance of the megalonychid Pliometanastes into North America was a separate dispersal event
that may or may not have included a second dispersal
into the Caribbean. A second dispersal from North
America would support the phylogenetic interpretation
of the Antillean sloths by White and MacPhee (2001),
that is, that there are two distinct lineages of megalonychids and their taxonomy reflects the independent
entrance of two different groups into the Antilles at
different times, as opposed to in situ divergence. The
recovery of additional material will be critical to addressing this question.
Miocene (Colhuehuapian to Huayquerian SALMA)
sloths
Sloths from the earliest Miocene include an enigmatic
form, Chubutherium, from the Colhuehuapian of
Chubut Province, Argentina (Cattoi 1962). Currently
known from only three skeletal elements (femur, tibia,
and astragalus), this genus is considered to be an early
scelidothere because of the astragalus, in which the articular surface for the cuboid is a large concavity. The
megalonychid Holomegalonyx is also known from this
site (Scillato-Yan 1979).
The earliest large assemblage of fossil sloths is that
recovered from the Miocene, particularly the Santacrucian SALMA of Patagonia. These remains have been
known for more than a century, mainly through the
work of Scott (19031904). Most specimens were referred to the Megalonychidae, with the most abundant
remains assigned to the genus Hapalops (figure 4.2A).
The long-standing confusion on the phylogenetic significance of these sloths (De Iuliis 1994) may be almost certainly traced back to Scotts (19031904, p. 161)
comments that the Santa Cruz ground-sloths appear
to have great phylogenetic significance; and forerunners, if not the actual ancestors of almost all the great
Pleistocene genera of North and South America, may
be observed in this fauna. Scott expressed some reservation about whether these sloths are or are not the
ancestors sought for but nonetheless asserted that for
all practical purposes, however, we may regard these
Santa Cruz genera as the actual ancestors, for they must
at least represent those ancestors in all the essentials
of structure. The recent recognition of the late Miocene and younger nothrotheres as a distinct family
resulted in the creation of the subfamily Schismotheriinae (McKenna and Bell 1997), which includes many
of the eight Santacrucian genera, including Hapalops,
all formerly considered early nothrotheres. No analysis
of this group to determine if it is truly monophyletic
has been undertaken and it has not been recognized in
more recent studies (Gaudin 2004a; Gaudin and McDonald this volume). Genera currently considered true
megalonychids in the Santacrucian are now restricted
to Eucholoeops (figure 4.2B) and Megalonychotherium.
In addition to the schismotheres and megalonychids,
the Santacrucian has two mylodontids, Nematherium
and Analcitherium. The largest of the Santacrucian
sloths were Planops (figure 4.2C) and Prepotherium,
which have been placed in their own subfamily, the
Planopsinae, a group that, because of their large size
compared to the other Santacrucian sloths, traditionally has been considered ancestral to the megatheres
(but see De Iuliis 1994).
It is mainly from Scotts views that xenarthran paleontologists clung for a long time to the idea that the
three main sloth cladesthe Mylodontidae, Megalonychidae (including the Nothrotheriidae), and Megatheriidae (traditionally regarded as families)developed
from Miocene forms, and that Hapalops could be considered as having a central position among Miocene to
Pleistocene sloths. De Iuliis (1994) noted this tendency
and argued it was unfounded. Among other things, it
ignored sloth remains from the Oligocene, discussed
above, relegating them for convenience (as is often the
case when material is not sufficiently well known) into
their own early side branch. More recent analyses
suggest the earlier sloths (except for the very unusual
genus Pseudoglyptodon) may well fall within the traditional lineages, and that the Santacrucian sloths are
not central. This is important because of the tacit assumption that early megalonychid morphology is the
basic form among sloths (e.g., the relatively simplified
caniniforms and molariforms and the low astragalus,
without a raised, peg-shaped odontoid process), a supposition that ignores features such as the complex trilobate dentition of Pseudoglyptodon and the pear-shaped
astragalus in Octodontotherium.
The Miocene Santacrucian sloths are nonetheless
a very important assemblage because they represent
the first major radiation among sloths. They therefore
have much to tell us about sloth evolution and diversity.
Unfortunately, despite an abundance of remains, little
47
C.
A.
D.
B.
Figure 4.2. Early and middle Miocene (Santacrucian and Collocuran) sloths. A. Hapalops. B. Eucholoeops. C. Planops. D. Megathericulus. AC after Scott 19031904.
has been accomplished over the last century in terms

of taxonomy. The reason for this is that the available
remains have not been studied as a whole. For example,
Mones (1986) listed 26 species of Hapalops, many of
which were from the same region. Twenty-one of these
species were named by Ameghino (see Mones 1986, pp.
24849 for details), mostly on partial or fragmentary
specimens. Scott (19031904) did attempt a synthesis
of the numerous specimens (for example, in synonymizing several of Ameghinos species), but he added
four new species himself. Clearly, there are differences
among the remains so that we would expect several
valid species. However, a review is clearly required,
particularly in view of the wide range of intraspecific
variation established for other sloth species (Stock 1925;
McDonald 1988, 1995; Cartelle and De Iuliis 1995, 2006;
De Iuliis 1996; Pujos 2002; Pujos et al. 2002; Pujos and
Salas 2004a).
The first certain megatheriines, represented by Megathericulus (figure 4.2D), appear somewhat later during
the Colloncuran SALMA (middle Miocene). Several
taxa have been named from the time between the appearance of Megathericulus and the taxa characteristic
of the Plio-Pleistocene (Megatherium and Eremotherium). Most of these are known from fragmentary
material of uncertain age and thus their taxonomy has
been particularly unstable (and so have been largely ex-
cluded from table 4.1). They include the following Miocene-Pliocene taxa from Argentina: Eomegatherium
(Mayoan SALMA of Patagonia), Plesiomegatherium
hansmeyeri (Chasicoan SALMA of Buenos Aires Province), P. halmyronomum (Jujuy Province), and the
suite of taxa known essentially from Entre Ros Province, presumably later Miocene to Pliocene in age, such
as Promegatherium, Megatherium antiquum, Eomegatherium nanum, and Pliomegatherium. Indeed, little
work has been published on these taxa since their first
descriptions (other than their repeated appearance in
faunal lists), although De Iuliis (1996) attempted a recent synthesis. De Iuliis et al. (2004) provided a partial
solution to the status of Plesiomegatherium, and the taxonomy and characteristics of Pyramiodontherium (figure 4.3A) were largely resolved by Carlini et al. (2002),
Brandoni et al. (2004), and De Iuliis et al. (2004). For
most of these taxa, however, additional material and a
reassessment of the provenance of the original material
(when possible) are required before much progress can
be made on this front. The Entre Ros megatheriines are
particularly problematic, but recent efforts by D. Brandoni may provide a much improved understanding of
these forms. Megatheriops is known from the Huayquerian of Mendoza Province, Argentina.
The family Nothrotheriidae sensu stricto appears in
the latest Miocene. One of its members, Thalassocnus
48
(figure 4.3B), known from the Mio-Pliocene Pisco Formation of Peru, has been interpreted as having evolved
to live in a marine environment and was at least semiaquatic in its habits (Muizon and McDonald 1995). An
evolutionary series of the Thalassocnus lineage is preserved in the Pisco Formation, with five recognized
species that demonstrate changes in the skull, dentition,
and postcranial skeleton from a terrestrial ancestor to
an aquatically adapted descendent (McDonald and
Muizon 2002; Muizon et al. 2003, 2004a,b). The diet of
the animal is interpreted as sea grass. The lineage became extinct in the Chapadmalalan and its disappearance seems to coincide with the final formation of the
Isthmus of Panama and the creation of the colder Humboldt Current along the west coast of South America.
Late Miocene mylodonts are represented by Pseudoprepotherium (figure 4.3C), Acremylodon, Urumacotherium, and Glossotheriopsis. While the last upper
and lower molariforms of these taxa are bilobate, they
are distinguished from all other mylodonts by having
molariforms with a single transverse loph. There is a
difference of opinion as to whether the first three genera are distinct or all represent the same taxon. Glossotheriopsis is distinguished from the others by having
the first tooth modified into a caniniform. Other late
Miocene mylodonts such as Sphenotherus, Ranculcus,
and Megabradys have molariforms with flat occlusal
surfaces seen in later mylodonts.
The dispersal of sloths into North America occurred
in the latest Miocene with the appearance of a megalonychid, Pliometanastes, and a mylodont, Thinobadistes
(figure 4.3D), in the Hemphillian. The Isthmus of Panama had not formed at this time, so it is presumed that
both taxa were able to disperse across some type of water barrier, perhaps by island hopping. The appearance
of these two sloths in North America is used in defining
the Clarendonian-Hemphillian NALMA (early late/late
late Miocene) boundary. It seems their appearance in
North America was essentially simultaneous. A partial
skeleton of Pliometanastes from the Siphon Canal, California (Mehrten Formation), has been dated at 8.19
0.16 mya (Hirschfeld 1981). The earliest appearance of
Thinobadistes is from the type locality, Mixons Bone
Bed, Florida, and considered to be about 8 mya. Both
genera are confined to the Hemphillian. Thinobadistes
became extinct without issue, but it appears Pliometanastes gave rise to Megalonyx.
While Pliometanastes is predominantly confined to
the early Hemphillian, there is a record of the genus
from the late Hemphillian Box T locality, Texas, dated

at 6.1 mya. Its descendent, Megalonyx, first appears in
the late Hemphillian, 6.7 mya, in the Lemoyne fauna of
Nebraska (Leite 1990). It is difficult to distinguish the
postcranial skeletons of Pliometanastes and Megalonyx;
positive identification to genus requires the preservation of the caniniform tooth. As a consequence, there
are numerous Hemphillian faunas that contain an indeterminate megalonychid.
Recent work in Cuba has resulted in the recovery
of terrestrial vertebrates from the late early Miocene
Lagunitas Formation near Domo de Zaza (MacPhee
and Iturralde-Vinent 1995; MacPhee et al. 2003). Along
with a megalonychid sloth, Imagocnus zazae, the fauna
includes an isolobodontine capromyid rodent, Zazamys
veronicae, and a platyrrhine primate, Paralouatta marianae, all derived from South American ancestors. The
age of the formation is considered Burdigalian (16.1
21.5 mya), based on the presence of marine invertebrate
taxa that correspond to the late early Miocene Miogypsina-Soritiidae zone. This makes the fauna roughly
equivalent to the Santacrucian SALMA. The amount
of material is slightly greater than the Oligocene record in Puerto Rico and includes a partial palate, some
molariforms, diaphysis of a humerus, a fragmentary
pelvis, and a proximal caudal vertebra. The phylogenetic relationship of Imagocnus to the later Quaternary
megalonychids is not clear, but it seems to more closely
resemble the Megalocninae. It is also possible that more
than one taxon is represented in the limited sample, but
this cannot be resolved at present.
Pliocene (Montehermosan, Chapadmalalan SALMA)
sloths
By the earliest Pliocene some sloth genera (e.g., Glossotherium, Scelidotherium, and Megatherium) appear that
continue into the Pleistocene, and it is at this time one
might consider that the sloth fauna starts to resemble
that of the late Pleistocene. The nothrotheres are represented by Pronothrotherium, which is present in both
the Montehermosan and Chapadmalalan and is distinguished from later forms by the retention of a caniniform, and the aquatic sloth, Thalassocnus, which has lost
the caniniform. The scelidothere, Proscelidodon, is also
known only from these two land mammal ages; later
forms are slightly larger than earlier ones. Diheterocnus
is the only known megalonychid and is restricted to the
Chapadmalalan. The Montehermosan is characterized
by three megatheres, Plesiomegatherium, Pyramiodon-
A.
49
C.
B.
D.
Figure 4.3. Late Miocene (Laventan to Huayquerian/Hemphillian) sloths. A. Pyramiodontherium. B. Thallasocnus. C. Pseudoprepotherium. D. Thinobadistes. B after Muizon and McDonald 1995, C after Hirschfeld 1985, D after Webb 1989.
therium, and the first appearance of Megatherium, but

none of these are known from the Chapadmalalan.
During the Pliocene, the Isthmus of Panama finally
provided a connection between South and North America and initiated the second major phase of the Great
American Biotic Interchange (GABI, Webb 1985b). In
temperate North America, many South American taxa
first appear around 2.7 mya, including the sloths Glossotherium and Eremotherium (figure 4.5D). The appearance of Eremotherium in North America in the Blancan
actually predates its earliest appearance in South America, which is not until the Ensenadan. Recent work in
the state of Guanajuato, Mexico, has documented the
appearance of South American taxa, including the sloth
Glossotherium, earlier than 2.7 mya, at around 4.74.8
mya (the Hemphillian-Blancan boundary, Flynn et al.
2005). Megalonyx is also present in the fauna but, as
previously discussed, had already evolved in situ from
its predecessor, Pliometanastes.
The fossil record of Central America is very poorly
known with regard to vertebrates. One exception is a
large megalonychid sloth, Meizonyx, from El Salvador
in deposits considered to be Pliocene (Blancan equivalent, Webb and Perrigo 1985). This is a large animal, but
is currently known only from the type specimen, half of
a lower jaw. The site also produced a second small megalonychid sloth, described as Megalonyx obtusidens.
These finds illustrate the great potential for new discoveries in the tropical parts of North and Central America
and the fact that we still have much learn about sloth
diversity.
Pleistocene (Uquian, Ensenadan, Lujanian SALMA)
sloths
Sloth diversity during the Pleistocene is greater than
during any other time in the Cenozoic except for perhaps the Santacrucian and Huayquerian. This diversity
reflects the greater number of sites and availability of
material that has been more intensively studied. Lower
sloth diversity during other land mammal ages is most
likely a taxonomic artifact, reflective of the more limited deposits of those ages and the fact that their faunas
have not been examined as thoroughly. If so, then it
does not seem there are any major patterns of increase
or decrease in taxonomic diversity in the group. We
would certainly caution that any study of changes in
sloth diversity through time is severely handicapped by
the lack of comprehensive alpha-level taxonomic syntheses for many groups.
While our knowledge of earlier sloths is definitely
geographically biased toward Argentina, at least by the
late Pleistocene the record is more balanced, with some
taxa found throughout the continent. This provides us
with a more reasonable interpretation of the actual dis-
50
C.
A.
D.
B.
Figure 4.4. Pleistocene (Lujanian) sloths from South America. A. Megatherium. B. Mylodon. C. Scelidotherium. D. Nothrotherium.
B after Reinhardt 1879.
tribution of these taxa. For example, the distribution of

Eremotherium as a tropical lowland form, compared to
the more temperate and high elevation distribution of
Megatherium (figure 4.4A), probably accurately reflects
differences in their ecology. Several new taxa, mainly
megatheres, have been recorded from the northwestern part of the continent (St. Andre and De Iuliis 2001;
Pujos 2002, this volume) and provide insights into the
distribution and possible migratory routes of these
sloths.
Another widespread form is Mylodon (figure 4.4B).
Although it has the most southern distribution of all
sloths, reaching 5135'S, it also ranges as far north as
Bolivia and is the one genus for which there appears
to be reasonable evidence that it was hunted by early
man in South America (Jackson 2003). Other taxa such
as the scelidotheres, Scelidotherium (figure 4.4C) and
Catonyx (= Scelidodon) are equally well represented.
Scelidotherium is restricted to the Pampas region of
Argentina, but Catonyx is more widespread, with spe-
cies both on the western side of the Andes and in the

tropical regions of Brazil. Other mylodonts include
Glossotherium and Lestodon, with Glossotherium having the wider distribution. Smaller sloths such as Nothrotherium (figure 4.4D) are known from multiple localities, although not as many as the larger sloths. This
relative rarity might represent a more limited habitat
preference and thus smaller distribution, or it could be
an artifact of their smaller size, making them more often overlooked. As might be expected, our knowledge
is biased toward sites in temperate regions. Sloths are
not common or well documented in the tropical parts
of South America, but records do exist, such as that of
the megalonychid Xenocnus from Brazil (Paula Couto
1980) and the megathere Eremotherium (figure 4.5D,
Cartelle and De Iuliis 1995).
By the Pleistocene, sloths had become well integrated into the North American mammalian fauna and
are often encountered in local faunas. Most of these
include only one or perhaps two sloths, indicative of
A.
51
C.
B.
D.
Figure 4.5. Pleistocene (Rancholabrean) sloths from North America. A. Megalonyx. B. Paramylodon. C. Nothrotheriops.
D. Eremotherium. B and C after Stock 1925.
the ecological distinctiveness of each genus and their

preferences for different types of habitat (McDonald
1996, 2003a). The record is fairly robust and inferences
about the distributions of Megalonyx (figure 4.5A),
Paramylodon (figure 4.5B), and Nothrotheriops (figure
4.5C) are based on large samples, although the sample
for Eremotherium (figure 4.5D) is much smaller. Megalonyx and Paramylodon were the most widespread
of the four genera, ranging from coast to coast, with
Megalonyx extending its range as far north as the Yukon of Canada, above the Arctic Circle (McDonald et
al. 2000). Nothrotheriops had a more restricted range,
although during the Irvingtonian its distribution did
extend from coast to coast, only to become restricted to
the western and southwestern United States and northern Mexico by the Rancholabrean (McDonald 1995).
Of all the North American sloths, Eremotherium laurillardi had the most restricted distribution, at least in
the United States. It is found only along the Gulf Coast
through Florida and along the Atlantic Coast, reflecting
its preference for subtropical to tropical habitat. However, it is one of the most widespread of all sloths because its range did extend southward through Mexico
and Central America into southern Brazil (Cartelle and

De Iuliis 1995, 2006).
Our understanding of the diversity, both taxonomic
and adaptive, of the Pleistocene sloths of the Caribbean
has greatly improved in recent years, particularly with
the recent studies in Hispaniola (White and MacPhee
2001). It appears that the radiation of megalonychid
sloths in the Caribbean resulted in two major clades,
represented by Megalocnus (figure 4.6A) and Acratocnus (figure 4.6B). One of the major recent discoveries
regarding this group is the post-Pleistocene survival of
numerous taxa into the Holocene (see below).
Paleoecology
Both extant genera of sloths, and presumably all fossil sloths, are herbivores, although Faria and Blanco
(1996) recently proposed that Megatherium may have
been carnivorous or at least scavenged carcasses like
some modern armadillos. The recent discovery of dung
referred to Megatherium (Carretero et al. 2004) supports the anatomical evidence that Megatherium was
an herbivore and a browser, feeding on Fabiana peckii,
52
B.
A.
Ephedra breana, Acantholippia seriphioides, Junellia

seriphioides, and Chuquiraga ruscifolia.
The diet of another browser, the North American
genus Nothrotheriops, has been well documented based
on the recovery of its dung from numerous caves in the
southwestern United States. Analyses of these samples
consistently show that the animal fed on desert vegetation (Laudermilk and Munz 1934, 1938; Thompson et
al. 1980). More than 72 genera of plants were identified in dung from Rampart Cave, Arizona (Martin et
al. 1961; Hansen 1978), indicating that while the diet of
Nothrotheriops was dominated by xeric plants, it fed on
what was available, including more mesic vegetation.
The advent of DNA analysis has expanded our ability to
identify plants in sloth dung beyond macroscopic identification, and when combined with AMS radiocarbon
dating, permits us to track changes in diet through time
(Hofreiter et al. 2000).
Studies of the dung of Mylodon darwini from the
cave at Ultima Esperanza in Chile support the interpretation that some mylodont sloths were grazers, as
the dominant vegetation has been identified as grasses
and sedges (Moore 1978). Based on skull morphology,
Naples (1989) suggested that Paramylodon harlani from
Rancho La Brea, California, may have been an intermediate, browsing as well as grazing. Recent stable isotope studies produced a 15N value in P. harlani that was
more positive than equids and true nonruminants but
less than that in ruminants, indicating that the sloth
could be considered a foregut fermenting nonruminant mammal (Coltrain et al. 2004, p. 209). A similar
study using bone collagen of Eremotherium laurillardi
from near Itaituba, Brazil, had a value of -26.9. This indicates that the animal was feeding on C3 plants, possibly from upper canopy trees, indicative of its habits
Figure 4.6. Late Pleistocene sloths from the Caribbean. A.

Megalocnus. B. Acratocnus. A after Matthew and Paula Couto
1959, B after Anthony 1926.
as a high browser (Rossetti et al. 2004). The increased

utilization of stable isotopes analysis should provide
additional dietary information on taxa for which we do
not have dung, and provide a better understanding of
the groups dietary diversity.
Recently, DNA analysis identified a sample of dung
from Cuchillo Cur in Neuqun Province, southwest
Argentina, as produced by a sloth. Plants in the dung
indicate that the animal was a browser and fed on Senecio as well as members of Sapindales, Lamiales, Fabales, Malpighiales, and Apiales (Hofreiter et al. 2003).
Based on the recovered DNA sequence, the sloth was
probably a nothrothere, but the location is outside the
known range of late Pleistocene nothrotheres. Additionally, there is currently no DNA sequence from
any of the known South American nothrotheres upon
which to base a comparison with the DNA sequence
obtained from the dung that might permit assignment
to a genus, let alone a species. We are therefore left with
the rather unusual situation of knowing the diet of an
extinct sloth, but not which one.
In addition to dietary information, preservation
of sloth dung in arid environments has permitted the
identification of parasites. The earliest study was by
Ringuelet (1957), who identified nematode eggs in the
dung of Mylodon darwini. Dung of Nothrotheriops shastensis from Rampart Cave contained nematode juveniles, helminth eggs, and coccidian cysts and included
Agamofilaria oxyura, Strongyloides shastensis, and two
species of Archeococcidia (Schmidt et al. 1992).
Aside from dung, many dry caves also preserved
soft tissues including skin, hair, ligaments, intervertebral discs, and the claws of ungual phalanges. Soft tissue remains are currently known from Nothrotheriops
shastensis and Megalonyx jeffersonii in North America
and Mylodon darwini in South America. Samples from

both locations include pieces of skin with hair, indicating fossil sloths were, like their modern counterparts,
covered with long coarse hair. Recently, Rodrguez and
Hernndez (1992) did a biochemical analysis of bone
samples of the Cuban ground sloth Megalocnus rodens
and were able to recover caprylic, caprilic, stearic, and
linoleic fatty acids, and amino acids including proline,
leucine, and tyrosine. Tyrosine is involved in hair pigmentation and allowed the inference that the hair color
was tawny or dark brown, while the stearic acid was
indicative of the animal being an herbivore.
Other studies of fossil sloth hair include those by
Ridewood (1901) for Mylodon darwini, and Hausman
(1936) for Nothrotheriops shastensis. Thermal conductance of hair from both species was higher than expected (McNab 1985). However, Nothrotheriops hair
had a much higher conductance than did that of Mylodon (McNab 1985). Based on the proportional abundance of the amino acid hydroxyproline in bone collagen fibers preserved in the asphalt at Rancho La Brea,
California, Ho (1967) calculated the core body temperature for Nothrotheriops as 34.4C. This is considered
very low for an animal with an estimated body weight
of 250 kg. The combination of low body temperature
and high thermal conductance suggests that, like modern sloths, Nothrotheriops was thermally sensitive and
cooler temperatures may have limited its distribution.
Mylodon was much larger in body mass than Nothrotheriops. McNab (1985) argued that larger body size and
lower thermal conductance permitted the mylodonts
to occupy seasonally cool to cold climates not available
to smaller sloths. This may explain why Mylodon was
able to extend its range into the southernmost parts of
South America.
Many Pleistocene sloths have been recovered from
cave deposits in South America, North America, and
the Caribbean, so it is likely that natural caves and
rock shelters played an important role in their ecology.
In those areas where this resource was limited, some
sloths may have created them by excavating burrowlike
structures (Vizcano et al. this volume).
Morphofunctional studies utilizing shape indices
of the elbow, hip, and knee joints have shown that not
all ground sloths stayed on the ground. In fact, the
group displays a variety of locomotor habits, including
arboreality, semiarboreality, and terrestriality (White
1993). For example, the small Antillean megalonychids Neocnus and Acratocnus appear to have been primarily arboreal. The divergent skeletal morphology of
53
Thalassocnus, indicative of its aquatic habits (Muizon

and McDonald 1995), further demonstrates the previously unappreciated diversity of locomotor abilities
in sloths and strengthens Whites observation that the
traditional distinction between tree sloths and ground
sloths is invalid. A more complete understanding of
morphological variation in fossil sloths, and its implications concerning locomotor modes, may shed light
on the evolution of the suspensory posture of extant
sloths.
Extinction
Despite the number of sloth genera present in the late
Pleistocene of South America (14), North America (4),
and three islands in the Caribbean (8), very few have
been directly radiometrically dated. Consequently,
while we know that sloths, along with much of the
New World Pleistocene megafauna, became extinct approximately 10,000 years ago, we have very little data by
which to examine the exact timing and pattern of this
extinction event in order to understand the role of climate or environmental change, as opposed to humans,
as a factor in their extinction.
North America
The 14C dates on surface dung of Nothrotheriops from
Rampart Cave in the Grand Canyon range from 10,400
275 to 11,480 200 radiocarbon years before present
(rcybp), while the oldest date is >40,000 rcybp at a
depth of 130 cm. Other dates in this time range are also
available for this species from other cave sites in the
southwestern United States. There are only two published 14C direct dates for Megalonyx. Four very similar
determinations on bone collagen, from a jaw found at
Lang Farm, Illinois, range from 11,430 60 to 11,710
80 rcybp, representing the youngest date for the genus
(Schubert et al. 2004). Chatters et al. (2004) reported a
date of 12,130 50 rcybp for a specimen from Bishop
Ranch in central Washington. There are a few published
dates for Paramylodon harlani from Rancho La Brea,
but not from any of the other North American records
of this taxon. There are no dates for North American
late Pleistocene specimens of Eremotherium laurillardi.
South America
Radiometric dates for Mylodon darwini from Mylodon Cave, Chile range from 10,200 400 to 12,270
350 rcybp. In addition to the indeterminate sloth
54
from Cuchillo Cur, Argentina, which has been dated

at 14,665 150 rcybp, there is a second indeterminate
form from Gruta del Indio, Argentina, for which a
number of dates are available, ranging from 10,200
300 to 24,730 860 rcybp. Until specific taxonomic
identifications can be made on these latter records, they
cannot be fully utilized to better understand either sloth
paleoecology or extinction. While both of these indeterminate sloths appear to be nothrotheres, the only
direct 14C date is 12,200 120 rcybp for Nothrotherium
maquinense from Gruta de Brejes, Brazil (Czaplewski
and Cartelle 1998). There is a single direct AMS 14C date
of 11,340 50 yr rcybp, based on purified collagen, for
Eremotherium laurillardi from the Itaituba Quarry, Par
State, Brazil, a humid tropical lowland setting (Rossetti
et al. 2004). The only directly 14C-dated remains of a
scelidothere sloth, Catonyx (= Scelidodon) chiliensis, is
from Pampa de Fsiles, Peru. Originally the age of this
specimen was first reported as mid-Holocene (Hoffstetter 1982), but redating of the specimen yielded an
estimate of 16,000 2,000 rcybp (Pujos 2000).
Caribbean
Direct dating has been done only recently on any of the
Caribbean sloths. From this work it appears that they
survived much later than their mainland counterparts,
even into the Holocene. This is supported by an AMS
14C date in Cuba of 6,250 50 rcybp for a humerus of
the largest West Indian sloth, Megalocnus rodens, from
Cueva Beruvides in Matanzas Province (MacPhee et al.
1999). An even younger date of 4,960 280 rcybp is
available for a smaller species, Parocnus brownii, from a
tar pit known as Las Breas de San Felipe, also in Matanzas Province (Jull et al. 2004). In Hispaniola, there are
several Holocene whole bone 14C dates based on bones
from Haitian cave sites (Morgan and Woods 1986). Unfortunately, these dates were not tied to individual taxa.
MacPhee et al. (2000) provided the first evidence that
sloths persisted into the Holocene on Haiti based on a
single whole-bone date of 8,120 216 rcybp from an
assortment of sloth bones. Additional dates associated
with specific taxa have since been reported by Steadman et al. (2005) from multiple localities in Haiti. These
include Neocnus comes (7 dates ranging from 4,391
42 to 27,720 560 rcybp), Neocnus dousman (single
date of 9,897 65 rcybp), and Parocnus serus (>14,200
rcybp). In both Cuba and Haiti none of the dated sloth
bones are associated with cultural features or artifacts.
Only a single sloth is known from Puerto Rico, Acratocnus odontrigonus, and none of its bones have been
successfully 14C dated, as apparently all currently available samples have suffered a total diagenetic loss of collagen.
Based on the available radiocarbon dates for late
Pleistocene sloths, it is clear that the extinction of sloths
in the Caribbean postdated their disappearance in both
South and North America. As with the rest of the Pleistocene megafauna, the explanation for the extinction
of sloths has centered on whether it was caused by climatic and related environmental change or by humans.
Their asynchronous extinction on the mainland versus
the Caribbean has been used to support the argument
that it was human induced (Steadman et al. 2005). The
human impact need not have been the direct result of
activities such as hunting, but could have occurred indirectly through modification of the environment. If
extinction was human caused, perhaps an indirect effect is more likely, because evidence for human-sloth
association is limited, whether in South America, North
America, or the Caribbean. In North America, only 2 of
62 early Paleoindian sites with faunal remains contain
sloth material in a cultural context: Aubrey, Texas, and
Kimmswick, Missouri (Cannon and Meltzer 2004). In
both cases, the sloth remains are restricted to the dermal ossicles of Paramylodon harlani in large localized
concentrations that seem to indicate a piece of skin;
there are no bones with butcher marks.
Evidence for direct association between humans
and sloths in South America is limited to one species,
Mylodon darwini, (Jackson 2003; Lpez and Jackson
2004). One of the better-known sites where bones of
M. darwini have been recovered in a cultural context
is Las Buitreras Cave (Scillato-Yan 1976a; Caviglia et
al. 1986). A site supporting the idea that at least Mylodon was hunted by Paleoindians has been found in
northern Chile at El Quereo (Dillehay 2000). However,
instead of active hunting, some have proposed that the
carcasses of Mylodon were only scavenged (Borrero et
al. 1988).
For those sloths for which we have direct evidence
of diet based on dung, most of the plants recovered still
live in the vicinity of the site where the dung was recovered. This has been used as evidence that climatic
change, at least with regard to impacting food supply,
could not have been a primary factor in sloth extinction. However, despite fairly robust data from numerous sites for some taxa (e.g., Nothrotheriops and, to a
lesser extent, Mylodon), we simply lack similar dietary
data for most of the other late Pleistocene taxa, as well
as good radiocarbon dates. Climatic and environmen-
tal changes affect more than the food supply, and for
animals with a low basal metabolism, such as sloths,
seasonal temperature extremes may be an equally critical limiting factor. Sloths have played a major role in the
debate on Pleistocene extinctions in the New World and
will probably continue to do so. However, our knowledge of the paleoecology, biogeography, and timing of
disappearance for all sloth taxa must improve before we
are able to determine whether environmental change or
humans were the primary cause of their demise.
Summary
One of the most important trends in recent studies of
fossil sloths has been the shift from pure taxonomic
55
description to an increased emphasis on their paleobiology. However, detailed taxonomic studies are still
needed, particularly synthetic studies that will revise
the alpha-level taxonomy and work out the broader relationships of the different taxa at all levels. Coupled
with morphofunctional studies, stable isotope analysis,
recovery of preserved DNA, and our improved understanding of sloth biogeography and paleoecology, taxonomic studies will be critical in providing the context
in which to trace the evolution of the various lineages,
their diversification and adaptations. We hope this paper will serve as more than strictly an overview of our
current knowledge of fossil sloths, but, more importantly, will point the way to future research needs in
this fascinating group of mammals.
5
Paleogeographic distribution and
anatomical adaptations in Peruvian megatheriine
ground sloths (Xenarthra: Megatherioidea)
Franois Pujos
Em regies localizadas ao Oeste da Amrica do Sul as

espcies de Megatherium, ao que parece, percorreram
rotas andinas migrando em direo norte para o Equador. Essas espcies compartilham caracteres tais como
tamanho mdio, cndilos occipitais destacados e orientados para trs, rosto curto, crista deltopeitoral do
mero reduzida, fmur pouco retorcido, hipsodontia
moderada, e locomoo quadrpede o que as diferenciam de M. americanum e E. laurillardi que eram mais
bem adaptados s plancies baixas. Um endemismo
andino, comparvel ao insular, poderia explicar a alta
diversidade taxonmica no gnero Megatherium assim como a locomoo mais quadrpede. A subfamlia
Megatheriinae a nica dos Phyllophaga que apresenta
diferenas morfolgicas entre as formas andinas e as de
plancie de baixa altitude.
Resumen
Se analiza la distribucin paleogeogrfica de los Megatheriinae en el Per y se compara con la de otros pases
sudamericanos. El gnero tropical Eremotherium est
representado por E. laurillardi nicamente en la Costa
norte peruana y posiblemente en la Amazonia pero
nunca en los Andes. Megatherium est ejemplificado
por cuatro especies halladas en los Andes y al nivel de
los ros costeros. Estas especies comparten un tamao
mediano, destacados cndilos occipitales dirigidos hacia atrs, rostro corto, cresta deltopectoral del hmero
reducida, fmur poco torcido, hipsodoncia moderada
y una locomocin cuadrpeda que les diferencia de M.
americanum y E. laurillardi, mejor adaptados a las pampas de poca altitud. Un endemismo andino comparable
al endemismo insular podra explicar la alta diversidad taxonmica en el gnero Megatherium, as como
una locomocin ms cuadrpeda. En la parte oeste de
Amrica del Sur, las especies de Megatherium parecen
haber usado una ruta andina para migrar en direccin
al norte y llegar a Ecuador. La subfamilia Megatheriinae
es la nica dentro de los Phyllophaga que presenta diferencias morfolgicas entre formas andinas y de bajas
altitudes.
Introduction
Because of continental drift, South America was isolated for much of the Cenozoic. In this continent isle,
described as a perfect laboratory of evolution comparable to Australia and Madagascar by Simpson (1980),
a peculiar mammalian fauna evolved and diversified.
Some new clades, such as Notongulata, Litopterna, and
Xenarthra, appeared and colonized a major part of the
continent, with xenarthrans eventually migrating into
North America.
Xenarthra constitutes one of the most unique vertebrate clades and probably the most characteristic of the
South American assemblage. Nowadays, the group is
represented by two genera of tree sloths, eight (possibly
nine, see Gardner 2005) genera of armadillos, and three
genera of anteaters (Delsuc and Douzery this volume;
Gaudin and McDonald this volume). Modern sloths are
Resumo
feita uma anlise da distribuio paleogeogrfica dos
Megatheriinae no Peru assim como uma comparao
com a constatada em outros paises sul-americanos.
O gnero tropical Eremotherium na costa do norte
peruano, e possivelmente na Amaznia, tem como
nico representante a espcie E. laurillardi que nunca
foi registrada nos Andes e em bacias de rios costeiros.
56
Peruvian megatheriine ground sloths
slow, medium-sized mammals that inhabit the Amazonian forest. They are almost exclusively folivorous and
their hanging arboreal locomotory mode is unique in
vertebrate evolution. In contrast, fossil sloths had distinct geographical distributions, diversity, and feeding
and locomotory modes (McDonald and De Iuliis this
volume).
The history of the Phyllophaga (= Tardigrada or
Folivora) probably began near the Eocene/Oligocene
boundary. The oldest fossil records of undoubted sloths
come from the Deseadan SALMA (upper Oligocene)
of Argentina and Bolivia (Hoffstetter 1954, 1956; Engelmann 1987; Carlini and Scillato-Yan 2004; Pujos
and De Iuliis 2007). All traditionally recognized sloth
families (see below) are found in the Miocene, but it is
during the Pliocene and Pleistocene that the number of
taxa dramatically increases (but this may be due to bias,
see McDonald and De Iuliis this volume). As a group, it
seems that fossil sloths were not subject to geographic
or ecological limitations, given that their remains have
been discovered from Alaska to Patagonia and from
Amazonia to the Andean peaks. Four clades, traditionally recognized as families, are well known: Mylodontidae, Nothrotheriidae, Megalonychidae, and Megatheriidae (Gaudin 2004a). According to Gaudin (2004a), the
clade Megatheriidae includes the largest forms, such as
Megatherium and Eremotherium, and the Santacrucian
SALMA form Planops Ameghino 1887. However, Pujos
et al. (2007) argue that Planops may not be a megatheriid. Pending a comprehensive phylogenetic analysis
of sloths, Planops and the largest Megatheriidae, for the
purposes of this review I will consider that the clade
Megatheriidae includes Planops and Megatheriinae
(i.e., Megatherium, Eremotherium, Megathericulus
Ameghino 1904, Megatheriops Ameghino 1921, Plesiomegatherium Roth 1911, and Pyramiodontherium Rovereto 1914).
Megatheriines appear in South America during the
Miocene. Megathericulus from the lower Miocene of
Argentina is the oldest form. During the Pleistocene
megatheriines are widely recorded in South America
and colonized North America (De Iuliis and Cartelle
1999) through the Panamanian isthmus during the
Great American Biotic Interchange. The early history of
Megatheriinae is not well known. The earliest remains,
from the Mio-Pliocene, are sparse, fragmentary, and located principally in the south of the continent (i.e., Argentina, De Iuliis 1996; Pujos 2006). Only in the Pleistocene do Megatheriinae remains become abundant.
It is among these megatheriines that the largest sloths,
57
which reached the size of a modern African elephant,

evolved. Weighing between three and six tons (Faria
et al. 1998), they were among the biggest mammals in
South America during the Quaternary. Two genera of
giant ground sloths are recognized in the Pleistocene.
Megatherium is considered the temperate form and
Eremotherium the tropical one because of their relative geographical distributions: Megatherium typically
in the south of South America and Eremotherium in the
north (De Iuliis 1996; Pujos 2002).
Peru is ideally located for understanding the distributional limits between the two genera (figure 5.1).
Moreover, the abundance and diversity of megatheriine
ground sloths in Peru, especially Megatherium (Pujos
and Salas 2004a,b; Pujos 2006), allow better insight into
their different modes of locomotion and dietary habits,
and how these features might have been influenced by
climate and topographic relief.
Paleogeographic background, anatomy, and

phylogenetic position of the Megatheriinae
Cuvier (1796) described the skeleton of the megatheriine ground sloth Megatherium americanum from Pleistocene (Lujanian SALMA) deposits of Lujn, Buenos
Aires province, Argentina. For more than two centuries
following this description, remains of gigantic megatheriines continued to be discovered in the Americas,
but their phylogenetic history, migration patterns, and
peculiar morphological adaptations remained unclear.
The main part of megatheriine history was played out
in South America and several Mio-Pliocene genera
are known, principally from Argentina (e.g., Pyramiodontherium, Plesiomegatherium, and Megatheriops).
During the late Pliocene, Eremotherium crossed the
Panamanian isthmus (De Iuliis and Cartelle 1999) and
colonized North America.
Like most ground sloths, Megatheriinae exhibit
a strong torsion of the pes with weight borne mainly
on the lateral part. The manus does not exhibit torsion; Megatheriinae walked on the lateral side of the
fifth digit, more or less like modern anteaters such as
Myrmecophaga (Tito and De Iuliis 2003). There is a reduction of manual and pedal medial digits, carpal and
tarsal bones, and the formation of osteological complexes in the manus and pes. The pes presents a mesoentocuneiform complex (MEC, fusion of digits I-II and
some cuneiform elements; see Brandoni et al. 2004)
and the manus a metacarpal-carpal-complex (MCC,
fusion of medial carpal bones and sometimes digit I
58
F. Pujos
Figure 5.1. Paleogeographic distribution

of megatheriine
ground sloths in
Peru.
and/or II; see De Iuliis and Cartelle 1993). Tito and De

Iuliis (2003) noted that the MCC probably functioned
as a keystone element in transferring force through the
carpus. Megatheriinae also exhibit fusion between tibia
and fibula (proximally and sometimes distally as well),
a very robust, flat, and anteroposteriorly compressed
femur, a massive skull, a long snout, and strongly hypsodont molariform teeth bearing two transverse crests.
Megatherium and Eremotherium present distinct MEC
and MCC morphologies, as well as differences in femur torsion, general robustness of the bones, and body
size. They are commonly considered as large herbivores (Bargo 2001b), and possibly scavengers (Faria
and Blanco 1996). Bargos (2001b) detailed analysis of
the masticatory apparatus of Megatherium americanum
indicates it was probably a browser in open habitats,
feeding on soft, tough foods.
Locomotion in Megatherium is controversial. It is
generally illustrated as bipedal, using its tail as a prop

and eating leaves. Aramayo and Manera de Bianco
(1996) suggested Megatherium was bipedal and quadrupedal, but Blanco and Czerwonogora (2003) viewed
it as mainly bipedal. For the Pampean region, Casinos
(1996) considered Megatherium as principally quadrupedal, but adopting bipedal postures during feeding.
This hypothesis has been applied uncritically to all
megatheriines. The megatheriine trackways from Pehun-C (Buenos Aires province, Argentina, Lujanian
age = 12,000 years before present [ybp]) are inconclusive as they preserve footprints alone or foot and hand
prints.
Phylogenetic affinities between Mio-Pliocene genera
and Quaternary genera are uncertain, principally because of a lack of good data on the earlier forms (McDonald and De Iuliis this volume). Nonetheless, there
is substantial evidence that Megatherium and Eremotherium are monophyletic sister taxa (Pujos 2006). Three
species of Eremotherium are commonly recognized: E.
laurillardi, the largest, which colonized northern South
America, Central America, and southern North America; E. eomigrans from the Pleistocene of Florida, U.S.A.
(De Iuliis and Cartelle 1999), and E. sefvei from the
Pleistocene of Bolivia (represented by a single femur,
De Iuliis and St-Andr 1997). In contrast, Megatherium
includes numerous taxa, such as M. altiplanicum, M.
americanum, M. celendinense, M. elenense, M. medinae,
M. sundti, M. tarijense, and M. urbinai (see Pujos 2006
and references therein). A phylogenetic analysis (Pujos
2006) suggests the existence of two Megatherium lineages: M. (Megatherium) includes the two first species
listed above and M. (Pseudomegatherium) includes the
others. Let us now consider the possible relationships
among the phylogenetic history, paleogeographic distribution, and anatomy of the Megatheriinae.
Anatomic analogies among Andean

Megatheriinae
Like E. laurillardi, M. (M.) americanum is one of the
largest South American ground sloths (>4,000 kg; Casinos 1996; Faria et al. 1998). In contrast, Andean species (e.g., E. sefvei and those belonging to Megatherium)
are medium sized (1,0001,500 kg based on the methodology of Faria et al. 1998) and exhibit characters
outside the range of intraspecific variability observed in
M. (M.) americanum and E. laurillardi (Pujos 2006). In
particular, Andean Megatherium species are characterized by (1) a 2060% reduction in body size compared
59
to M. (M.) americanum, (2) occipital condyles well

extended posteriorly and prominent, (3) a short rostrum, (4) strong reduction of the humeral deltopectoral
crest, (5) a moderately twisted femur, and (6) moderate
hypsodonty relative to M. (M.) americanum (see figure 5.2). Reduced size and a relatively untwisted femur
characterize not only all members of the Andean Pseudomegatherium lineage except M. (P.) celendinense (a
particularly large Megatherium species from the northern Peruvian Andes, see Pujos 2006), but also the Montehermosan SALMA (Pliocene) Andean form M. (M.)
altiplanicum (St-Andr and De Iuliis 2001), which is the
sister species of M. (M.) americanum (Pujos 2006).
Limb morphology and characteristics of the occipital bones may indicate a more quadrupedal stance in
Andean taxa than in E. laurillardi and M. (M.) americanum (Pujos et al. 2002; figure 5.3). These latter two species mainly occupied regions of lower altitude with less
topographic relief (i.e., Pampean grasslands) in northern and southern areas of South America respectively.
Smaller size and a more quadrupedal stance may have
been responses related to a less uniform landscape,
compared particularly to the Pampean lowlands. In
addition, small size might have resulted from the lack
of abundant vegetation at high altitudes during glacial
periods.
It is tempting to link Megatherium lineages with quality of food and habitat. The M. (Megatherium) lineage is
noted for its very high hypsodonty (Bargo et al. 2006),
while Eremotherium teeth are the least hyposodont and
M. (Pseudomegatherium) are intermediate between
these two. The degree of hypsodonty may be related to
feeding behavior and habitat (Bargo et al. 2006). For example, M. (Megatherium) species occur at high altitude
(M. (M.) altiplanicum) and in low-lying pampas (M.
(M.) americanum), suggesting that increased vegetation hardness and open habitat, respectively, promoted
increased hypsodonty. M. (Pseudomegatherium) species are found in intermediate habitats, such as M. (P.)
medinae from lower altitudes, mainly near coastal areas
of Chile, and M. (P.) tarijense from intermediate altitude Tarija (Bolivia) localities. Eremotherium is known
mostly from the lowest areas, typically near-coastal
and mainly closed habitats (but see below). However,
such correlations are far from straightforward because
many exceptions occur. For example, M. (M.) americanum is also known from Tarija (De Iuliis 1996) and
the Yantac of Peru (Pujos and Salas 2004b), M. (P.) urbinai has been recovered from Tres Ventanas cave, M.
(P.) elenense from Cerro de Pasco is the highest altitude
Figure 5.2. Skull and dentary (A. [MLP 2-64], B. [UNI 1]), humerus (C. [MLP 2-201], E. [FMNH P14216]),
and femur (D. [MLP 44-XII-28-1], F [FMNH P14216]) of Megatherium (Megatherium) americanum
(A, C, D) and Megatherium (Pseudomegatherium) tarijense (B, E, F). dc = deltopectoral crest.
61
Figure 5.3. Hypothetical reconstruction of Megatherium (Pseudomegatherium) by R. Salas.
megatheriine known (Pujos and Salas 2004b), and E.

sefvei (though its skull is not known) and M. (P.) sundti
are both known from Ulloma, Bolivia. In addition, the
present level of paleoecological knowledge for many of
the localities yielding megatheriine remains is insufficient to allow confident statements about the relationships between environmental factors and the overall
diversity of megatheriines.
Diversity of the Megatheriinae and possible

migrations
Mio-Pliocene Megatheriinae are not very abundant.
Several genera represented by fragmentary specimens
are recorded in Colombia (Hirschfield 1985), Bolivia
(St-Andr and De Iuliis 2001), Venezuela (Linares
2004), and Argentina (Kraglievich 1940a). The large
Pleistocene Megatheriinae were important faunal elements until their extinction before the Montehermosan
(early Pliocene).
During the late Pliocene and Pleistocene, Megatherium occupied the temperate areas of South America
(Argentina, Uruguay, Paraguay, Bolivia, Chile, Peru,
and Ecuador) and Eremotherium the more tropical areas (Brazil, Ecuador, Venezuela, Colombia, and Peru;
Cartelle and De Iuliis 1995; Pujos and Salas 2004b). But
this geographic distribution, accepted by most xenarthran paleontologists (e.g., Hoffstetter 1982), excludes
the northwestern presence of megatheriines, particularly in the Andes, where, ironically, taxonomic diversity was highest. In fact, the high taxonomic diversity
of Megatherium may be directly related to Andean oc-
cupation, presumably because of aspects of the region

itself and habitat isolation.
Both Pleistocene genera are present in Peru (Marshall et al. 1984; Pujos and Salas 2004b; figure 5.1). Taking Eremotherium first, it is present in coastal Peru,
from Tumbes, Talara, and Piura in the north, to the
Cupisnique desert (Trujillo area) in the south, but absent from the Peruvian Andes. It may be represented
in Amazonia by finds from the Ucayali Basin (Buffetaut and Hoffstetter 1977) and by a large astragalus
from Tarapoto (Pujos and Salas 2004b). All Eremotherium specimens from Peru are referred to E. laurillardi (except the doubtful specimen from Tarapoto).
These specimens further highlight the high degree of
intraspecific variation observed within the Megatheriinae (Cartelle and De Iuliis 1995; De Iuliis 1996; Pujos
2002). Eremotherium is commonly associated in Peru
and Ecuador with faunas characteristic of open areas
(e.g., Equus [Amerhippus], Paleolama, Stegomastodon,
and Pampatherium). Along the Ecuadorian coast at
La Carolina it is also found in association with a small
Megatherium species, M. (P.) elenense, previously considered by most authors as an Eremotherium species
(Hoffstetter 1952; Marshall et al. 1984). Megatheriinae
are recorded in the Ecuadorian Andes, but without detailed studies (Marshall et al. 1984). The available data
suggest Eremotherium is absent from the Andes of Ecuador, Peru, and Chile; the only record is a single femur
of E. sefvei from the Bolivian Altiplano. Interestingly,
this species is clearly distinct from E. laurillardi (De
Iuliis and St-Andr 1997).
In contrast, Megatherium in Peru exhibits higher di-
62
F. Pujos
versity, which extends into other Andean areas (Ecuador, Bolivia, and Chile). At least eight species, seven in
the Lujanian (and maybe Ensenadan, including M. (M.)
americanum) and one in the Montehermosan, are recognized in the Andes. The large M. (M.) americanum,
so abundant in the Argentinian pampas, is recorded
close to the Andes (e.g., Neuqun, Mendoza, and Salta)
and the Bolivian Altiplano (Tarija). M. (P.) medinae is
in Chile (Casamiquela and Sepulveda 1974), M. (P.) elenense in Ecuador and Peru (Hoffstetter 1952; Pujos and
Salas 2004b), M. (P.) sundti in Bolivia (De Iuliis 2006),
M. (P.) tarijense in Bolivia and Peru (Pujos and Salas
2004b), and M. (P.) urbinai and M. (P.) celendinense
in Peru (Pujos and Salas 2004a; Pujos 2006). As stated
above, the Andean lineage M. (Pseudomegatherium)
is phylogenetically distinct from M. (M.) americanum
(Pujos et al. 2002; Pujos 2006). However, M. (M.) altiplanicum is also a small-sized Andean Megatheriinae,
but phylogenetically close to M. (M.) americanum (StAndr and De Iuliis 2001; Pujos 2006).
The geographical origin of Megatherium is unclear,
although three hypotheses are possible based on the
distribution of the fossil remains. One is that Megatherium originated in the southern part of the continent
and spread northward, with extensive taxonomic diversification occurring once the lineage reached the Andes. This scenario, which reflects the more traditional
view that much of the endemic South American fauna
had its origins in or around the south of the continent
(i.e., Argentina), is supported by the presence of a majority of pre- Pleistocene remains in southern South
America and by the rarity of Megatherium in northern
South America (but see the discussion by McDonald
and De Iuliis this volume on biases in the distribution
of fossil sloths). Under this scenario, the presence of M.
(M.) americanum and M. (P.) tarijense in the Lujanian
beds of Tarija may represent the trail of this northwestern spread, and the beginning of the radiation of
this genus.
A second possibility is that the genus originated in
the Andes. This view is supported by the presence of
the oldest certain member of the genus, M. (M.) altiplanicum, in the Andes (St-Andr and De Iuliis 2001).
The considerable subsequent diversity of the genus in
this area would then suggest the Andean region, with
its particular topography and varied ecological niches,
as a center of radiation. The remains in Tarija and, more
distantly, Argentina may then represent migration away
from the Andean center.
The third possibility is that the ultimate ancestor of

Megatherium was present in the northern part of South
America. Although this alternative is in sharp contrast to the traditional view of Argentina as the center
of megatheriine radiation, pre-Pliocene fossil remains
are known from Venezuela and Colombia. Given the
presence of the basal megatheriine genus Megathericulus from the middle Miocene of Colombia (Hirschfeld
1985), this hypothesis must remain a possibility.
The Andes have been commonly considered a route
of migration and dispersal used by many mammalian
groups (Alberdi et al. 2004). For example, the gomphothere Cuvieronus hyodon is first recognized in the
Pleistocene of the Peruvian Andes, while Stegomastodon waringi is recorded later from the coast (Alberdi
et al. 2004). Apparently, Cuvieronus and Stegomastodon used the Andean and coastal routes from north
to south. Megatherium species may have used the same
routes, but at a later time and in the opposite direction
(i.e., from south to north, assuming the clade arose in
southern South America), with occasional movements
down to coastal swamps (figure 5.1).
The presence of several taxa of Andean Megatheriinae on the Pacific coast is not necessarily paradoxical. Indeed, these large ground sloths could have used
the Andean valleys to move from the Andes to coastal
oases like the Paleo-Ro Ica, a Ocucaje Late Pleistocene
site where M. (P.) urbinai is recorded (Pujos and Salas
2004a; figure 5.1). Pujos (2002) suggested the existence
of local oases, marked by the development of lagoons
and grasslands favorable for mammals, in the coastal
Sacaco and Aguada de Lomas areas. Salas et al. (2004)
provided paleontological and geological evidence for
the existence of swampy regions within the desert along
the southern Peruvian coast ca. 400,000200,000 years
ago. Further, the presence of various articulated skeletons in Ocucaje, Sacaco, Aguada de Lomas, and the
Cupisnique desert preclude the possibility of a postmortem fluvial transport of these specimens from the
Andes, meaning the individuals lived near their final
resting places. These places may have been refuges during the harsher periods in glacial times. As in the present, abundant vegetation occurred near Pacific coastal
rivers during the Quaternary, which could explain the
presence of such large mammals. For example, in the
case of the Ocucaje site where M. (P.) urbinai is present,
it is accompanied by Stegomastodon waringi.
Does endemism affect all groups of fossil

Xenarthra?
Nowadays, mammal assemblages are distinct in coastal
versus Andean Peru (see Eisenberg and Redford 1999).
Differences also occurred during the Lujanian SALMA
(Marshall et al. 1984; Hoffstetter 1986). For example,
during the Quaternary, native ungulates were absent
from coastal sites of Ecuador (e.g., La Carolina), Peru
(Cupisnique, Ocucaje, Sacaco), and, apparently, Chile,
but abundant in the Andes (e.g., Tarija in Bolivia, Marshall et al. 1984). The gomphotheriid Stegomastodon
was abundant on the coast, but Cuvieronus was more
common in the Andes. Similarly, the equids Hippidion and Onohippidion are typically recorded from the
Andes, but Amerhippus is the only fossil horse of the
Peruvian coast. Among Xenarthra, Holmesina cf. paulacoutoi (Dasypodoidea: Pampatheriidae) is the only
fossil cingulate (armored xenarthran) present on the
Peruvian coast. It is generally replaced in the Andes by
the glyptodonts Glyptodon sp. (Glyptodontinae) and
Panochthus sp. (Hoplophorinae), which are extremely
common there (e.g., Cuzco in Peru and Cochabamba in
Bolivia, personal observations). Adaptations to mountainous relief and coastal plains are clearly observable in
South American horses (Eisenmann 1984; Eisenmann
and Sondaar 1998). This might lead one to predict that
other groups (e.g., xenarthrans) might differ between
the two regions as well.
Aside from the megatheriines described above,
other sloths in Peru are less diverse and consist of one
Scelidotheriinae (Scelidodon chiliensis) and one Mylodontinae (Glossotherium sp.) from the Ensenadan and
Lujanian SALMA, principally on the coast but also in
the Andes (Pujos and Salas 2004b). Further, intraspecific variation in Scelidodon chiliensis is extremely low
among the Peruvian specimens observed. Possible explanations for the distribution patterns of these sloth
clades may be that the Megatheriinae arrived earlier in
the Andes (with M. altiplanicum in the Montehermosan SALMA) or were more highly adaptable than other
fossil ground sloths. The existence of many Megatherium species in the Andes could reflect an Andean
endemism, comparable to the insular endemism of
megalonychoid sloths in the Greater Antilles. Andean
valleys could have served as islands for Megatherium
populations. Another possibility is that we have overestimated species diversity in the Andes. However, the
present knowledge of Megatheriinae (De Iuliis 1996,
2006; De Iuliis and St-Andr 1997; St-Andr and De
63
Iuliis 2001; Pujos 2002, 2006; Pujos and Salas 2004a,b)

does suggest a multitude of megatheriine species in the
Andes and westward. If so, then a task for the future
will be to test hypotheses, such as those presented here,
to explain the diversity of high altitude forms.
Summary and conclusions

Over the last few years, the discovery of various Megatheriinae specimens in Andean and Pacific coastal areas
of South America has yielded new information on locomotion and adaptations of Quaternary ground sloths.
One species of Eremotherium is recorded in Peru
and does not range below the latitude of 8S (figure 5.1).
This tropical genus is abundant in Brazil, Ecuador, Colombia, Venezuela, and the northern coast of Peru, but
practically absent in the Andes (the exception being the
poorly known E. sefvei from Bolivia; De Iuliis and StAndr 1997). In contrast, Megatherium (Pseudomegatherium) is abundant in the Andes of Chile, Peru, Bolivia,
and Ecuador. Like most mammals during the Quaternary, Megatherium used an Andean route to migrate,
most probably northward (figure 5.1). In some cases,
Andean Megatherium species are recorded from the
Peruvian and Ecuadorian coasts in close proximity to
large rivers. This may have occurred as sloths followed
Andean valleys to descend to the coast and reach local
oases that served as refuges. Remote Andean valleys
may have functioned as islands, resulting in local endemism. Andean megatheriines appear closely related,
as confirmed by their inclusion in the subgenus Megatherium (Pseudomegatherium) (Pujos 2006). They were
probably more quadrupedal than larger forms such as
M. americanum and E. laurillardi. Lastly, the degree of
diversity witnessed among megatheriines does not occur in other sloth groups. The phylogenetic history of
the Megatheriinae seems to be closely associated with
paleogeography and has apparently been strongly influenced by tectonic and climatic environmental modifications.
Acknowledgments
I thank S. F. Vizcano and J. Loughry for inviting me
to contribute to this special volume on Xenarthra, V.
Eisenmann, who commented on an earlier version of
this work, and R. Salas for the illustration of figure 5.3.
I acknowledge the two reviewers, G. De Iuliis and C.
Cartelle, and the editors for critical review of the manuscript.
6
Skeletal anatomy and the fossil history
of the Vermilingua
H. Gregory McDonald, Sergio F. Vizcano, and M. Susana Bargo
terrestres para arborcolas ou vice-versa. Nesta contribuio apresentamos uma sntese das caractersticas
mais salientes do esqueleto dos Vermilingua atuais e o
estado presente do conhecimento das fsseis, incluindo
uma relao sinptica dos especimes disponveis em
colees de museus.
Resumen
La diversidad taxonmica de los osos hormigueros
de Amrica del Sur o Vermilingua (Familias Myrmecophagidae y Cyclopedidae) est limitada a solo
tres gneros y cuatro especies. El registro certero ms
temprano de Vermilingua corresponde al Oligoceno
tardo-Mioceno temprano, aunque hay algunas propuestas dudosas ms antiguas. Los restos son escasos y
muchos de los taxones fsiles estn basados en huesos
aislados. Este registro tan parcial limita la capacidad de
entender algunas de las tendencias evolutivas dentro
del grupo, como sus especializaciones craneanas para
la alimentacin basada en insectos coloniales y la transicin de hbitos terrestres a arborcolas o viceversa.
En esta contribucin presentamos una sntesis de las
caractersticas ms salientes del esqueleto de los Vermilingua vivientes y el estado actual del conocimiento de
los fsiles, incluyendo una lista sinptica de los especimenes disponibles en colecciones de museos.
Introduction
An insectivorous diet is common in many extant mammals and is implied as the primitive condition in the
earliest eutherian mammals, as all proximal outgroups
to the Eutheria were apparently insectivorous (KielanJaworowska et al. 2004). However, specialization for
feeding on colonial insects such as ants and termites
(myrmecophagy) has evolved in only a limited number of mammals, although trace fossils considered to
be termite nests are known from the Triassic (Hasiotis
and Dubiel 1995) and body fossils of colonial insects are
known from the Cretaceous (Grimaldi and Agosti 2000;
Krishna and Grimaldi 2003). This dietary specialization is reflected in numerous morphological features,
resulting in convergent evolution among a number of
unrelated groups. Many of the skeletal characters associated with myrmecophagy first appear in the Jurassic in the enigmatic Fruitafossor (Luo and Wible 2005).
Modern forms displaying these adaptations (to various
degrees) include monotremes (the echidna, Murray
1984), marsupials (the numbat, Myrmecobius fasciatus,
Reiss 2001), the pholidotes, tubulidentata, xenarthrans,
and a carnivore (the aardwolf, Proteles cristatus). The
greatest number of similar morphological characters
is shared by the xenarthran anteaters (Vermilingua)
and pangolins (Pholidota) and these similarities have
been interpreted as indicative of a close phylogenetic
Resumo
A diversidade taxonmica dos tamandus na Amrica
do Sul (Vermilingua: Famlias Myrmecophagidae e Cyclopedidae) est limitada apenas a trs gneros e quatro
espcies. O registro confivel mais antigo dos Vermilingua corresponde-se com o Oligoceno superior-Mioceno
inferior, embora haja propostas duvidosas anteriores.
Os restos achados tm sido escassos, e muitos dos txons fsseis baseiam-se em ossos isolados. Este registro
to parcial limita a nossa capacidade de entendermos
algumas das tendncias evolutivas dentro do grupo,
como suas especializaes craneanas para a alimentao
baseada em insectos coloniais e a transio de hbitos
64
Skeletal anatomy and the fossil history of the Vermilingua
relationship (Englemann 1985; Reiss 1996; Asher et al.

2003) or that at least the two lineages should be considered sister groups (Novacek and Wyss 1986).
Contrary to these earlier systematic investigations,
recent morphological studies of the South American
anteaters (Vermilingua) and the Eurasian pangolins
(Pholidota) have shown that the two groups are not
closely related (Rose and Emry 1993; Gaudin and Wible
1999; Gaudin and McDonald this volume). Instead, the
anatomical similarity represents an outstanding example of convergent evolution reflecting common dietary
specialization on colonial insects. This more distant
relationship between the Vermilingua and Pholidota
has been supported by recent molecular studies (Delsuc et al. 2001, 2002, 2003; Springer et al. 2005; Delsuc and Douzery this volume), which place Xenarthra
(including the Vermilingua) as one of the four major
subdivisions of placental mammals and the Pholidota
within the Laurasiatheria. The term anteater in this paper will refer to dietary specialization/habits for feeding
on colonial insects such as termites and ants, whereas
the term Vermilingua is used taxonomically to refer to
the South American xenarthrans with this dietary specialization and includes all members, fossil and living,
of the two currently recognized extant families, Myrmecophagidae and Cyclopedidae, and possibly one or
two poorly known taxa from the Miocene (Santacrucian). The highly specialized feeding apparatus of the
Vermilingua has been studied in detail (Naples 1985,
1999; Reiss 1997) and the reader is referred to these articles for details concerning modifications to the cranium and its musculature.
F. Ameghino (1894) described the family Adiastaltidae to include his genera Adiastaltus 1893 and Plagiocoelus 1894. Originally Ameghino (1893) considered
Adiastaltus and Anathitus to be closely related to monotremes, with the latter possibly a link between theriodonts and monotremes. In 1894 he included Adiastaltus and Plagiocoelus (Fam. Adiastaltidae), Anathitus
(Fam. Anathitidae), Scotaeops (Fam. Scotaeopsidae; an
armadillo now known as Stegotherium) and Dideilotherium (Fam. Dideilotheridae) within the Monotremata.
An evaluation of the status of these families is being
undertaken, but for the purposes of this review, based
on the morphology of the distal humerus, we provisionally consider the families Adiastaltidae and Anathitidae as junior synonyms of the Myrmecophagidae
and all members to be poorly known vermilinguans. A
complete analysis is not possible at this time given the
paucity of material available for these taxa, which lim-
65
its a comprehensive morphological assessment of their

relationships to either the Vermilingua or any other
group of South American mammals.
The relationship of Eurotamandua from Germany,
originally considered a member of the Vermilingua
(Storch 1981) has been thoroughly discussed elsewhere
(Gaudin and Branham 1998; Szalay and Schrenk 1998;
Delsuc et al. 2001; Storch 2003; Rose et al. 2005) and
will not be considered here, except to note that we do
not consider it a member of the Xenarthra, let alone
a Vermilinguan, but merely another outstanding example of the convergent evolution that has occurred in
mammals adapted for anteating.
Morphology of the Vermilingua

Members of the Vermilingua possess numerous distinctive modifications of the skeleton and soft anatomy
indicative of their adaptive specializations for feeding
on colonial insects. While a comprehensive description
is not possible, a brief overview of the major modifications of the skeleton of the living forms is provided
to illustrate features useful for classifying fragmentary
fossil material.
Cranial morphology
Gaudin and Branham (1998, see also appendix 3.1 of
Gaudin and McDonald this volume) list several cranial
synapomorphies of Vermilingua, including curvature
of the basicranial/basifacial axis, reduced maxillary
orbital exposure, and the posteromedial orientation of
the eustachian tube. Other specializations in the skull
are related to the use of the slender, elongated, sticky
tongue, the feature from which the group derives its
name (Vermilingua = worm tongue), as the primary
means of capturing prey. These include an elongation
of the rostrum and the hard palate and a complete
loss of dentition (figure 6.1A,B,C). The zygomatic arch
is incomplete and the jugal is reduced to a small process attached to the zygomatic process of the maxilla
in Tamandua and Myrmecophaga, while in Cyclopes it
has been lost. A septomaxilla has been documented in
Tamandua (Zeller et al. 1993) and Cyclopes (Gaudin
and Branham 1998). The premaxilla remains small
and lengthening of the skull is accomplished primarily
through an elongation of the nasal and maxilla.
In Tamandua and Myrmecophaga the hard palate is
elongated as a result of a slight elongation of the palatines and a joining of the pterygoids at the midline, so
that the internal nares are shifted posterior to the gle-
66
H. G. McDonald, S. F. Vizcano, and M. S. Bargo
noid fossa. The elongated secondary palate allows these

genera to accommodate the retracted tongue within
the oropharynx without interfering with the ability to
breathe (Naples 1999). In this regard these two genera
are more derived than Cyclopes, where, like the others,
the pterygoids and soft palate are elongated posteriorly,
but the internal nares remain in the same plane as the
orbit (Patterson et al. 1992; Reiss 1997). In all Vermilingua the pterygoids extend posteriorly to form the
anterointernal portion of the tympanic cavity, whereas
the basioccipital forms its posterointernal wall. The
contribution of these two bones to the formation of the
tympanic cavity is a distinctive feature of the Vermilingua that distinguishes them from other Xenarthra
(Patterson et al. 1992). The anterior edge of the tympanic is well behind the glenoid in Cyclopes but in Tamandua and Myrmecophaga the anterior margin of the
tympanic forms its posterior edge. There is no mastoid
process. The stapes in Myrmecophaga is bicrurate while
in Cyclopes it is columelliform and lacks a perforation,
which presumably has been lost secondarily (Patterson
et al. 1992; Gaudin et al. 1996).
Whereas Cyclopes retains prominent coronoid and
angular processes on the mandible, in the other two extant genera they are reduced. The mandibular symphysis is unfused in all three genera. This permits mandibular rotation during feeding and effectively increases
the gap of the mouth opening (Naples 1999).
Postcranial morphology
The primary postcranial modifications of the skeleton
are in the forelimbs, which are modified for digging.
These modifications cannot be viewed as fossorial in a
strict sense because the forelimbs are used only for digging earth in Myrmecophaga, whereas in Cyclopes and
Tamandua they are used to remove bark from trees or
rip open the hard nests of ants or termites in order to
feed. Detailed descriptions of various postcranial features follow in the next sections.
Scapula
The scapula has a coracoid (or scapular) foramen, a
character shared with sloths and used to define the
Pilosa (Flower 1882), although the coracoid is greatly
reduced in Cyclopes and the foramen has been secondarily lost. The foramen is formed between the procoracoid (= anterior coracoid), which is a separate center
of ossification, and the supraspinous fossa. Characteristic of all Xenarthra is a secondary scapula spine, with
the attachment area for the teres major muscle greatly
enlarged in Myrmecophaga and Tamandua, but not

as much in Cyclopes. The teres major serves to retract
the humerus and the enlargement of its origin on the
scapula and its more distal insertion on the humerus
increases its moment arm, thus increasing the amount
of force it contributes during digging (Taylor 1978). The
acromion process is well developed in all forms and the
metacromion process is present in Myrmecophaga and
Tamandua but absent in Cyclopes. The acromion process extends more medially than in other mammals and
increases the capability of the acromiodeltoid muscle to
protract the humerus because it positions the muscle in
front of the shoulder and not to the side. This may aid
in climbing or defensive posturing (Taylor 1978).
Humerus
The humerus is robust in the Vermilingua, with welldeveloped muscle scars and ridges for the origin of
muscles that both extend and flex the forearm. The deltoid tubercle is well developed and curves posteriorly
in all three extant genera. An entepicondylar foramen is
present in all forms. The distal end is greatly expanded,
primarily by the enlargement of the supinator crest,
which provides the origin of the extensor carpi radialis,
extensor carpi ulnaris, and extensor digitorum communis. One head of the supinator, along with the extension
of the medial epicondylar process, provides the origin
of the pronator teres, flexor carpii radialis, flexor digitorum sublimis, flexor carpi ulnaris, and flexor digitorum profundus muscles on the anterior surface and the
epitrochleo-anconeus muscle on the posterior surface
(Taylor 1978). In Tamandua the medial head of the triceps is fused with the flexor digitorum profundus by a
common tendon that passes under the distal border of
the medial epicondyle to insert on the third digit. The
enlargement of the entepicondylar process and notch
can be used to infer this unique muscle arrangement in
fossil Vermilingua, but both characters are also present
in a variety of other forms (e.g., pangolins and palaeanodonts) that lack this specialized muscle arrangement.
The supinator crest may be further extended proximally
by a delto-epicondylar ligament that connects the crest
with the deltoid tubercle (Taylor 1978).
Ulna
The olecranon process is enlarged compared to nonfossorial mammals, but not to the extent seen in armadillos (Vizcano et al. 1999). The index of fossorial
ability (IFA), calculated as the length of the olecranon
process divided by the difference between ulnar length
Figure 6.1. Skulls of extant (AC) and fossil (DF) Vermilingua. A. Myrmecophaga tridactyla (MACN 49.405).
B. Tamandua tetradactyla (MACN 33.255). C. Cyclopes didactylus (no catalog number). D. Protamandua rothi
(YPM-PU 15267). Note that because this specimen is extremely flattened dorsoventrally, we decided to
include a ventral view, which is more informative in terms of showing various vermilinguan features.
E. Neotamandua conspicua (MACN 8097). F. Palaeomyrmidon incomptus (MACN 8098). Scale bar 5 cm. See
Appendix 6.1 for full names of museum acronyms.
68
and olecranon length x 100, is 24.2 in Myrmecophaga,

Tamandua, and Cyclopes. This is considerably less than
the value of 101.17 34.5 in the most fossorial armadillo, Priodontes, and not even half that obtained for the
least fossorial and most cursorial armadillo, Tolypeutes
(58.02 9.16, Vizcano and Milne 2002).
White (1993) noted that climbing and digging mammals often have a rather long, proximally extended
olecranon, with little dorsal extension and often a
slight ventral orientation, whereas in more terrestrial
mammals the olecranon tends to project dorsally. She
developed an index based on the length and depth of
the trochlear notch to clearly distinguish the three extant vermilinguan genera with relation to their locomotor habits. Cyclopes had the lowest index (ca. 1.1),
Myrmecophaga the highest (ca. 1.7), and Tamandua was
intermediate (ca. 1.4).
In all extant forms, the anconeal process is only
moderately developed, permitting a greater extension
of the forearm, and there is no coronoid process. Both
Myrmecophaga and Tamandua have a well-developed
styloid process that is absent in Cyclopes.
Radius
The radius in the Vermilingua has a distinct curvature
so that the distal end is offset anteriorly relative to the
proximal end, and the distal anterior edge is flared. This
increases the moment arm of axial rotation by the pronator teres and hence the force generated, even though
the range of pronation and supination is more limited
than in other mammals (about 50 in Tamandua and
40 in Myrmecophaga, Taylor 1978). This modification
is less pronounced in Cyclopes. Orr (2005) noted that in
Myrmecophaga there is a distal projection of the dorsal
ridge of the distal end and a deeper concavity of the
distal articular surface that limits the range of extension of the antebrachiocarpal joint to 1015, compared
to 60 in Tamandua. This ridge acts as a brace, and Orr
considered this to be a hallmark of a locomotor style
that involves the manus being held vertical relative to
the substrate. The presence or absence of this feature in
fossil Vermilingua can be used to infer terrestrial versus
arboreal habits in extinct forms.
Manus
While metacarpals two and three are roughly the same
length and slightly shorter than metacarpals four and
five, the manus is dominated by the third metacarpal,
which is roughly rectangular in morphology and more
robust than the other metacarpals (Taylor 1978). The
distal end is strongly keeled, with the keel angled to the

axis of the bone so that it slopes medially. Its distinctive morphology in vermilinguans has been discussed
by Hirschfeld (1976), Shaw and McDonald (1987), and
Gaudin and Branham (1998). The fifth metacarpal
is reduced, lacks an ungual (present on all other digits), and is bound in a pad of fat and connective tissue with the fourth metacarpal. The proximal phalanx
is proximodistally shortened and there is no flexion
or extension between it and either the metacarpal or
second phalanx. The animals weight is carried on the
back of the unguals, primarily the third (the largest)
and fourth, which can be flexed at ninety degrees to
the manus in Myrmecophaga (Orr 2005). In contrast,
Tamandua walks on the sides of its hands, supported
by the padded fourth and fifth digits.
Femur
The head of the femur in all three extant genera extends
above the greater trochanter. The angle at which the
head of the femur is oriented relative to the shaft varies, though, with the highly arboreal Cyclopes having
the smallest angle (extending more proximally); in the
fully terrestrial Myrmecophaga it extends more medially, and the angle is intermediate in Tamandua (White
1993). The lesser trochanter is prominent in all three
taxa. There is no distinct third trochanter in Cyclopes,
but it is represented by a thin lip of bone along the distal
half of the lateral edge of the shaft that is continuous
with the lateral epicondyle. The third trochanter in Tamandua is a slightly raised ridge on the midshaft of the
femur. In Myrmecophaga, the third trochanter is represented as a thickening of a prominent crest that extends
from the greater trochanter to the lateral epicondyle.
The distal end of the femur in Cyclopes is flattened more
than in the other two genera, reflecting the greater flattening of the diaphysis. There is a prominent epicondylar process above the medial condyle in Tamandua and
Myrmecophaga that is absent in Cyclopes. The patellar
surface is continuous with the articular surfaces of the
condyles in all three genera.
Because the shape of the distal femur is highly correlated with the mode of locomotion in many mammals,
White (1993) utilized an index of distal femur shape
(maximum depth of the distal end divided by its width)
that clearly demonstrated differences between the
modes of locomotion in the three extant genera. Cyclopes had the lowest value (ca. 0.65) and Myrmecophaga
the highest (ca. 1.05). This index is partially determined
by the degree to which the medial side of the patellar
surface is raised relative to the lateral edge. There is no

raised edge in Cyclopes but there is a prominent one
in Myrmecophaga. This index, along with the femoral
neck angle, can be applied to fossil forms as a means of
determining the extent to which they were arboreal or
terrestrial.
Tibia
The shaft of the tibia is flattened in Cyclopes, and the
distal articular surface is also anteroposteriorly flattened. The distal tibial articular surface is shallow and
the grooves for the tendons of the tibialis caudalis,
flexor digitorum longus, and flexor hallucis longus are
greatly reduced in Cyclopes but are large and prominent
in Myrmecophaga and Tamandua. In Tamandua the diaphysis of the tibia is more oval, but it tends to be more
triangular in Myrmecophaga. The articular surface for
the astragalus is deeper in Myrmecophaga than in the
other two genera, thus restricting lateral movement.
Pes
The pes is plantigrade in Myrmecophaga and Tamandua but highly modified for grasping branches in Cyclopes, as the derivation of the generic name suggests. All
Vermilingua have an astragalus with a concavity on the
navicular facet and a navicular possessing a complementary mamillary process. There is a tibial sesamoid
next to the entocuneiform.
Vertebral column
In Myrmecophaga there are four sacral vertebrae, with
the first three articulating with the ilium and the fourth
articulating with the ischium, as does the first caudal.
In Tamandua the number of sacral vertebrae is three,
with the first two articulating with the ilium and the
third having no contact with the pelvis, while the first
caudal articulates with the ischium to form the synsacrum. The number of sacral vertebrae in Cyclopes is
four, with the first three articulating with the ilium, the
fourth having no contact with the pelvis, and the first
caudal vertebra articulating with the ischium.
The caudal vertebrae differ between Myrmecophaga, Tamandua, and Cyclopes, reflecting the presence
or absence of a prehensile tail. In Myrmecophaga the
anterior caudal centra are short, robust, and round,
with a massive neural arch and neural spine, and the
zygapophyses are deep limiting lateral movement. The
anterior caudal series in Tamandua have centra with
flat ends, short neural arches, and zygapophyses that
are broad and shallow. Tamandua and Cyclopes have bi-
69
furcate transverse processes on the mid and posterior

caudals, indicative of their prehensile tails. The bifurcation of the transverse process is absent in the caudals
of Myrmecophaga. The transverse processes of the caudals in Tamandua are longer anteroposteriorly than in
Myrmecophaga for attachment of the muscles that flex
the tail. The posterior caudal vertebrae in Tamandua
are phalanxlike, with convex posterior ends that allow
for flexibility in the end of the tail (Hirschfeld 1976).
Gaudin and Branham (1998) note that bifurcated, elongated transverse processes of the caudal vertebrae are
found in Protamandua (FMNH 13134), implying that
this taxon, like Cyclopes and Tamandua, had a prehensile tail.
Ribs
All three extant genera of Vermilingua have anteroposteriorly expanded ribs. A similar pattern of rib expansion is present in some genera of armadillos. In Cyclopes
the expansion of the ribs is more pronounced than in
the other two genera so that successive ribs imbricate.
The expansion of the ribs may increase the stability of
the thorax and vertebral column and can facilitate both
arboreal and fossorial habits (Jenkins 1970).
Fossil record of the Vermilingua

Given the relative rarity of fossil Vermilingua, we have
provided a synoptic list (appendix 6.1) of all specimens either described in the literature or known to us
through our visits to various collections. That it is even
possible to provide such a list indicates just how few
specimens have been collected.
Except for an enigmatic metacarpal from the early
Eocene of Patagonia that shows some similarities with
Vermilingua (Carlini et al. 1989), the earliest certain
fossil of the group is from the Colhuehuapian (late
Oligocene to early Miocene) from the Atlantic coast of
Chubut province, Argentina (Carlini et al. 1992). The
specimen consists of half a mandible, numerous vertebrae, ribs, scapula, a fragment of the distal humerus,
an almost complete manus, both femora, the proximal
portion of the tibia, and an almost complete pes. The
specimen has not been formerly described but is similar in size to Tamandua, although more robust. What
makes the discovery of this specimen even more remarkable is that it was recovered from the marine sediments of the Gaiman Formation.
A large number of taxa from the Santacrucian
(early-middle Miocene) have been referred to the Ver-
70
milingua, including Protamandua rothi, Promyrmephagus dolichoarthrus, Promyrmephagus euryarthrus,

Adiastaltus habilis, Adiastaltus procerus, Plagiocoelus
obliquus, Anathitus revelator, Argyromanis patagonica,
and Orthoarthrus mixtus. The large number reflects the
generally fragmentary nature of the types, and the fact
that each new discovery was described as a new taxon.
Because different skeletal elements often were utilized
as a type, it has not been possible to make any comparisons between taxa that would permit their synonomy.
Some taxa, among them Adiastaltus habilis, Adiastaltus
procerus, Anathitus revelator, Argyromanis patagonica,
and Orthoarthrus mixtus, were considered Xenarthra
incertae sedis by Mones (1986), whereas Plagiocoelus
obliquus was considered Mammalia incertae sedis by
the same author, reflecting the need for a critical reexamination of this material. A preliminary step in this
direction was undertaken by Hirschfeld (1976), who
noted that the type specimen of Protamandua rothi (a
calcaneum) easily articulated with the type of Promyrmephagus euryarthrus (an astragalus), indicating that
not only were the taxa the same but the two specimens
probably represented the same individual. She proposed that P. euryarthrus and P. dolichoarthrus be considered junior synonyms of Protamandua rothi. Carlini
et al. (1993) considered Promyrmephagus euryarthrus
to be congeneric with Protamandua but retained it as a
distinct species.
There has been no recent study of the family Adiastaltidae and its genera, Adiastaltus and Plagiocoelus,
or the Anathitidae with Anathitus. As indicated earlier,
for the practical purposes of this review, we consider
them to be synonymous with the Myrmecophagidae,
as was preliminarily proposed by Vizcano, Bargo, et al.
(2004), based on the similarity of the distal humerus
to other Vermilingua. However, these genera require
re-examination to determine their validity. Likewise,
a re-examination of Argyromanis patagonica and Orthoarthrus mixtus is needed to determine their validity,
possible synonomies, and systematic relationships.
Pending further study of the Santacrucian material, only one taxon can be confidently referred to the
Vermilingua, Protamandua rothi (figure 6.1D). This is a
small animal, intermediate in size between Tamandua
and Cyclopes. The morphology of the manus and pes
are similar to those of Tamandua and Myrmecophaga,
but lack any of the specializations in the limbs for either
arboreal or terrestrial locomotion seen in these genera.
The only specialization seen in Protamandua is modifi-
cation of the caudals, indicating that it had a prehensile

tail (Hirschfeld 1976; Gaudin and Branham 1998).
Neotamandua? australis was described by ScillatoYan and Carlini (1998) from the Colloncuran (their
Friasian, Colln Cur Formation) of Ro Negro province, Argentina. The type specimen is a complete right
humerus. A distal left humerus was also found. Assignment of the new species to Neotamandua was tentative
given the limited amount of material. The humerus is
similar in size to Tamandua but about 50% wider for its
length. Scillato-Yan and Carlini (1998) considered its
overall structural appearance to be most similar to Tamandua and N. conspicua. They considered Neotamandua to be ancestral to Tamandua and Myrmecophaga
and postulated without discussion that the two forms
evolved allopatrically, with Myrmecophaga evolving in
the northern part of South America and Tamandua in
the south.
In the late Miocene (Laventan), Neotamandua is
represented by N. borealis, based on a partial skeleton
from La Venta, Colombia. This taxon is intermediate
in size between Tamandua and Myrmecophaga and is
also intermediate in most characters of the manus and
pes (Hirschfeld 1976). Preserved caudal vertebrae indicate that the tail was not prehensile as in extant Tamandua. Hirschfeld (1976) considered N. borealis a possible
common ancestor to Tamandua and Myrmecophaga.
Both Engelmann (1985) and Gaudin and Branham
(1998) considered Neotamandua to be the sister taxon
to Myrmecophaga, that is, to be more closely related to
Myrmecophaga than to Tamandua, rather than the ancestor to both modern genera. Patterson et al. (1992)
suggested that Neotamandua (N. conspicua, FMNH
P14419) was congeneric with Myrmecophaga, although
Gaudin and Branham (1998) treated it as a separate
taxon and recognized at least one autapomorphy not
shared with the modern genus. This taxon occurs during the interval identified by Delsuc et al. (2001) as the
time of divergence of Tamandua from Myrmecophaga.
Neotamandua conspicua was described by Rovereto
(1914) based on a posterior part of a skull from the
Huayquerian (late Miocene) of Catamarca province,
Argentina (figure 6.1E). Based on some skull measurements, this species is slightly larger than Tamandua.
The type specimen of Neotamandua greslebini
from the Huayquerian (late Miocene) of Topunco
in Tucuman province, Argentina, consists of a partial skeleton that includes the proximal third of both
femora, the distal end of the left femur, the body of the
right astragalus, the left calcaneum, distal left humerus,

proximal right ulna, and distal right radius (Kraglievich
1940b). Unfortunately, the specimen cannot be located
at the moment (but see notes in synoptic list of specimens).
Also present in the Huayquerian is one of the betterknown fossil Vermilingua, Palaeomyrmidon incomptus
(figure 6.1F), represented by a nearly complete skull
(without a mandible), and a few postcranial fragments
that are tentatively assigned to the genus. This small
taxon is generally considered closely related to the extant Cyclopes by most workers (Hirschfeld 1976; Gaudin
and Branham 1998). Delsuc et al. (2001) calculated that
the pygmy anteater diverged from other anteaters 35
45 mya; Barros et al. (2003) place the divergence at 32
mya. Either way, these results indicate a long hiatus in
the fossil record between the proposed origin of the lineage and the first appearance of a member in the fossil
record between 7 and 9 mya. This may not be surprising given the high degree of arboreality in Cyclopes,
which would not lend itself to being readily preserved,
particularly if this adaptation became established early
in the groups history.
Patterson and Pascual (1968, 1972) cite a cranium
with a partial rostrum they referred to Myrmecophaga
from the Montehermosan (early Pliocene) of Catamarca province, Argentina. Hirschfeld (1976) noted that
this specimen is larger than the living form. However,
there is no formal description and no photos or illustrations have been published. The specimen, MLP 31-XI12-15, is currently cataloged as Neotamandua. The skull
is from a juvenile individual; the frontals and parietals
have separated along their common suture and other
sutures are visible. A detailed examination of the specimen indicates that the suture patterns and shape of the
cranial bones do not match any known vermilinguan,
and the dorsal profile of the skull is not compatible with
its interpretation as a member of this group. We do not
consider this specimen to be a vermilinguan.
Nunezia caroloameghinoi was also described from
the Montehermosan (Kraglievich 1934). Nunezia comes
from the type locality for this age and was recovered
from lithostratigraphic unit 1, which Tonni et al. (1992)
consider the type for the Montehermosan. This taxon
is currently considered a synonym of Myrmecophaga
(McKenna and Bell 1997). By extrapolation, C. Ameghinos (1919) Neotamandua magna should also be considered Myrmecophaga because Kraglievich (1934) transferred this taxon to Nunezia. If these synonomies are
71
correct, then the earliest record of Myrmecophaga is in

the Huayquerian. The next record of Myrmecophaga is
from the early Irvingtonian (roughly equivalent to the
Uquian) locality of El Golfo in Mexico (Shaw and McDonald 1987). The El Golfo specimen is a third metacarpal that permits direct comparison with the type of
Nunezia caroloameghinoi, also a third metacarpal, but
not with Neotamandua magna, which is known only
from a pelvis.
In the Lujanian (late Pleistocene/early Holocene),
we have good records of both Tamandua tetradactyla
and Myrmecophaga tridactyla, based primarily on specimens recovered from the caves of Minas Gerais, Brazil,
by P. W. Lund (Winge 1915), with additional records of
Myrmecophaga from Bahia, Brazil (Cartelle and Leite
1989), and Uruguay (Mones and Ximenez 1980). Tamandua tetradactyla is known from three other localities (all caves) in Bahia, Brazil (Cartelle and Leite 1989).
There are no fossil records of either T. mexicana or Cyclopes didactylus.
Phylogeny of the Vermilingua

The most recent phylogeny of the Vermilingua is the
cladistic study by Gaudin and Branham (1998) utilizing 107 osteological characters of the skull and skeleton (figure 6.2). Along with the three extant genera,
they were able to include only three extinct genera in
their analysis (Palaeomyrmidon, Protamandua, and
Neotamandua) because most of the other putative vermilinguan taxa consist of single bones. Their analysis
essentially confirms the conclusions of previous workers such as Hirschfeld (1976) and Engelmann (1985), in
which Palaeomyrmidon is the sister group of Cyclopes,
Protamandua is the sister group to the other Vermilingua taxa, and Neotamandua is the sister group to Myrmecophaga. It is unfortunate that more fossil taxa were
not included in the analysis. As already mentioned, this
was due partly to the general paucity of fossil Vermilingua, but inclusion was further hampered by lack of any
recent detailed studies of existing material or re-evaluations of alpha-level taxonomy.
Using a molecular clock, Delsuc et al. (2001) calculated divergence times for various stages in the evolution
of the Xenarthra, including events within Vermilingua
(table 6.1). While it may never be possible to obtain
DNA from the older Vermilingua taxa in order to refine this phylogeny using molecular data, their work
does provide a hypothesis that can be tested by the fos-
72
Figure 6.2. Hypothesized phylogeny of the Vermilingua based on PAUP analysis utilizing 107 osteological
characters. Nodes are: 1-Vermilingua, 2-Cyclopedidae,
3-Myrmecophagidae, 4-Tamandua, Neotamandua, and
Myrmecophaga, 5-Myrmecophaga and Neotamandua.
Modified from Gaudin and Branham (1998).
sil record. However, this is difficult at present because,

as can be seen in table 6.1, there are significant gaps
in the vermilinguan fossil record. Thus, at this time,
fossil data are insufficient to answer some very critical
questions about vermilinguan evolutionary history. For
example, we cannot say at what stage the Vermilingua
lost their dentition and the pattern (if any) of tooth reduction. Regarding Tamandua, an unresolved question
is whether its semiarboreal habits represent a primitive
retained condition or were derived from a terrestrial
ancestor. Neither is it known whether the terrestrial
habits of Myrmecophaga represent a derived condition
from a semiarboreal ancestor. Continued field work
and detailed description and analysis of new finds will
be necessary in order to utilize the fossil record to understand the evolutionary history and phylogenetic relationships of these taxa, including the timing of major
divergences and the acquisition of major adaptations.
Despite these problems, the historical record does
suggest certain plausible evolutionary scenarios. Although there is considerable variation in size and lo-
comotor habits (terrestrial to fully arboreal), all extant

forms possess a generally uniform morphology, suggesting a strong phylogenetic constraint on the groups
musculoskeletal anatomy (Taylor 1985). The presence of
a prehensile tail like that of Cyclopes and Tamandua in
the Santacrucian Protamandua, the oldest well-known
described anteater, plus the fact that two of the three
modern genera have arboreal adaptations, led Gaudin
and Branham (1998) to infer that arboreality is likely
the primitive condition for vermilinguans. If the ancestral form was semiarboreal, like extant Tamandua,
then its distribution was probably restricted to closed
canopy forests. Therefore, the evolutionary appearance
of Myrmecophaga may be closely tied to the development of savanna and open-country habitats in South
America, and with the appearance of ground-dwelling
termites and other colonial insects in these same areas.
Woodland savanna was present in South America from
the late Paleocene to early Miocene, but by the Eocene
open savanna woodlands with abundant grasses and
shrubby elements had differentiated from tropical
Table 6.1. Chronologic Distribution of Fossil Vermilingua

South American
Land Mammal
Age (SALMA)
Duration
(mya)
Taxa present
Present
Myrmecophaga tridactyla

Tamandua tetradactyla

Tamandua mexicana

Cyclopes didactylus
Lujanian
0.50.01

Ensenadan
10.5

GAP
Irvingtonian
3.81.5
(Uquian)

GAP
Chapadmalalan
43.8
Montehermosan
74
Myrmecophaga (= Nunezia)

caroloameghinoi
Huayquerian
97
Palaeomyrmidon incomptus

Neotamandua greslebini

Neotamandua conspicua
Chasicoan
109
Mayoan
1210
Laventan
14.212
Neotamandua borealis

GAP
Colloncuran
15.514
Neotamandua? australis
Friasian
16.515.5
Santacrucian
17.516.5
Protamandua rothi

GAP
Colhuehuapian
2119
Undescribed taxon

GAP
Deseadan
2925

GAP
Tinguirirican
3631.5

GAP

4240
Divisaderan

GAP
Mustersan
4845

GAP
Casamayoran
5451

GAP
54.855.5
Riochican

GAP
Itaboraian
5957.5

GAP
Peligran
62.561

GAP
64.563
Tiupampan
Time since divergence
Giant anteater from collared

anteater 13 (1512) mya
Pygmy anteater from other

anteaters 38 (4535) mya
Vermilingua from sloths 54

(6551) mya
Cingulata and Pilosa diverge

63 (7659) mya
Sources: Time of divergence of major taxa based on the molecular data of Delsuc et al. (2001) and Barros et al. (2003). South American
Land Mammal Age chronology based on Flynn and Swisher (1995).
74
rainforest (Webb 1978). Some open park savanna was

present by the earlymiddle Miocene (Santacrucian ca.
17.516.5 mya), and Webb (1978) noted that at this time
anteaters (Vermilingua) were abundant, along with the
myrmecophagus armadillo, Stegotherium, thus suggesting abundant termitaria. This slightly precedes the
estimate of Delsuc et al. (2001) for the divergence of
the Tamandua and Myrmecophaga lineages (table 6.1),
but is suggestive that expansion of South American savanna habitat played an important role in the evolution
of Myrmecophaga. Given the importance of colonial insects living in open savanna habitat in the diet of extant
Myrmecophaga (Rodrigues et al. this volume), it is safe
to assume its evolutionary history is closely tied to the
evolution of this ecosystem in South America.
The fossil record provides little help in understanding other aspects of the evolution of Myrmecophaga. For
example, the earliest preserved cranial material of the
giant anteater already resembles the modern form, so
provides no insight into how the extreme modification
of the skull was achieved or how rapidly it occurred.
Likewise, except for the late occurrence of Palaeomyrmidon in the Huayquerian, very little is known about
the ancestry of Cyclopes and when it became an obligate arboreal form. Given the high degree of specialization in the living form, its evolution must be closely
tied to the history of the Neotropical rain forest. Both
Cyclopes and Tamandua have evolved prehensile tails
as part of their adaptation to an arboreal habitat. This
may be a primitive feature retained by these two taxa
from Protamandua but secondarily lost in the terrestrial Myrmecophaga, or, given their early divergence, it
is possible this feature was independently acquired in
each lineage.
An equally fruitful area of potential research is to
thoroughly review the numerous Santacrucian taxa
currently placed in the families Adiastaltidae and Anathitidae or as Xenarthra incertae sedis, although this
will definitely require the discovery of additional material. In the past these taxa have been variously referred
to Pholidota and Monotremata, as well as to Vermilingua. Although the sample is limited, the robustness of
the preserved humeri suggest there may have existed
in the Santacrucian a branch within Vermilingua that
was more fossorially adapted than any other members
of the group (Vizcano, Bargo, and Kay 2004). A better
understanding of the range of morphological adaptations within Vermilingua and their evolutionary history, particularly as related to the utilization of colonial
insects, can certainly enhance our ability to compare
them with myrmecophagous mammals from other

parts of the world, and consequently provide more insight into this fascinating example of convergent evolution.
Acknowledgments
Mary Ann Turner kindly provided the locality information on the Yale Peabody Museum specimen of
Protamandua. Bill Simpson of the Field Museum of
Natural History generously provided the locality information on their anteater specimens. HGM extends his
appreciation to Kim Aaris-Srensen and Knud Rosenlund of the Zoological Museum of the University of
Copenhagen for their hospitality during his visits to
study the Lund collection and Knud Rosenlund and Per
Christiansen for providing additional locality information on Vermilinguan specimens in the Lund collection. HGM also wishes to express his appreciation to
Castor Cartelle for his help and hospitality during visits
to Belo Horizonte to study those collections. SFV and
MSB are grateful to Walter Joyce and Dan Brinkman
from Yale Peabody Museum, and Alejandro Kramarz
(Vertebrate Paleontology Department) and Olga Vaccaro (Mammalogy Department) of the Museo Argentino de Ciencias Naturales de Buenos Aires for allowing
study of the collections under their care. The authors
wish to express their appreciation for the insightful
comments of the two reviewers, T. J. Gaudin and G. J.
Scillato-Yan, which added to the quality of this review.
This is a contribution to the projects of SFV from Universidad Nacional de La Plata N 474, PIP-CONICET
5240, and PICT 26219.
Appendix 6.1. Synoptic list of fossil Vermilingua

An extinct taxon is indicated by . Institutional Abbreviations: FMNH = Field Museum of Natural History,
Chicago; MACN = Museo Argentino de Ciencias Naturales, Buenos Aires; MCL = Museu de Cincias Naturais, Belo Horizonte; MLP = Museo de La Plata, La
Plata; MNHN = Museo Nacional de Historia Natural,
Montevideo; SDMNH = San Diego Museum of Natural History, San Diego; MPT = Museo Provincial de
Tucumn, San Miguel de Tucumn; UCMP = University of California Museum of Paleontology, Berkeley;
YPM-PU = Yale Peabody Museum, Princeton University collection, New Haven; ZMUC = Zoological Museum of the University of Copenhagen, Copenhagen.
Family Cyclopedidae
Cyclopes Gray 1821
Species: Cyclopes didactylus Linneaus 1758
No fossil record
Palaeomyrmidon Rovereto 1914

Species: Palaeomyrmidon incomptus Rovereto 1914
Locality: Andalhuala, Catamarca Province, Argentina
Material: Type = MACN 8098 nearly complete skull
Age: late Miocene, Huayquerian, Araucano Formation
Reference: Rovereto 1914
Species: Palaeomyrmidon?
Locality: Valley of Santa Mara, Catamarca Province,
Argentina
Material: MACN 2934 left proximal tibia, and both distal
portions; right proximal femur and left distal femur.
Age: late Miocene, Huayquerian (= Araucanense)
Reference: unpublished
Comment: collected by J. Mndez, 1908.
Family Myrmecophagidae
Myrmecophaga Linneaus 1758
Species: Myrmecophaga tridactyla Linneaus 1758
Locality: Lapa da Escrivania No. 1, Lagoa Santa, Minas
Gerais, Brazil
Material: ZMUC 3407 posterior part of skull, 3742 right
astragalus, 3743 second phalanx digit III manus, 4681
proximal caudal, 4677 patella
Age: Pleistocene
Reference: Winge 1915
Locality: Lapa da Cerca Grande No. 2, Minas Gerais, Brazil
Material: ZMUC 11007 distal epiphysis left tibia, 11041 neural arch of thoracic, 11043 distal end metapodial, 3572
second phalanx digit III manus, 11042 left scaphoid
Age: Pleistocene
Locality: Lapa da Escrivana No. 5, Lagoa Santa, Minas
Gerais, Brazil.
Material: Probably a single individual represented by
ZMUC 5243 left femur, 5622 right third metacarpal,
5247 left second metatarsal, 5672 left tibia, 5689 proximal left humerus, 5362 distal right radius, 4245 distal
right fibula, 5244 distal half left tibia, 8885/7234 left acetabulum, 12382 right ulnar, 5245 left third metacarpal,
11890 right third metatarsal, 5673 thoracic, 5361, 11889
lumbar, 5625 caudal, No # proximal right femur, 8652
diaphysis, 5246 left fourth metatarsal, 11888 head femur,
5623 second phalanx, 9743 distal end right second
metacarpal, 5624 ungual; 10761 posterior skull of juve-
75
nile, 9733 left astragalus juvenile, 8858 left calcaneum,

-12 distal epiphysis left humerus
Age: Pleistocene
Locality: Lapa da Escrivana No. 11, Lagoa Santa, Minas
Gerais, Brazil
Material: ZMUC 7354 distal left humerus, No # proximal
left ulna, 10759 midshaft right ulna, No # ungual digit
III manus, 11891 proximal phalanx digit III manus, 6711
second phalanx digit III manus, 6580 distal epiphysis
third metacarpal, 10762 neural arch of thoracic, 10760
proximal epiphysis radius, 12552 right third metatarsal,
12549 right fourth metatarsal, 12481 left fourth metatarsal, 12548 right fourth metacarpal, 12551 left second
metacarpal, 12553 right second metacarpal
Age: Pleistocene
Locality: Lapa dos Tatus, Lagoa Santa, Minas Gerais, Brazil
Material: ZMUC 9410 partial right pelvis, 9385, 9386, 9387,
thoracic, 9330 caudal, 11913 thoracic neural arch, 11892
patella, 9422 right calcaneum, 2071 partial sacrum
Age: Pleistocene
Locality: Lapa Vermelha No.1, Lagoa Santa, Minas Gerais,
Brazil
Material: ZMUC 930 left second metacarpal, 931 second
phalanx, 1916 left astragalus
Age: Pleistocene
Locality: Gruta dos Brejes, Municpio de Morro do Chapu, Bahia, Brazil
Material: MCL 1603 partial skeleton
Age: late Pleistocene, Lujanian
Reference: Cartelle and Leite 1989
Locality: El Golfo, Sonora, Mexico
Material: SDMNH 20323 right third metacarpal
Age: Pleistocene, Irvingtonian
Reference: Shaw and McDonald 1987
Locality: Departamento de Rocha, Costa del Ro Cebollat
(Arrocera de Saglia), Uruguay
Material: MNHN-M 2573 posterior half of skull
Age: Pleistocene
Reference: Mones and Ximenez 1980
Neotamandua Rovereto 1914

Species: Neotamandua? australis Scillato-Yan and Carlini
1998
Locality: Commallo, Rio Negro Province, Argentina
Material: Type = MLP 91-IX-6-5 complete right humerus,
and MLP 91-IX-4-23 distal third of left humerus
Age: middle Miocene, Collocuran (originally Friasian),
Colln Cur Formation
Reference: Scillato-Yan and Carlini 1998
76
Species: Neotamandua conspicua Rovereto 1914

Locality: Andalhuala, valley of Santa Maria, Catamarca
Province, Argentina
Material: Type = MACN 8097 posterior part of cranium
Reference: Rovereto 1914
Species: Neotamandua conspicua Rovereto 1914
Locality: Corral Quemado, Department of Belen, Catamarca Province, Argentina
Material: FMNH P14419 skull, jaws, ulna
Age: late Miocene-Pliocene, Corral Quemado Formation
Reference: Gaudin and Branham 1998, Patterson et al. 1992
Comment: Patterson et al. (1992) suggest this material is
congeneric with Myrmecophaga, but Gaudin and Branham (1998) provisionally accept it as a separate genus.
Species: Neotamandua greslebini Kraglievich 1940b
Locality: Topunco, Tucuman Province, Argentina
Material: MACN no holotype designated and the material
cannot currently be found in the collections; referred
specimens proximal third of both femora, distal end left
femur, body of right astragalus, left calcaneum, distal
fragment (condyle and epicondyle) left humerus, proximal right ulna, distal right radius
Reference: Kraglievich 1940b
Comment: No catalog number was given at the time the
species was described and no mention of this species
is available in the museum catalog; it is possible that
the material listed above is the same as Neotamandua?
MACN 2403, 2405, 2406, 2407, 2408, and Paleomyrmidon? MACN 2934 listed below, but it is not currently
possible to confirm this.
Species: Neotamandua borealis Hirschfeld 1976
Locality: San Nicholas locality, Department of Huila, La
Venta, Colombia
Material: Type = UCMP 39847 incomplete manus, pes, and
fragmentary postcranial elements of a mature individual
Age: middle Miocene, Laventan, Honda Formation
Reference: Hirschfeld 1976
Species: Neotamandua magna C. Ameghino 1919
Locality: Topunco, Tucuman Province, Argentina
Material: Type = MPT 58 pelvis
Reference: C. Ameghino 1919, p. 152, lam. VI
Comment: Kraglievich (1934) transferred this to his genus
Nunezia
Species: Neotamandua?
Material: MACN 2403 7 fragments of bone (includes frag-
ments of diaphysis and fragment of the olecranonean

apophysis, which is left; Kraglievich indicated it as right).
Comment: collected by J. Mndez, 1911
Material: MACN 2405 left caput humeralis
Material: MACN 2407 left proximal femur.
Material: MACN 2408 right distal radius
Material: MACN 2406 right astragalus, incomplete
Material: MACN 2411 left calcaneum, incomplete
Nunezia Kraglievich 1934

Species: Nunezia caroloameghinoi Kraglievich 1934
Locality: Monte Hermoso, Buenos Aires Province, Argentina
Material: Type = MACN 6215 left third metacarpal
Age: late Miocene/early Pliocene, Montehermosan
Reference: Kraglievich 1934, Tonni 1974
Comment: McKenna and Bell (1997) include Nunezia as a
junior synonym of Myrmecophaga
Promyrmephagus Ameghino 1904

Species: Promyrmephagus dolichoarthrus Ameghino 1904
Locality: Santa Cruz, Argentina
Material: Type = MLP ? right astragalus (specimen lost).

Cast of the type in MACN A-4621b.
Age: early/middle Miocene, Santacrucian
Reference: Ameghino 1904, Hirschfeld 1976
Species: Promyrmephagus euryarthrus Ameghino 1904
Material: Type = MACN A-11530 right astragalus
Reference: Ameghino 1904, Hirschfeld 1976
Protamandua Ameghino 1904

Species: Protamandua rothi Ameghino 1904
Material: Type = right calcaneum (specimen lost). Cast of
the type in MACN A-10901b.
Age: early/middle Miocene, Santacrucian, Santa Cruz
Formation
Reference: Ameghino 1904
Comment: Hirschfeld (1976) noted that the type of P. rothi
articulates with the type of Promyrmephagus euryarthrus and is not only the same taxon but probably also
the same individual.
Species: Protamandua rothi
Locality: La Costa farm, 8 miles south of Coyle, Santa Cruz
Province, Argentina
Material: FMNH P13134 partial skull with postcranial
elements including calcaneum and astragalus (material
currently missing)
Formation
Reference: Patterson et al. 1992, Gaudin and Branham 1998
Comment: collected by G. F. Sternberg, 1923
Species: Protamandua rothi
Locality: Ger Aike Department, 10 miles south of Coy
Inlet, Santa Cruz Province, Argentina
Material: YPM-PU 15267 partial skull
Formation
Reference: Patterson et al. 1992, Gaudin and Branham 1998
Comment: collected by O. A. Petersen
Tamandua Gray 1825

Species: Tamandua tetradactyla Linneaus 1758
Locality: Toca dos Ossas, Municpio de Jacobina, Bahia, Brazil
Material: MCL 1604 skull, 30 vertebrae, 23 rib fragments,
partial scapula, right radius, both ulnae, partial pelvis,
left femur, both tibiae, ungual from digit 3 pes
Age: Pleistocene, Lujanian/Holocene
Locality: Gruta dos Brejes, Municpio de Morro do Chapu, Bahia, Brazil
77
Age: Pleistocene
Locality: Gruta do Impossvel, Municpio de Palmeiras,
Bahia, Brazil
Age: Pleistocene
Locality: Lapa do Bahu No. 1, Lagoa Santa, Minas Gerais,
Brazil
Material: ZMUC 780 left fourth metatarsal
Age: Pleistocene
Species: Tamandua mexicana Saussure 1860
No fossil record
Myrmecophagidae indeterminate
Locality: Santa Cruz Province, Argentina
Material: MLP 69-IX-8-8a left humerus lacking the proximal epiphysis.
Formation.
Reference: Vizcano, Bargo, et al. 2004
Comment: Lydekker (1894) attributed this specimen to the
armadillo Peltephilus (Peltephilidae). Recently it was
referred to Myrmecophagidae by Vizcano, Bargo, et al.
2004.
Family Adiastaltidae
Adiastaltus Ameghino 1893
Species: Adiastaltus habilis Ameghino 1893
Locality: Corriguen Aike, Santa Cruz Province, Argentina
Material: Type = MACN A-7772 almost complete left humerus with the deltoid tuberosity broken, and A-7773
left ulna.
Formation
Reference: Ameghino 1893, 1894
Comment: collected by C. Ameghino, 18911892. The size
and morphology of this specimen fits with the description of the type of Adiastaltus habilis, Ameghino 1893.
Locality: Killik Aike, Rio Gallegos, Santa Cruz Province,
Argentina
Material: MACN A-5801 distal humerus
Reference: Ameghino 1894, Hoffstetter 1958
Comment: collected by C. Ameghino, 1891. This specimen
was mislabeled as the type of Adiastaltus habilis.
Species: Adiastaltus procerus Ameghino 1894
Locality: Monte Observacin, Santa Cruz Province, Argentina
78
Material: Type = MACN A-7779, distal humerus

Comment: collected by C. Ameghino, 18901891
Plagiocoelus Ameghino 1894

Species: Plagiocoelus obliquus Ameghino 1894
Material: Type = MACN A-7780, distal humerus
Formation
Species: Plagiocoelus? latus
Material: MACN A-7781, distal humerus
Formation
Comment: collected by C. Ameghino, 18901891. It appears that this species was never formally described.
The specimen cannot be assigned to Plagiocoelus and is
probably not a vermilinguan.
Family Anathitidae
Anathitus Ameghino 1894
Species: Anathitus revelator Ameghino 1894
Material: Type = MACN A-7782 distal right humerus
(pathological), A-7783 lumbar (missing)
Reference: Ameghino 1893, 1894
Family Manidae Ameghino 1904,

Mammalia incertae sedis Mones 1986
Argyromanis
Species: Argyromanis patagonica Ameghino 1904
Locality: Santa Cruz, argentina
Material: Type = MACN A-11687 left astragalus
Orthoarthrus Ameghino 1904

Species: Orthoarthrus mixtus Ameghino 1904
Locality: Argentina
Material: Type = MLP 69-IX-5-16 left astragalus
Comment: originally described as a pangolin.
7
The evolution of armored xenarthrans
and a phylogeny of the glyptodonts
Juan Carlos Fernicola, Sergio F. Vizcano, and Richard A. Faria
filtico, con la inesperada estrecha relacin de Neosclerocalyptus y Panochthus por un lado y de Doedicurus
y Glyptodon, por el otro. Futuros anlisis filogenticos
deberan incluir ambos tipos de informacin, es decir
caracteres craneanos y poscraneanos. Sin embargo, la
fuerte sustentacin de algunos clados (como Panochthus + Neosclerocalyptus) permite sugerir que las propuestas tradicionales deben ser modificadas.
Resumen
Los xenartros con coraza (Cingulata) son uno de los
grupos ms caractersticos de los mamferos fsiles sud
americanos desde el principio del Terciario y tambin
son conocidos en Amrica Central y del Norte. Considerados un grupo natural, han sido tradicionalmente
divididos en armadillos (incluyendo a los extintos
pampaterios) y los completamente extinguidos gliptodontes. El primer anlisis cladstico reconoci una dicotoma basal, con los pampaterios, armadillos eutatinos y gliptodontes como el grupo hermano del resto de
los armadillos. Un anlisis ms reciente, basado en caracteres craneodentales, no sustenta esta dicotoma ni
la monofilia de los diferentes agrupamientos clsicos.
Adems, la clasificacin y la filogenia tradicionales de
los gliptodontes estaban basadas en el estudio de la superficie exterior de las placas seas y del tubo caudal, al
contrario que en el enfoque comn para los mamferos,
en el cual los caracteres craniodentales son fuente privilegiada de informacin. Un reciente anlisis cladstico
de este destacablemente uniforme grupo monofiltico,
basado en 84 caracteres craniodentales, desafi la divisin tradicional en hasta cinco grupos a los niveles
de familia o subfamilia: Propalaehoplophoridae/nae,
Glyptatelinae, Hoplophoridae/nae (= Sclerocalyptidae/
nae), Glyptodontidae/nae y Doedicuridae/nae. En este
enfoque, se encontr una dicotoma basal de los gliptodontes, basada en cinco sinapomorfas inequvocas, con
la subfamilia Propalaehoplophorinae, tradicionalmente
considerada parafiltica, como el grupo hermano del
resto. Ms aun, la monofilia de la subfamilia Hoplo
phorinae y de sus tribus no fue sustentada. Los gneros
Neosclerocalyptus, Plohophorus, Doedicurus, Glyptodon,
Panochthus y Urotherium conforman un grupo mono-
Resumo
Os xenarthros com carapaa (Cingulata) so um dos
grupos mais caractersticos dos fsseis de mamferos
sul-americanos desde o incio do Tercirio e tambm
so conhecidos na Amrica Central e do Norte. Considerados um grupo natural, tm sido tradicionalmente
divididos em tatus (incluindo os extintos pampatrios)
e os completamente extintos gliptodontes. A primeira
anlise cladstica reconheceu uma dicotomia basal,
com os pampatrios, tatus eutatinos e gliptodontes
como o grupo irmo do resto dos tatus. Uma anlise
mais recente, baseada em caracteres crnio-dentais no
sustenta esta dicotomia nem a monofilia dos diferentes
agrupamentos clssicos. Tambm, a classificao e a
filogenia tradicionais dos gliptodontes estavam baseadas no estudo da superfcie exterior das placas sseas e
do estojo caudal, ao contrrio que no enfoque comum
para os mamferos, segundo o qual os caracteres craniodentais so fonte privilegiada de informao. Uma
recente anlise cladstica deste muito uniforme grupo
monofiltico, baseada em 84 caracteres crnio-dentais,
desafiou a diviso tradicional em at cinco grupos nos
nveis de famlia ou subfamlia: Propalaehoplophoridae/nae, Glyptatelinae, Hoplophoridae/nae (= Sclerocalyptidae/nae), Glyptodontidae/nae e Doedicuridae/
79
80
J. C. Fernicola, S. F. Vizcano, and R. A. Faria
nae. Nesta aproximao, encontrou-se uma dicotomia

basal dos gliptodontes, baseada em cinco sinapomorfias inequvocas, com a subfamlia Propalaehoplophorinae, tradicionalmente considerada parafiltica, como
o grupo irmo do resto. Ainda mais, a monofilia da
subfamlia Hoplophorinae e de suas tribos no foi
sustentada. Os gneros Neosclerocalyptus, Plohophorus, Doedicurus, Glyptodon, Panochthus e Urotherium
conformam um grupo monofiltico, com a inesperada
estreita relao de Neosclerocalyptus e Panochthus, por
uma parte, e de Doedicurus e Glyptodon, pela outra. Futuras anlises filogenticas deveriam incluir ambos os
tipos de informao, isto , caracteres cranianos e pscranianos. Porm, a forte sustentao de alguns clados
(como Panochthus + Neosclerocalyptus) permite sugerir
que as propostas tradicionais devem ser modificadas.
Introduction
Armored xenarthrans (Cingulata) constitute one of
the most characteristic groups of fossil mammals in
South America, although they are also found in Central and North America. Traditionally they have been
divided into two groups, informally known as armadillos and glyptodonts. The former include living and
many fossil forms classified as dasypodids and pampatheres. In contrast to the modest current representation of the cingulates (8 or possibly 9 genera; Wetzel
1985; Vizcano 1995; Gardner 2005), the fossil record
is very rich. Indeed, more than 100 genera have been
described, 65 corresponding to glyptodonts (McKenna
and Bell 1997). The oldest glyptodont remains, assigned
to the genus Glyptatelus, include isolated scutes from
the middle Eocene of Patagonia, Argentina (Pascual et
al. 1996), while Riostegotherium yanei, from the middle
Paleocene of So Jos de Itabora, Brazil, is not only the
oldest armadillo, but also the oldest xenarthran known
(Bergqvist et al. 2004).
No doubt, glyptodonts represent one of the most
bizarre groups of mammals ever to have evolved. During the terminal Pleistocene, these herbivorous animals
reached gigantic sizes, with forms that varied between
one and two tons of body mass, yet they were still able
to adopt bipedal postures to perform strenuous activities such as intraspecific fights (Faria 1995; Faria et
al. 1998; Alexander et al. 1999). Their most remarkable
anatomical features are the presence of an essentially
immobile dorsal carapace, different numbers of trilobate teeth, elephantine hind-limbs, and a masticatory
apparatus telescoped well below the cranium (Hoffstetter 1958; Faria and Vizcano 2001).
Historical background
The first mention of a South American armored xenarthran goes back to the mid-eighteenth century. It was
Thomas Falkner (1774), an English Jesuit missionary,
who reported discovering a great carapace at least three
meters long made up of hexagonal scutes and comparable to that of armadillos, in clear reference to what we
today know as a glyptodont.
Albeit indirectly, Dmaso Antonio Larraaga provided the first scientific description of a great armored
extinct animal. This fossil material, composed of a femur, carapace fragments and a caudal tube was assigned
to Dasypus (Megatherium Cuv.) by Larraaga, and as
such included by Cuvier (1823) in his second edition
of Recherches sur les ossements fossiles. This subgeneric
assignment was advocated because the original material of Megatherium americanum (a giant fossil sloth,
see McDonald and De Iuliis this volume) was collected
jointly with armor fragments (see Mndez Alzola 1950,
and references therein). Beyond the reasons that induced Larraaga to recognize the existence of armored
megatheriids, the fact that his opinion was published
in Cuviers work (1823) led to the similar categorization
of several incomplete fossils described later (e.g., Weiss
1830; Clift 1835).
In 1838, Richard Owen named the genus Glyptodon
from material collected in Buenos Aires and shipped to
London by the English Secretary Woodbine Paris. One
year later, Owen revised all the references to armored
megatheres and established that the big armored animals were glyptodonts, related to the armadillos, and
that megatheres did not possess a carapace (Owen
1839). Also during this time, knowledge and collection of large armored mammals were increasing in the
Ro de la Plata region. The pioneering descriptions and
functional studies by Teodoro Vilardeb and Bernardo
Berro in Uruguay and Francisco P. Muiz in Argentina, although not very well known, are of particular
interest. They collected many fossil mammals, including glyptodonts (Ameghino 1889; Mndez Alzola 1950),
which were sent to France during the first half of the
nineteenth century. These fossils, along with material
from other sources, ultimately became known to science through the extensive and definitive monograph
published by Nodot (1857), who described glyptodonts
The evolution of armored xenarthrans and a phylogeny of the glyptodonts
in great detail. Some years later, Burmeister (18701874)

wrote a detailed monograph on the comparative anatomy of glyptodonts, taking a complete specimen of
Panochthus as the main reference.
In regard to armadillos, the first record corresponds
to a giant armadillo, Chlamydotherium (= Pampatherium), discovered and named by Lund (1839) in Brazil. During most of the nineteenth century, the scarce
information about fossil armadillos strongly contrasts
with the detailed anatomical and morphological knowledge available on glyptodonts. It is possible that this
disparity was due to the greater interest in large specimens during those times. In any case, it was Florentino
Ameghino (1889, 1891) who increased our knowledge
of fossil armadillos considerably. Prior to Ameghino,
several authors (Darwin 1839; Pictet 1853; Flower 1882;
Schmidt 1887) recognized glyptodonts as ancestors of
recent armadillos. This relationship was rejected by
Ameghino (1884, 1889), who maintained the opposite,
based on a set of morphological and evolutionary inferences. In short, arguing that the fusion of bony elements should be considered a derived characteristic,
Ameghino (1889) postulated that the rigid carapace of
the glyptodonts should have derived from the flexible
armor of the dasypodids, and hence the inverse evolutionary relationship was proposed. At the same time,
but using different characters of the dermal carapace,
Ameghino (1889) presented the first phylogeny of the
glyptodonts, including almost all known taxa. Despite
some later anatomical descriptions (e.g., Scott 1903
1904), this seminal work shifted the study of cingulates
to phylogeny and notably influenced all genealogic and
classificatory proposals during the twentieth century
(Castellanos 1932; Simpson 1945; Hoffstetter 1958; Paula
Couto 1979; McKenna and Bell 1997).
Cingulate lineages
Because of the unique postcranial morphology of Cingulata, these peculiar mammals were initially considered a natural group (Ameghino 1889; Hoffstetter
1958; Engelmann 1978, 1985). Many of these characteristics (e.g., dermal armor) were discussed recently by
McDonald (2003b). In addition, Gaudin (1995, 2004a)
and Gaudin and Wible (2006) have described a group
of derived cranial features that identify the Cingulata.
All these data, plus recent molecular studies (see Delsuc
et al. 2003 and references therein) unambiguously support the monophyletic origin of Cingulata.
81
Traditionally, cingulates have been split into two taxonomic groups: the dasypodids, including armadillos
and pampatheres, and glyptodontids (Ameghino 1889;
Simpson 1945; Hoffstetter 1958; Paula Couto 1979; Scillato-Yan 1980; see figure 7.1). Engelmann (1978, 1985)
was the first author to employ cladistic methodology
to work out cingulate phylogeny. He examined a wide
range of dasypodids, among them pampatheres, and
glyptodonts, with the latter treated as a terminal taxon.
Engelmann (1978, 1985) recognized a basal dichotomy
of Cingulata but with a different taxonomic composition. He included pampatheres in his Glyptodonta
group as the sister group to Glyptodontoidea, formed
by eutatine armadillos and glyptodonts, although he
recognized that the hypothesis that joins pampatheres
and glyptodonts was also possible. The remaining living and extinct armadillos were included in the group
Dasypoda. Recently, Gaudin and Wible (2006) reanalyzed the genealogical relationships between extant and
extinct armadillos, pampatheres, and glyptodonts, with
the latter represented by just the basal Miocene genus
Propalaehoplophorus. This cladistic analysis was based
on a morphological analysis of 163 craniodental characters. Their proposed phylogenetic scheme did not
support any basal dichotomy nor the monophyly of
different classic clusters (e.g., eutatine and euphractine
armadillos) recognized by earlier authors (Simpson
1945; Hoffstetter 1958; Paula Couto 1979; Scillato-Yan
1980; Engelmann 1978, 1985; Carlini and Scillato-Yan
1996; McKenna and Bell 1997). On the other hand, the
alliance of Dasypus and Stegotherium was clearly supported, a relationship proposed by Engelmann (1978,
1985) and Patterson et al. (1989).
With respect to the sister group of glyptodonts,
Gaudin and Wible (2006) found strong support for a
clade formed by pampatheres and glyptodonts. This
hypothesis, discarded by Engelmann (1978, 1985) and
Hoffstetter (1958), had been previously outlined by
Ameghino (1889), Castellanos (1937), Patterson and
Pascual (1968, 1972), and Patterson et al. (1989).
The genus Palaeopeltis has been controversial. It is
the only representative of the family Palaeopeltidae,
and was considered closely related to glyptodonts by
Ameghino (1895), based on scutes. Later, Kraglievich
and Rivas (1951) described parts of a skull and scutes
from the head shield and carapace. These authors believed that the teeth of Palaeopeltis agreed with the holotype of Orophodon, previously described as a sloth by
Ameghino (1895). Thus, they relegated Palaeopeltis to
82
Figure 7.1. Drawing of cingulates.

A. Glyptodon.
B. Giant armadillo
Pampatherium.
C. Armadillo
Eutatus.
synonymy and founded the superfamily Orophodontoidea, placing it in an intermediate position between
dasypodids and glyptodonts within Cingulata. Later,
Hoffstetter (1954, 1956) showed that Orophodon was a
ground sloth and assigned the original scutes of Palaeopeltis to two ground sloths: Orophodon and Octodontotherium. He created a new genus, Pseudorophodon, for the specimen described by Kraglievich and
Rivas (1951). Hoffstetter (1958) recognized Pseudorophodon as a terminal branch to the horned peltephiline armadillos. Patterson and Pascual (1968, 1972) discussed the problematic Palaeopeltidae and proposed
that the original scutes and the specimen described by
Kraglievich and Rivas (1951) corresponded to the genus
Palaeopeltis, and that Pseudorophodon was its synonym.
Patterson and Pascual (1968, 1972) did not support the
relationship proposed by Hoffstetter (1958), and placed
this group as a basal component to their pampatheresglyptodonts dichotomy.
The Palaeopeltidae have a very poor fossil record.
Until now, this group has not been considered in any
cladistic analysis. However, the presence of a continuous
dental series, similar to that of the horned armadillo

Peltephilus, supports the opinion of Hoffstetter (1958).
If this is corroborated in the future, then the relationship between Palaeopeltidae and glyptodonts should be
discarded, following Gaudin and Wibles (2006) view,
who placed Peltephilus as the sister taxon to all remaining cingulates.
Phylogeny of glyptodonts
Traditionally, the classification and phylogeny of the
glyptodonts were based on study of the external surface of bony scutes and the morphology of the caudal
tube. This contrasts with other mammals, where skull
features have played a central role. This oversight was
addressed by Fernicola (2005), who studied the craniodental morphology (n = 84 characters) of 12 genera
in order to describe glyptodont phylogeny using cladistic methodology. The analyses included representatives of the following taxa: Propalaehoplophorinae
(Propalaehoplophorus and Eucinepeltus), Doedicurinae (Doedicurus), Glyptodontinae (Glyptodon), and
Hoplophorinae with the following tribes: Hoplophorini

(Neosclerocalyptus, Eosclerocalyptus, and Hoplophractus), Plohophorini (Plohophorus, Stromaphorus, and
Pseudoplohophorus), Panochthini (Panochthus), and
Lomaphorini (Urotherium). In accord with Engelmann
(1978, 1985) and Gaudin and Wible (2006), the genera
Pampatherium, Eutatus, Euphractus, and Dasypus were
included as outgroups.
Historically, glyptodonts have been considered a
natural group characterized by a remarkable morphological uniformity (Engelmann 1978, 1985). This author
recognized as unique characters of the group the absence of mobile transverse bands in the dorsal carapace, strongly trilobated last cheek teeth, presence of
a bony groove defining the course of the optic nerve,
and a very short face. These last three characteristics,
together with 12 others, were identified by Fernicola
(2005) as unambiguous synapomorphies that partially
support the monophyly of glyptodonts. Support was
only partial because Fernicola (2005) did not include
representatives from all taxa of rank higher than genus
in his analyses.
In the past, glyptodonts were split into at least four
groups at the familial (Ameghino 1895; Castellanos
1932, 1940) or subfamilial level (Simpson 1945; Hoffstetter 1958; Paula Couto 1979; McKenna and Bell 1997):
Propalaehoplophoridae/nae, Hoplophoridae/nae (=
Sclerocalyptidae/nae), Glyptodontidae/nae, and Doedicuridae/nae. However, a fifth subfamily, Glyptatelinae,
erected by Castellanos (1932, 1940), was also recognized
by Hoffstetter (1958), Scillato-Yan (1977), Carlini et al.
(1997), and McKenna and Bell (1997), while Simpson
(1945) and Paula Couto (1979) included the corresponding genera in the Propalaehoplophorinae.
The taxonomy of the oldest glyptodont group, the
glyptatelines, was discussed recently by Vizcano,
Rinderknecht, and Czerwonogora (2003). These authors included in the glyptatelines the genera Clypeotherium, Glyptatelus, Neoglyptatelus, and, tentatively,
Pachyarmatherium, a taxon previously described by
Downing and White (1995) as a dasypodid. With regard
to the last three taxa, some cranial elements have been
described, although of doubtful taxonomic assignment.
Among the elements confidently assigned to these taxa,
scutes revealed a carapace without mobile bands (Vizcano, Rinderknecht, and Czerwonogora (2003). Although not always set in a phylogenetic context, this
character has been considered diagnostic of glyptodonts by several authors (e.g., Burmeister 18701874,
Flower 1882, Ameghino 1889, Hoffstetter 1958, Engel-
83
mann 1978, 1985, Vizcano, Rinderknecht, and Czerwonogora (2003). The presence in Pachyarmatherium
and Neoglyptatelus of free, dasypodid-like dorsal vertebrae (Carlini et al. 1997; Vizcano, Rinderknecht, and
Czerwonogora (2003) contrasts with the different fusion grades observed in the vertebral column in the remaining glyptodonts (Hoffstetter 1958; Gillette and Ray
1981). This mixture of glyptodont-like and dasypodidlike characters suggests placement of the Glyptatelinae
in its own taxonomic rank and a possible basal position
within the glyptodonts. This hypothesis needs to be further assessed in future phylogenetic studies.
Phylogenetic relationships among the remaining
glyptodont groups, as proposed by Ameghino (1889),
Castellanos (1931, 1932), and Hoffstetter (1958), were
tested by Fernicola (2005) and the results do not support the monophyly of the classic groups. Fernicolas
(2005) study yielded three hypotheses of maximum
parsimony. This was attributed to the fluctuating position of the genus Urotherium in the derived portion of
the tree (figure 7.2). The exclusion of this taxon yielded
only one hypothesis of phylogenetic relationships,
which will be the basis of the following discussion. It is
not the aim of this chapter to provide a comprehensive
analysis, as in Fernicola (2005). Instead, the approach
will involve examination of the most strongly supported groups. However, it is worth mentioning that
trees consistent with the classifications of Castellanos
(1932), Simpson (1945), Hoffstetter (1958), and Paula
Couto (1979) were as much as 23 to 25 steps longer than
the most parsimonious tree of Fernicola (2005).
Within the glyptodonts, Fernicola (2005) proposed,
for the first time, the existence of a basal dichotomy,
with the subfamily Propalaehoplophorinae positioned
as the sister group of the remaining glyptodonts (figure
7.2). This contradicts the traditional paraphyly of the
group proposed by, among others, Ameghino (1889),
Castellanos (1931, 1932), and Hoffstetter (1958). These
authors concurred, although with slight differences, in
deriving the different groups of glyptodonts from different genera of Propalaehoplophorinae. In Fernicolas
analysis, five unequivocal synapomorphies support
Clade A (figure 7.2). Within this, four genera, included
by Hoffstetter (1958) in the tribes Hoplophorini and
Plohophorini, were positioned basally. The successive
arrangements that define the positions of these genera
(Stromaphorus, Hoplophractus, Eosclerocalyptus, and
Pseudoplohophorus) were weakly supported. However,
the monophyly of the two tribes, as well as that of the
subfamily Hoplophorinae, was contradicted by the al-
84
Figure 7.2. Phylogenetic relationships of glyptodonts (modified from Fernicola 2005).

Possible allocations of the genus Urotherium in solid squares.
location of the genera, as was the derived position of

the Hoplophorini Neosclerocalyptus and Plohophorini
Plohophorus. These last two genera, together with Doedicurus, Glyptodon, Panochthus and Urotherium, formed a
monophyletic group. Within this nameless group (Clade
B), none of the recovered groupings had been proposed previously. In particular, the close relationship
between Panochthus and Neosclerocalyptus was quite
unexpected. Clade B had the strongest support in the
analysis, with 6 unequivocal synapomorphies.
Ameghino (1889) viewed panochthines and doedicurines as independently derived groups of plohophorines. Castellanos (1931, 1932, 1940) and Simpson (1945)
partially supported this hypothesis, putting doedicurines and panochthines in the family/subfamily Doedicuridae/nae, while the plohophorines were positioned
in the family/subfamily Hoplophoridae/nae. Hoffstet
ter (1958) and Paula Couto (1979) included panoch
thines in Hoplophoridae/nae, relating them to the plohophorines, while putting doedicurines in a separate
subfamily.
The relationship between plohophorines and doedicurines as originally proposed by Ameghino (1889) was
partially recovered in Fernicola (2005) in that the genus Plohophorus was identified as the sister taxon of the
clade formed by Glyptodon and Doedicurus. However,
a clade of Glyptodon and Doedicurus had not been proposed previously because the Glyptodontinae/dae was
always interpreted as a group with several primitive
characters (e.g., absence of a caudal tube) and as such
derived directly from the Propalaehoplophoridae/nae.
Discussion and conclusions

For many years after the remarkable phylogenetic synthesis of Hoffstetter (1958), the study of glyptodonts
did not include cladistic methodology. This could be

explained by the fact that the taxonomy of these mammals was historically dominated by the external morphology of carapace scutes and the caudal tube. In this
respect, it is interesting to carefully recall the opinion
of Hoffstetter (1958: 572): Le sectionnement en sousfamilles et en tribus est plus dlicat. Beaucoup de
formes sont trs imparfaitement connues, certaines par
une seule plaque. Dans ces conditions, il est illusoire
dessayer dtablir des relations phyltiques et des divisions naturelles . . .
The first phylogenetic study using a cladistic methodology, and carried out at the generic level, was that
of Fernicola (2005). This author questioned virtually all
the previously proposed clusters. As described above,
he found monophyletic groups previously considered
paraphyletic (e.g., Propalaehoplophorinae), while others, formerly considered monophyletic, were solved
otherwise (e.g., Hoplophorini, Plohophorini). Recent
studies (Gaudin and Wible 2006) have reinforced the
monophyly of the clade pampatheres + glyptodonts,
a hypothesis clearly influenced by Ameghinos ideas.
Fernicolas study showed partial, but strong, support
for the monophyly of glyptodonts.
The scarce and dubious cranial record of the
Glyptatelinae prevented them from being phylogenetically assessed by Fernicola (2005). In this chapter, we
consider them as glyptodonts because of their immobile armor, but the presence of free dorsal vertebrae,
85
as in dasypodids, may indicate a basal position within

glyptodonts. However, at present it is not possible to
rule out a closer relationship with the dasypodids.
It is worth pointing out that, just as demonstrated
by Fernicola (2005) for the glyptodonts, the results obtained by Gaudin and Wible (2006) for the dasypodids
show a marked discrepancy between postcranial and
cranial data. Because of this, future phylogenetic analyses should include both sources of characters. Even so,
the data presented in this chapter, and in particular the
strong support for some clades (e.g., Panochthus + Neosclerocalyptus), suggests that traditional phylogenies
should be reconsidered (Fernicola 2005).
Acknowledgments
Alejandro Kramarz and Olga Vaccaro (Museo Argentino de Ciencias Naturales Bernardino Rivadavia, Argentina), Marcelo Reguero and Susana Bargo (Museo
de La Plata, Argentina), and Mario Clara, Daniel Perea,
and Martn Ubilla (Universidad de la Repblica, Uruguay) facilitated access to the specimens that formed
the basis of this study. We are especially indebted to
Marcelo Canevari for the illustrations in figure 7.1. JCF
is also indebted to PEDECIBA and the Universidad
de la Repblica, Uruguay. This is a contribution to the
projects UNLP N474, PIP-CONICET 5240, and PICT
26219.
8
Form, function, and paleobiology in xenarthrans
Sergio F. Vizcano, M. Susana Bargo, and Richard A. Faria
cofagia, necrofagia, alimentacin selectiva o al bulto en

herbvoros y omnivora). Las investigaciones sobre la
relacin entre el rea de la superficie oclusal dentaria
(OSA) y el tamao corporal ha proporcionado interesantes resultados sobre la dieta y metabolismo; los
perezosos milodntidos tienen OSAs mucho menores
que las esperadas para animales de su tamao, lo que
indica bajos metabolismos, mientras que Megatherium
tiene una OSA an mayor que la predecible, indicando
mayor capacidad de procesamiento del alimento en la
cavidad oral, baja capacidad de fermentacin intestinal
y/o mayores requerimientos metablicos. Con respecto
a las interpretaciones paleoecolgicas de las faunas de
Amrica del Sur, han sido abordadas desde varios aspectos, incluyendo los hbitos alimenticios, locomotores y relaciones trficas. Este ltimo estudio propuso
la novedosa hiptesis de que algunos xenartros, previamente considerados herbvoros estrictos, podran
haber tenido una dieta ms amplia, lo que despert un
inters renovado hacia otros aspectos ecolgicos, como
la particin de nichos durante el Pleistoceno, reinterpretaciones de la sistemtica de algunos Carnivora sudamericanos y otras ms.
Resumen
La historia de los estudios de la biologa de los xenartros fsiles es muy rica y se inicia con la descripcin del
perezoso terrestre gigante Megatherium americanum
realizada por Cuvier en 1796. Los xenartros fueron
ejemplos destacados de las colecciones realizadas por
Darwin y de los estudios de Owen. El grupo ya se re
gistra en Amrica del Sur en los comienzos del Terciario,
en sedimentos del Paleoceno, es decir, alrededor de 55
millones de aos antes del presente, convirtindose luego en uno de los elementos ms abundantes y caracte
rsticos de las faunas fsiles de ese continente y llegando
a Amrica Central y del Norte a partir del Pleistoceno
y hasta la actualidad. El registro fsil muestra una gran
diversidad morfolgica, mucho ms acusada que en la
actualidad, que incluye mamferos de tamao corporal
gigantesco y con rasgos muy peculiares, especialmente
los de la megafauna del Pleistoceno de Amrica del
Sur. Debido a su morfologa tan peculiar y a la falta de
anlogos modernos, los xenartros fsiles han incentivado una serie de estudios paleobiolgicos y propuestas
novedosas. En este captulo se resume cmo se ha estimado el tamao corporal en xenartros fsiles, mediante
el uso de modelos a escala y geomtricos generados por
computadora, as como a travs de ecuaciones alomtricas. Algunos estudios biomecnicos contribuyeron a
la interpretacin de la capacidad de soportar flexiones
de los huesos de los miembros de los gliptodontes, la
velocidad de extensin del antebrazo en Megatherium,
el bipedalismo o las habilidades cavadoras de otros
perezosos terrestres. El aparato bucal ha sido estudiado
en una amplia variedad de xenartros y ha permitido
comprender diversas posibilidades adaptativas en cingulados y tardgrados (incluyendo carnivora, mirme-
Resumo
A biologia dos xenartros fsseis mostra uma rica
histria, comeando pela descrio da preguia gigante
Megatherium americanum por Cuvier, em 1796. Os xe
nartros foram exemplos proeminentes nas colees de
Darwin e nos estudos de Owen. Os xenartros aparecem
na Amrica do Sul no comeo do Tercirio, em sedimentos do Paleoceno, isto , cerca de 55 milhes de anos
antes do presente e se tornaram membros muito abun86
dantes e caractersticos das faunas de mamferos neste

continente, com mais de 150 gneros, e alcanaram as
Amricas Central e do Norte desde o Pleistoceno at o
presente. Seu registro fssil mostra uma maior diversidade morfolgica da observada no presente incluindo mamferos de grande tamanho corporal e caracte
rsticas muito peculiares, especialmente na megafauna
pleistocnica da Amrica do Sul. Sendo to peculiares
e faltando os anlogos modernos, os xenartros fsseis
tm estimulado os enfoques paleobiolgicos criativos.
Neste captulo, exploramos como o tamanho corporal
tem sido estimado nos xenartros fsseis, incluindo o uso
de modelos de escala e geomtricos gerados por computador, e tambm equaes alomtricas. Ainda mais,
os estudos biomecnicos realizados em grandes gliptodontes e preguias terrcolas permitiram compreender
a capacidade dos ossos longos das extremidades em
gliptodontes de resistir foras de flexo, extenso do antebrao e velocidade em Megatherium e bipedalismo ou
capacidade de excavar em outras preguias terrcolas.
O aparato mastigador tambm tem sido estimado em
diversos xenartros, o que levou a entender as variadas
possibilidades adaptativas nos cingulados e tardigrados
(incluindo mirmecofagia, necrofagia, carnivoria, herbivoria seletiva e no-seletiva e omnivoria). Entre outros enfoques, a relao entre a rea oclusal (OSA) e a
dieta tem sido pesquisada com resultados interessantes
sobre dieta e metabolismo; a maioria das preguias tm
OSAs menores do que o esperado, o que sugere um
metabolismo baixo, ao passo que M. americanum tem
a OSA esperada para um mamfero do seu tamanho,
ou ainda maior, o que indica um maior processamento
oral do alimento, menor capacidade de fermentao e/
ou um maior metabolismo. Quanto s interpretaes
paleoecolgicas, houve enfoques diferentes no estudo
das faunas do Cenozico sul-americano, incluindo
hbitos alimentares, locomoo e relaes trficas. Este
ltimo tipo de estudos levou idia contra-intuitiva de
que alguns dos xenartros que tinham sido considerados herbvoros estritos possam ter tido uma dieta mais
ampla, e isto, pela sua vez, aumentou o interesse em
outros tpicos ecolgicos, como partio de nichos no
Pleistoceno, reinterpretao da sistemtica de alguns
Carnivora da Amrica do Sul e outros.
Some history
The biology of fossil xenarthrans has deservedly received the attention of many scientists since the description of the giant sloth Megatherium americanum
87
by Cuvier in 1796. This species has played a significant

role in the history and development of vertebrate paleontology and evolutionary science (De Iuliis et al. 2005).
The first specimen was a nearly complete individual
that, as usual for the period, was packaged and shipped
to a European center of study. The remains, considered
marvelous and aberrant, made a great impression on
the scientific community. Its gigantic size and strange
anatomical features defied explanation in the hands of
the specialists. These extraordinary remains arrived
at the Real Gabinete de Historia Natural de Madrid,
where they were mounted and illustrated in a lifelike
position, the first fossil skeleton to be so mounted, and
illustrated by Juan Bautista Bru de Ramn (figure 8.1).
His illustrations reached Cuvier, at the Musum National dHistoire Naturelle in Paris, who described and
named the remains. Cuvier considered the giant sloth
important for several reasons. One of these dealt with
the idea of extinction. Using the powerful new methods
of comparative anatomy, Cuvier clarified phylogenetic
relationships of Megatherium, for it was he who realized
that M. americanum, aberrant as it must have seemed,
was a giant ground sloth and thus its closest living relatives were tree sloths. But Cuvier went further, for he
was able to reconcile the odd anatomy of the gigantic
beast, demonstrating that its parts contributed to a
harmonious whole, in contrast to the opinions of other
naturalists who viewed the anatomy of M. americanum
as proof that it was ill adapted for life and therefore led
a miserable existence. Several interesting exchanges followed and led ultimately in 1831 to a famous Academy
dispute between Cuvier and tienne Geoffroy Saint-Hilaire, primarily about their particular interpretations of
animal form and the effects of function on structure
(Ramrez Rozzi and Podgorny 2001).
Additional large extinct mammals from South
America, including M. americanum and other xenarthrans, were first found by Darwin or made better
known through his efforts while collecting in South
America between 1832 and 1833. As usual, the specimens were sent back to England. Darwin sent them for
study primarily to Richard Owen, who made magnificent descriptions and illustrations of M. americanum
(Owen 1842, 1851, 1856) that became a model for the
presentation of information on fossils. More than 150
years later, this is still the standard text on the anatomy
of the giant sloth. Owen considered giant sloths terrestrial forms, barely able to climb, with their weight resting on the outer margins of the manus and pes. They
were able to raise themselves on their hind limbs, using
88
S. F. Vizcano, M. S. Bargo, and R. A. Faria
Figure 8.1. Early drawing of Megatherium americanum by Juan Bautista Bru de Ramn in 1793, first published by Cuvier in
1796. Modified from Lpez Piero and Glick 1993. Scale bar: 100 cm.
the tail as a support, so that the forelimbs were freed for

other activities such as feeding on the foliage of trees.
Such an image has had a profound influence on the way
the habits of this species have been, and are, conceived.
Indeed, the dominant reconstruction shows the great
beast as an indolent herbivore that needed to bear its
weight on the branch of a tree whose leaves were torn
off with a long tongue. As we will see below, new studies based on functional morphology and biomechanics,
joined with evidence coming from other sources, are
providing us with a very different view.
Diversity of fossil xenarthrans

Xenarthra represents one of the four major clades of
placental mammals (see Delsuc and Douzery this volume). The other three, Afrotheria, Euarchontoglires,
and Laurasiatheria, were previously believed to be a
single clade, Epitheria (McKenna 1975).
As described by Vizcano and Loughry (this volume) xenarthrans have a long history in South America, beginning about 55 mya during the Paleocene and
becoming one of the most prominent and abundant
fossils of South America. During the latest Tertiary and
the Pleistocene, xenarthrans also became abundant in

Central and North America. Today, there are 31 species of living xenarthrans, a diversity that cannot be
regarded as very numerous. They include armadillos
(Cingulata), digging mammals with a complete, flexible armor and diets ranging from omnivory to myrmecophagy (McDonough and Loughry this volume),
the strictly myrmecophagous anteaters (Vermilingua),
which range from fully terrestrial to fully arboreal in
habits (Rodrigues et al. this volume), and the almost
completely arboreal, largely folivorous tree sloths (Tardigrada, also known as Phyllophaga or Folivora; see the
preface for a discussion of this nomenclatural issue; see
Chiarello this volume for more discussion of sloth behavior).
On the other hand, the fossil record of the group
shows a striking morphological diversity. It includes
mammals of very peculiar features and great body size,
particularly in the Pleistocene megafauna. Among them
are large cingulates, including armadillos (Dasypodidae), pampatheres (Pampatheriidae) and glyptodonts
(Glyptodontidae), and tardigrades, with four families
(Mylodontidae, Megalonychidae, Nothrotheriidae,
and Megatheriidae), some of them with body masses
greater than a ton (see below).
Reconstructing paleobiology: form and function

in xenarthrans
Paleobiologists are interested in reconstructing the external form of fossils as living animals and describing
their habitat, ecological role, behavior, and basic biology. As direct observation of behavior or the measurement of physiological variables is impossible in extinct
species, a different way of inferring habits must be devised. One important tool is the principle of actualism,
according to which past events are surmised by analogy with currently observable processes. Mammals are
very diverse in present times and most of them have
fossil forms. Reciprocally, most fossil mammals can be
assigned to extant orders. Perhaps this is the reason paleomammalogists have applied actualism with a more
restrictive criterion than other vertebrate paleontologists, by assuming that fossil species had similar habits
to their current relatives.
However, this methodology is of little use when
phylogenetic affinity is not very close or fossil lineages
possess morphologies not represented in extant species (Vizcano, Faria et al. 2004a). Organisms reflect a
duality between adaptation to environmental requirements and an inherited tendency to remain constant,
that is, the historical contingency of evolutionary constraint. Therefore, if a lineage develops a phenotype
that is adapted to a certain environmental condition, its
species will not necessarily be identical to the current
models. This circumstance is particularly applicable
to those mammalian faunas that evolved in isolation
in South America during a good part of the Tertiary.
For instance, xenarthrans show important dental, skeletal, and muscular peculiarities that make them differ
greatly from other mammals (Vizcano and De Iuliis
2003, and references therein).
In vertebrate paleontology, the main sources of information are fossilized bones and teeth. Therefore, even
though indirect evidence can be used (ichnites or fossil
tracks, taphonomic data, etc.), the majority of the paleobiological information comes from these data. Thus,
the form-function correlation approach (Radinsky
1987) applies. According to this principle, function can
be inferred from form. If appropriate homologies are
lacking, biological analogs must be used; if biological
ones are not available, then mechanical analogs can be
helpful.
Although various features of skeletal anatomy, some
of them very peculiar, characterize xenarthrans (McDonald 2003b) and make them ideal for studies relat-
89
ing form and function, some authors (Hirschfeld 1976;

Webb 1985a; Mendoza et al. 2002; Vizcano and De
Iuliis 2003) have recognized that xenarthrans have not
only historical, but also functional and biomechanical constraints. For instance, the dentition (teeth are
hypselodont and lack enamel, but are composed of
osteodentine, a biomaterial with different structural
properties) probably imposed a severe functional and
biomechanical constraint in subsequent adaptations to
different diets among the various clades.
Several methods can be applied to interpret formfunction relationships: functional morphology, biomechanics, and ecomorphology. Functional morphology
analyzes how form causes, allows, or restricts the functions an organism can perform; biomechanics studies
the relationships between form and function using the
principles of physics and engineering; finally, ecomorphology investigates form and function in regard to environment (Plotnick and Baumiller 2000). All of these
can be complemented by other techniques, such as
morphogeometrical analyses. This particular approach
was developed in the early 1980s (Benson et al. 1982)
and is especially useful for comparing morphologies in
the context of evolutionary transformations (Chapman
1990) and has proved applicable in morphofunctional
interpretations of xenarthrans (Vizcano and Bargo
1998; Vizcano et al. 1998; De Iuliis et al. 2000; Bargo
2001a,b; Vizcano and De Iuliis 2003).
A basic protocol in paleobiological studies implies
the definition of three biological attributes that are essential for each taxon: size, diet, and usage of substratum or type of locomotion. Morphological study of the
masticatory and locomotor apparatuses should permit
predictions about the movements they were designed
for performing. In addition, analyses of mastication are
useful for formulating hypotheses about the diet of the
organism, while analyses of the locomotor apparatus allow inferences about the type of locomotion or preferences in the use of substratum: runner, hopper, digger,
burrower, etc. Obviously, these two aspects, added to
body size, yield much relevant data for the interpretation of an organism in a paleobiological context.
Although actualism is problematic in studies of
some lineages of fossil xenarthrans (e.g., ground sloths
and glyptodonts, figures 8.2 and 8.3), relationships
between form and function in living species are still
needed. It was not until the second part of the 1990s
that some authors began to apply biomechanical, morphogeometrical, and ecomorphological methods to the
study of morphology in living xenarthrans as part of
Figure 8.2. Skeletons of fossil and extant tardigrades. A. Scelidotherium leptocephalum (MLP 3-401)
from Lydekker 1894, Pl. LVI. B. Megatherium americanum (MLP 27-VII-1-1) on display at the exhibition
hall of Museo de La Plata. By permission of Museo de La Plata, La Plata, Argentina. C. Bradypus tridactylus (MACN 4412).
91
Table 8.1. Estimations of Body Mass (Kg) of Some Fossil Xenarthrans

Species

Estimation Technique (reference)

Allometric equation
Geometric model
Glyptodon clavipes
Glyptodon reticulatus
862 (3)
Panochthus tuberculatus
1,061 (3)
1,468 (3)
Doedicurus clavicaudatus
Neosclerocalyptus ornatus
Glossotherium robustum
1,713 (3)
1,200 (4)

1,376a,b (6)
750b (6)
Lestodon armatus
3,397 (3)
3,750 (4)
Mylodon darwini
1,650c (6)
Scelidotherium leptocephalum
1,057 (3)
830 (4)
Megatherium americanum
6,073 (3)
3,950 (5)
Scale model
2,000 (1)
1,100 (1)
1,400 (1)
280 (1)
1,500 (4)
4,100 (4)
850 (4)
3,800 (2)
Sources: (1) Faria 1995, (2) Casinos 1996, (3) Faria et al. 1998, (4) Bargo et al. 2000, (5) Blanco and Czerwonogora 2003, (6) Christiansen and Faria 2003.
a. Mean value of craniodental and limb measurements.
b. Values for the gracile morph of G. robustum, proposed to have been the female (Christiansen and Faria 2003).
c. As mentioned in the text, craniodental measurements yield underestimates of body mass in ground sloths. This value was corrected
from appropriate factors (see Christiansen and Faria 2003 for further discussion).
a major project aimed at understanding the great paleobiological diversity of the South American extinct
forms (but see Naples 1987, 1989). These results allow
the development of novel interpretations of xenarthran
modes of life that, coupled with paleoenvironmental
data (geology, paleoclimatology, and reconstruction of
paleovegetation), provide insightful information on the
paleoecological context in which these animals existed.
Recently Bargo (2003) provided a review of paleobiologic studies in xenarthrans. The following sections
of this chapter constitute an updated summary of our
knowledge, primarily on sloths and cingulates because
the fossil material of vermilinguas is too limited (see
McDonald et al. this volume).
Body size in fossil xenarthrans
Body size has a remarkable influence on an animals life
because it can be correlated, among other features, with
metabolism, limb bone dimensions and biomechanics
of locomotion, or particular solutions for food intake.
Body size in living xenarthrans ranges from 100 g in
the fairy armadillo (Chlamyphorus truncatus Harlan)
to 50 or 60 kg in the giant armadillo (Priodontes maximus Kerr) and giant anteater (Myrmecophaga tridactyla
Linn). Fossil forms range from about 1 kg to almost
6 tons in M. americanum. The oldest xenarthrans recorded during the Paleocene were relatively small armadillos similar in size to the living Patagonian pichi,
Zaedyus pichiy (Desmarest). Like other mammalian
lineages (Bargo 2004), there was a tendency to increase
body size all through the Cenozoic, particularly since

the late Miocene and Pliocene. Within the Pleistocene
fauna, the abundance and diversity of large-sized forms
(armadillos, pampatheres, glyptodonts, and ground
sloths) represent a fertile soil for biomechanical studies.
This section is a brief overview of the approaches
used in estimating body size in xenarthrans. The results
obtained are then related to the remarkable adaptations
of the locomotor and masticatory apparatuses, as well
as to the paleoecology of xenarthrans. These issues will
be explored in the following sections, where a complex,
interrelated landscape of conclusions will be drawn.
Three different methods have been used to estimate
body mass in fossil xenarthrans: scale and computergenerated (geometric) models, and allometric equations (table 8.1). Scale models were applied by Faria
(1995) in some large Pleistocene species of Glyptodontidae and, coupled with computer-generated models, by
Bargo et al. (2000) in Pleistocene mylodontid ground
sloths. The same geometric model was used by Blanco
and Czerwonogora (2003) for assessing body mass in
the Pleistocene megatheriid M. americanum. Casinos
(1996) estimated the body mass of M. americanum by
means of a scale model and allometric equations. Faria
et al. (1998) estimated the body masses of 7 species of
xenarthrans (3 glyptodonts and 4 ground sloths) and
7 other large placental species, using a set of allometric equations previously developed from craniodental
(Damuth 1990; Janis 1990) and limb bone dimensions
92
(Scott 1990) in modern mammals. Christiansen and

Faria (2003) also calculated body masses of three
specimens of mylodontid ground sloths using a geometric model and allometric equations.
However, estimation itself is not without difficulties.
For example, Faria et al. (1998) demonstrated that dental measurements are not appropriate for xenarthrans
because homologies within the group are unclear, and
many xenarthrans have significantly reduced and/or
homodont dentitions, producing clear underestimates
of body size. Some postcranial measurements, such
as femoral transverse diameter in ground sloths, had
to be discarded too because they produced ridiculous
overestimates of one order of magnitude. As arithmetic
means yielded larger estimates, Faria et al. (1998) suggested that this statistic should be used to estimate body
size only in studies in which larger size is the conservative hypothesis.
Limbs and function
In the last decade limb proportions have been used to
interpret adaptation in fossil xenarthrans. Biomechanical studies performed on large glyptodonts and ground
sloths provided insight into the capacity of the limb
bones to withstand bending forces, forearm extension
and velocity, and bipedalism or digging abilities.
Within cingulates, Faria (1995) analyzed limb bone
strength and locomotor habits in some glyptodonts, using a strength indicator developed by Alexander (1983,
1985). The strength of a bone is obtained from its crosssectional area and is inversely related to the bones
length and to the fraction of the weight of the animal
that is supported by the corresponding pair of limbs.
Hence, as a first step, Faria estimated body masses
and centers of mass, and then calculated the distribution of weight between the anterior and posterior limb
pairs. The femur strength indicators of large Pleistocene
glyptodonts were equivalent to those of large living
mammals capable of galloping (i.e., buffalos and rhinos), but values of the humerus were similar to those of
elephants, which cannot gallop. Although the anatomy
of their long bones is not consistent with such activity (see figure 8.3A), the muscular insertions suggest
that glyptodonts were able to adopt bipedal postures to
perform strenuous activities, such as the intraspecific
fights proposed by Faria (1995). Consistent with this
hypothesis, Alexander et al. (1999) estimated the energy that the tail muscles could have supplied for a blow
and the energy required to fracture the carapace in a
large glyptodont. They interpreted observed damage in

carapaces as resulting from fights.
Studies of living armadillos provide the larger sample sizes necessary for statistical and comparative studies on the correlations between limb proportions, body
size, and behavior. Faria and Vizcano (1997) found
that armadillos scale their humeri, ulna, and, partially,
tibia like other digging mammals and their femora as
generalized mammals. Efforts to identify limb variables that correlate with fossorial ability demonstrated
that forelimb indices correlate well with digging habits (particularly the development of the olecranon, see
figure 8.3B, Vizcano et al. 1999). Hindlimb indices do
not, but seem more related to weight bearing (Vizcano
and Milne 2002). These results have been applied to
the study of different extinct armadillos. For example,
Vizcano and Faria (1997) used olecranon length as indicative of fossorial ability for the previously proposed
cursorial Miocene armadillo Peltephilus Ameghino. All
limb indices were applied to the Pleistocene form Eutatus seguini Gervais, which led to the conclusion that
its fossorial ability was comparable to that of living euphractines (Vizcano, Milne, and Bargo 2003).
White (1997) investigated the influence of a large set
of morphological characters on locomotor behavior of
early sloths. She found that most of the Santacrucian
(early/middle Miocene) sloths were semiarboreal; the
smaller forms must have been mostly climbers, while
the largest were more terrestrial. This study provides
useful evidence for the interpretation of adaptations in
the larger Pleistocene ground sloths.
The pes of tardigrades, and particularly of Megatheriinae, is very different morphofunctionally from that
of other mammals and merits distinctive terminology,
including: rotated, pedolateral (McDonald 1977,
Webb 1989), and traviportal (both quadrupedal and
bipedal stance with obligate slow moving habits, Toledo
1996) postures. Less attention has been paid to the manus, despite the presence of equally anomalous morphofunctional features (Tito and De Iuliis 2003). For
the ground sloths, different specializations may have
derived from a primitive quadruped way of locomotion in both main lineages of Pleistocene forms, megatheriids and mylodontids. The giant sloth M. americanum has been formally proposed as bipedal based on
ichnologic and biomechanic evidence (Aramayo and
Manera de Bianco 1996; Casinos 1996; Blanco and Czerwonogora 2003; see figure 8.2B). The latter includes
analyses of body size, speed, Froude number, indicator
93
Figure 8.3. Skeletons of fossil and extant cingulates. A. Glyptodon clavipes (MLP 16-40) from Lydekker 1894,
Pl. V; by permission of the Museo de La Plata, La Plata, Argentina. B. Tolypeutes matacus (MLP 1217).
of athletic ability, and bending and resistance moments

of the vertebral column, as well as a complete geometric
and biomechanical analysis of the footprints assigned
to this species found in Pehun-C, Buenos Aires Province, Argentina. Bipedalism also implies that the forelimb could have been free to perform activities other
than locomotion. Faria and Blanco (1996) tested the
possibility that the forearms of M. americanum were
designed for optimizing speed rather than strength of
extension, and concluded that such a trait must have
been associated with a potentially aggressive use of the

animals large claws.
Tito and De Iuliis (2003) provided an analysis of
the manus of the tropical ground sloth Eremotherium
laurillardi (Lund) that lends support to the traditional
view that it was normally inverted during locomotion,
with weight borne mainly on the lateral margin of digit
V. The manus of E. laurillardi was unsuited for digging for reasons intrinsic to the manus itself, as well
as those associated with the structure of the forelimb,
94
which is different from the forelimbs of other typical

digging animals. In general, the manus was a fairly rigid
structure. Movement among its elements, except for the
ungual phalanges, was severely restricted. Its structure
and form seem ill-suited for use during locomotion,
except perhaps for occasional use in uneven terrain;
otherwise, we would expect only slow and awkward
movement. Similarly, the forelimb was probably not
used for manipulating objects, even though it is marked
by adaptations allowing a high degree of pronation and
supination. More likely, the forearm, capable of striking severe blows, was effective in defense. In addition
to being able to resist stresses along the mediolateral
plane, the form of the forelimb seems suited to securing
overhead branches with the palms facing medially.
Most of the approaches mentioned above were also
applied to mylodontid ground sloths. Bargo et al. (2000)
analyzed limb proportions and resistance to bending
forces in Scelidotherium leptocephalum Owen, Glossotherium robustum Owen and Lestodon armatus Gervais
to infer their locomotor adaptations. Allometric equations were calculated from the long bones and used to
predict linear dimensions from body mass; the strength
indicator was estimated for humeri and femora. This
analysis, coupled with the anatomical evidence, indicated that S. leptocephalum and G. robustum were well
adapted for strenuous activities in which force was
enhanced over velocity, such as digging (figure 8.2A).
Based on this work, Vizcano et al. (2001) considered
these taxa as possible builders of large Pleistocene
burrows present in the Pampean region of Argentina.
These tunnels, discovered over the past several decades,
were interpreted as paleoburrows built by fossil mammals (Zrate et al. 1998). The diameters of the burrows
(ranging from 80 to 180 cm) are consistent with the
diameters of the sloths, and the tracks found on the
roof and walls of the tunnels coincide with the form
of the hand skeleton. Although limb proportions of L.
armatus are comparable to those of the others, its low
strength indicator suggests that its limbs were not well
designed to perform such demanding activity.
Semi-aquatic habits have been proposed for some
nothrotheriid ground sloths (Thalassocnus) from late
Miocene/late Pliocene marine deposits of Peru (McDonald and Muizon 1995; Muizon and McDonald
1995). Thalassocnus probably entered the near shore
waters (as suggested by the morphology of the femur
and patella), swam by paddling of the hind limbs (according to the morphology of the femur head and tibia
length), and moved the tail dorsoventrally (as inferred
from the morphology of the caudal vertebrae) to help

maintain a head-down feeding position.
The feeding apparatus
As mentioned above, functional morphology and biomechanics were applied to the study of feeding in a
wide range of xenarthrans. Although Smith and Redford (1990) did not continue their investigations on the
mechanics of the masticatory apparatus in living armadillos, several studies of fossil armadillos and ground
sloths were conducted by others using modifications of
their method, which involved calculating the moment
arms of the masticatory musculature to infer relative
bite forces and velocity (Vizcano and Bargo 1998; Vizcano et al. 1998; De Iuliis et al. 2000; Bargo 2001a,b;
Vizcano and De Iuliis 2003).
Within cingulates, the great diversity of forms suggests a number of adaptive possibilities, ranging from
specialized myrmecophagous species to carrion eaters
or predators among the animal eaters; selective browsers to bulk grazers among herbivores; and some omnivores as well (Vizcano, Faria et al. 2004a). Whereas
armadillos developed varied habits (mostly animalivorous but also including omnivores and herbivores),
pampatheres and glyptodonts were herbivores (figure
8.4). Morphofunctional and biomechanical studies
permitted a reevaluation of previous hypotheses based
solely on comparative morphology. As a consequence,
specialized carnivory in peltephiline armadillos was
refuted (Vizcano and Faria 1997), but corroborated
in armadillos of the genus Macroeuphractus Ameghino
(Vizcano and De Iuliis 2003). Plant-eating was confirmed in eutatines (Vizcano and Bargo 1998), pampatheres (Vizcano et al. 1998; De Iuliis et al. 2000) and
glyptodonts (Faria and Vizcano 2001), although different kinds of herbivory may have occurred in each
group (Vizcano and Bargo 1998; Vizcano et al. 1998;
De Iuliis et al. 2000). Vizcano, Faria et al. (2004a)
proposed that such morphological and adaptive diversity suggests a more extensive cladogenesis than that
reflected by current systematic schemes. These analyses
also revealed that some cingulates evolved mechanical
solutions that are not shared by closely related taxa,
and do not have current analogs that can be used as
models to investigate and interpret adaptations of lineages without living representatives. For instance, the
masticatory apparatus in glyptodonts underwent a telescoping process that placed it well below the cranium
(Faria 1985, 1988), creating problems in the way that
stresses produced by mastication were absorbed by the
95
Figure 8.4. Skulls and mandibles of various cingulates. A. Eutatus seguini (MLP 69-VIII-1-5). B. Macroeuphractus outesi (MLP 69-IX-9-3). C. Stegotherium tessellatum (YUPM-PU 15565). D. Peltephilus
pumilus (MACN 7784). E. Vassallia maxima (FMNH P14424). F. Doedicurus clavicaudatus (MLP 16-24).
By permission of Asociacin Paleontolgica Argentina, Buenos Aires. Scale bar: 10 cm.
mandible and implying unusual jaw mechanics (Faria

and Vizcano 2001).
Using information obtained from living sloths, Naples (1987, 1989) analyzed in detail the morphology and
function of the feeding apparatus of two fossil North
American ground sloths, proposing generalized anteromedially directed masticatory movements for both.
Bargo (2001a,b) studied the masticatory apparatus of
the large South American Pleistocene ground sloths
(G. robustum, L. armatus, Mylodon darwini Owen, S.
leptocephalum, and M. americanum) with extensive
morphological, biomechanical, morphogeometrical,
and ecomorphological methods (figure 8.5). Jaw mechanics, morphogeometric analyses, and the correlation between craniodental variables (hypsodonty,
dental occlusal surface area, and relative width of the
muzzle) and diet all suggested probable niche differentiation among ground sloths based on dietary categories. While the masticatory pattern of mylodontids
is rather generalized with a clear anteromedial powerstroke, as previously proposed by Naples (1987, 1989),
M. americanum was well adapted for strong, mainly
vertical biting. This information, in addition to tooth
shape, suggests that teeth were used mainly for cutting,
rather than grinding, and that fibrous food was not the
main dietary component.
Muizon et al. (2004b) analyzed the masticatory apparatus of the five species of aquatic sloths, Thalassocnus, from Peru. They showed different feeding adaptations: the three oldest species were probably partial
grazers, feeding on stranded seaweeds or sea grasses or
in very shallow waters, while the younger were more
Figure 8.5. Skulls and mandibles of various tardigrades. A. Glossotherium robustum (MLP 3-137). B.
Mylodon darwini (skull MLP 3-764; mandible MACN 991). C. Lestodon armatus (MLP 3-29). D. Scelidotherium leptocephalum (MMP 458-S). E. Megatherium americanum (MLP 2-64). Scale bar: 10 cm.
specialized grazers and probably fed exclusively in the

water and at greater depth.
Some other specific traits of the feeding apparatus are
also being considered. One of them is hypsodonty, the
relative increase in crown height of a tooth, which has
been traditionally viewed as a response to dietary shifts
toward abrasive vegetation. But recent work indicates
that evolution of hypsodonty is also due to the higher
prevalence of grit and dust in more open environments
(Janis 1988, 1995; Janis and Fortelius 1988). The teeth
of sloths are both high crowned and open rooted, that
is, hypselodont. A comparative analysis performed in
eleven species of Pleistocene sloths by Bargo, De Iuliis,
and Vizcano (2006) suggests that differences in hypsodonty may be explained by diet, habitat, and behavior.
Among mylodontids, morphologic and biomechanical analyses indicate that hypsodonty was unlikely to
be due solely to dietary preferences such as grazing.
As mentioned above, some mylodontids (e.g., S. leptocephalum, L. armatus, G. robustum, M. darwini) were
capable diggers that likely dug for food, and ingestion
of abrasive soil particles probably played a considerable
role in shaping their dental characteristics. Increased
hypsodonty over time in Paramylodon harlani Owen in
North America, however, is apparently due to a change
in habitat from closed to more open environments (McDonald 1995). Geographical distributions of the megatheriids Eremotherium Spillmann and Megatherium
indicate differing habitats as possible factors in hypsodonty differences. In summary, although Tardigrada
hypsodonty is apparently affected by diet, habitat and
behavior, as well as the absence of enamel (which must
be responsible for much of the hypsodonty observed in
xenarthrans), it is difficult to identify the contribution
of each factor.
Fossil xenarthrans are also interesting for the different degrees of lobation in their cheek teeth. Vizcano et
al. (2006) investigated the relationship between dental
occlusal surface area (OSA) and diet, and other physiological factors in fossil xenarthrans. For most fossil
xenarthrans OSA is smaller than expected compared
to extant herbivorous mammals of equivalent body
size. Within xenarthrans, cingulates show the highest
OSA values, suggesting more extensive oral food processing than in tardigrades. Among ground sloths, mylodontids have extremely low OSA values, suggesting
low efficiency in oral food processing that was probably
compensated for by high fermentation in the digestive
tract, and/or lower metabolic requirements. On the
97
other hand, M. americanum has an OSA expected for,

or even higher than, a mammal of its size, which indicates higher oral food processing, lower fermentation
capacity, and/or higher metabolic requirements.
Other features besides teeth are important in reconstructing feeding preferences and behaviors. For
instance, muzzle shape and facial musculature are important features to consider in relation to food intake, a
problem that seems to be emphasized in sloths (Bargo,
Toledo, and Vizcano 2006) because of the edentulous
nature of the muzzle and its varied morphology. The relationship between dietary habits and shape and width
of the muzzle is well documented in living herbivores,
particularly in ungulates (see Janis and Ehrhardt 1988
and references therein), and has been considered an important feature for the inference of alimentary styles in
fossils (Solounias et al. 1988; Solounias and Moelleken
1993a,b; Janis 1995). Bargo (2001a,b) and Bargo, Toledo,
and Vizcano (2006) used this interesting methodological tool to develop models of food intake in five species
of South American Pleistocene giant ground sloths by
reconstructing the appearance and shape of the muzzle.
Using features of the skull such as the areas of origin
and insertion of the musculature, and the patterns of
musculature in modern tree sloths and some living
ungulates, they reconstructed the nasal cartilages and
facial muscles involved in food intake. The preservation
of the nasal septum in some fossils, and the scars for
muscular attachment in the rostral part of the skulls,
permit a conservative reconstruction of muzzle anatomy in fossil sloths. Ground sloths with wide muzzles
(G. robustum and L. armatus; figure 8.5A,C) had a
square, nonprehensile upper lip and were mostly bulkfeeders. The lips, coupled with the tongue, were used
to pull out grass and herbaceous plants. Sloths with
narrow muzzles (M. darwini, S. leptocephalum, and M.
americanum; figure 8.5B,D,E) had a cone-shaped and
prehensile lip and were mixed or selective feeders. The
prehensile lip was used to select particular plants or
plant parts.
Characteristics of the tongue may also be important for food intake and intraoral processing, a feature
underlined in the aforementioned reconstructions of
Megatherium tearing off leaves with its purported long
tongue. Recently, Tito and De Iuliis (2003) proposed
that E. laurillardi used its hands to bring overhead
branches toward a powerful prehensile tongue that
could strip the leaves from the branch. They also stated
that it is not unreasonable to presume similar feeding
98
as well as locomotory habits for Megatherium, given the

strong similarities in skeletal structures between the
hands of these closely related genera.
Of course, these interpretations are based on limited samples because preservation of tongue muscle is
extremely rare. However, it is possible to recover information from the hyoid complex, its bony attachment.
Recently, functional studies on the hyoid apparatus
have been performed in both living and fossil species
of xenarthrans (Naples 1986, 1999; Prez et al. 2000;
Prez 2001). Although fossil preservation of the hyoid
complex is itself unusual, studies in progress describe
some fine specimens of Paramylodon, Scelidotherium,
Glossotherium, Megatherium, and Glyptodon.
The hyoid apparatus of Glyptodon shows a marked
fusion of the stylo-, epi-, and ceratohyoid bones of each
side into a vertical rod, which is notable also for its general robustness and large areas for muscular insertion.
This modification apparently occurred in response to
the migration of the masticatory apparatus below the
neurocranium that occurred in the evolution of glyptodonts, as described by Faria (1985, 1988). While other
muscles are reduced, the geniohyoideus and ceratohyoideus acted antagonistically, producing powerful
movements of protraction and retraction of the tongue,
respectively. However, as the geni process is quite far
back, the tongue would not have protruded a long distance. Hence, it would have been more important in
the reworking of food in the oral cavity than in food
intake.
Within Tardigrada, Glossotherium has mobile joints.
The relationship of the hyoid bones to the skull suggests
that these animals were able to make agile movements
with their tongues. Scelidotherium is somewhat different in that the protrusion capability of the tongue is
interpreted as unimportant. In Megatherium, the parts
of the hyoid show limited ability of movement among
themselves or for the positioning and stretching adopted as a whole, following modification in the height
of the skull. These features indicate that the classical,
pervasive image of the animal tearing off leaves and
branches and driving them into its mouth with a long
prehensile tongue should be reconsidered. As a matter
of fact, current morphofunctional analyses contradict
the notion that Megatherium was even capable of sticking out its tongue.
Paleoecological interpretation
As described in a recent review (Bargo 2003), studies
on body size and locomotor and feeding apparatuses
are among the most relevant data for the interpretation
of the main aspects of the biology of a fossil mammal
(paleoautoecology) and its classification in a paleoecological framework (paleosinecology).
There have been several paleoecological approaches
to the study of Cenozoic South American faunas, including xenarthrans (Faria 1996; Kay and Madden
1997; Tauber 1997; Croft 2001, Vizcano, Faria et al.
2004b). Most examined body mass and feeding habits;
locomotion was usually considered only qualitatively
or ignored. Faria (1996) analyzed the trophic relationships between the South American Lujanian and North
American Rancholabrean (both ages being late Pleistocene/early Holocene) megamammals from the point
of view of body mass, and the ecological implications
of these body sizes. According to the author, the fauna
contained significant diversity of large herbivores, but
did not contain a proportionally diverse suite of large
carnivores. In addition, he suggested that the coexistence of so many large herbivores in a poor environment
led to strong competition for resources. He concluded
that some of the mammals previously considered strict
herbivores might have been carnivores to some degree,
and proposed that ground sloths were opportunistic
carrion eaters.
This challenging view promoted new lines of research. For instance, Vizcano (2000) briefly analyzed
plant resource exploitation among sympatric Lujanian
herbivorous armored xenarthrans (Cingulates). The
morphofunctional and biomechanical studies discussed
above revealed that the main dietary difference among
these cingulates was the coarseness of the vegetation
they were capable of processing. On that basis, Vizcano
(2000) concluded that the coexistence of many herbivorous cingulates could have been sustained through differential exploitation of resources: eutatine armadillos
were mainly browsers, and the larger pampatheres and
glyptodonts represented increasing degrees of grazing habits. Vizcano (2000) has suggested that similar
niche differentiation may apply to other herbivores as
well. Bargo (2001a) and Bargo, Toledo, and Vizcano
(2006) proposed a niche differentiation among Lujanian ground sloths. G. robustum and L. armatus were
bulk feeders, whereas M. darwini was a mixed feeder.
S. leptocephalum was a selective feeder that dug using
muzzle and claws to search for food (roots and tubers)
or browsed on shrubs and grasses at ground level. M.

americanum was probably a generalized selective feeder;
that is, it fed on a mixture of leaves, shrubs, and fruits,
although it may have been more selective, and capable
of consuming flesh as an alternative energy source (see
above). In all these analyses, in addition to diet, the use
of substrate should be considered in order to provide a
more complete paleoecological interpretation.
The notion that the review of hypotheses in the light
of new evidence is a healthy practice in science applies
to these lines of study. For instance, a recent revision of
the ursids from South America proposed that during
the late Pleistocene there were three species of bears
(Soibelzon 2004), instead of one as proposed by Faria
(1996). These species may have acted as large scavengers, which was probably true for other carnivores
such as felids and canids as well. If true, this forces reexamination of Farias estimates of trophic diversity, at
least to a certain extent. In fact, a recent analysis of the
paleoecology of the large carnivore guild from the late
Pleistocene of Argentina suggests that if high herbivore
biomass occurred during the Lujanian, it could have
supported a higher density of carnivores (Prevosti and
Vizcano 2006). As another example, the identification
of sexual dimorphism (see Christiansen and Faria
2003), might diminish the number of herbivore species. Finally, the proposed very low metabolism of the
mylodontids (Vizcano et al. 2006) suggests that they
were probably not as abundant nor required as much
food as originally calculated.
Final remarks
Paleobiology reaches its highest degree of interest when
it deals with the challenging tasks of interpreting fossil
forms having unusual and bizarre morphologies (Hickman 1988), especially those that lack modern analogs.
As presented in this chapter, the use of different methodological tools (biomechanics, morphogeometry, ecomorphology) allows a better understanding of the biology of the fossil xenarthrans. Indeed, through the past
two centuries, fossil xenarthrans have intrigued many
scientists who applied the modern points of view of the
day to understand their peculiar morphology.
In recent years, xenarthran paleoecology has been
based on the estimation of body size, which was rather
impressive in many cases, as in the giant, elephant-sized
ground sloths and in many glyptodonts whose body
99
masses exceeded one ton. Such results have allowed

the interpretation of autecological, synecological, and
morphofunctional aspects of the life history of many
members of the group.
Further, limb dimensions have been used to assess
locomotor habits, and other biomechanical models
yielded results important for inferring mastication and
hence diet and metabolism. Other biological traits remain to be addressed. For instance, based on energetic
considerations, Faria (2002) concluded for M. americanum that the hide of this giant sloth, like that of modern mammals of similar size, must have been hairless,
which represents a significant change in the appearance
of the most frequent reconstructions.
These studies demonstrate that extinct xenarthrans
were peculiar mammals, not only in their anatomy, but
also in their diverse alimentary habits and, in some
cases, even in the arrangement of their jaws and in the
possible usage of the hands for food processing.
Science is, however, an endless game, and paleontology is no exception. Other kinds of evidence will appear,
such as biochemical, palynological, and sedimentological. They will augment our current reconstructions and
eventually modify our points of view about the behavior of this striking fossil fauna.
Acknowledgments
The authors are grateful to Olga Vaccaro (Mammalogy Department) of the Museo Argentino de Ciencias
Naturales de Buenos Aires, who kindly provided the
skeleton of Bradypus, to Jim Loughry for reading critically an early draft, and to the reviewers Ross MacPhee
and Darin Croft for their valuable comments. This is a
contribution to the projects of SFV from Universidad
Nacional de La Plata N474, PIP-CONICET 5240, PICT
26219.
Note
Abbreviations used in figures: FMNH = Field Museum of Natural History, Chicago; MACN = Museo
Argentino de Ciencias Naturales Bernandino Rivadavia, Buenos Aires; MLP = Museo de La Plata; MMP
= Museo Municipal de Ciencias Naturales de Mar del
Plata; YUPM-PU = Yale Peabody Museum, Princeton
University collection, New Haven.
3
Living Xenarthra
Physiology and Genetics
9
Armadillos and dimorphic pathogenic fungi
Ecological and evolutionary aspects
Eduardo Bagagli and Sandra de Moraes Gimenes Bosco
presenta una estrategia de bajo costo para monitorear

diversas importantes enfermedades infecciosas de potencial zoontico en humanos.
Resumen
El armadillo de nueve bandas Dasypus novemcinctus ha
mostrado ser importante en el estudio de algunos patgenos dimrficos pertenecientes al Orden Onygenales,
especialmente al Paracoccidioides brasiliensis, agente
etiolgico de la paracoccidioidomicosis (PCM), la ms
importante micosis sistmica de Amrica Latina. La
PCM es ms frecuentemente observada en trabajadores
de zonas rurales pobres, el tratamiento es difcil y prolongado, las recidivas son frecuentes y puede llevar a
la muerte o dejar secuelas incapacitantes, provocando,
de ese modo, elevado impacto social en estas comunidades. Medidas preventivas contra nuevas infecciones
son inaplicables debido a que el nicho ecolgico del
patgeno an permanece desconocido, pues el hongo
es slo raramente aislado del ambiente, la enfermedad
tiene un prolongado perodo de latencia y sin brotes
epidmicos y, adems, ha sido muy poco estudiada en
los animales domsticos o silvestres. La observacin de
que el armadillo de nueve bandas se presenta naturalmente infectado por P. brasiliensis en reas endmicas
trajo nuevas posibilidades para los estudios eco-epidemiolgicos y tambin para el mejor conocimiento de
su interaccin patgeno-hospedador. Los armadillos
se han mostrado tiles como animales centinela para
delimitar reas donde el P. brasiliensis ocurre ms frecuentemente en la naturaleza, como en matas riparias
hmedas y modificadas debido a la accin humana.
El frecuente aislamiento del hongo en animales aparentemente sin sntomas de la enfermedad indica que
el patgeno tiene una biologa evolutiva en asociacin
a los hospedadores animales y presentan adaptaciones
marcadas tanto para su vida saproftica como para la
parasitaria. El estudio de patgenos en armadillos re
Resumo
O tatu de nove bandas Dasypus novemcinctus mostrou
ser particularmente importante no estudo de alguns
fungos patognicos dimrficos pertencentes ordem
Onygenales, em especial o Paracoccidiodes brasiliensis, causador da paracoccidioidomicose (PCM), a mais
importante micose sistmica da Amrica Latina. A
PCM acomete principalmente trabalhadores rurais pobres, o tratamento difcil e prolongado, recidivas so
freqentes, e podem levar ao bito ou deixar seqelas
incapacitantes, causando assim elevado impacto social
nestas comunidades. A adoo de medidas preventivas
contra novas infeces prejudicada uma vez que o
nicho ecolgico do patgeno ainda pouco conhecido,
pois o fungo difcil de ser isolado a partir de amostras
ambientais, a doena tem prolongado perodo de latncia e sem surtos epidmicos, e foi pouco pesquisada em
animais domsticos ou silvestres. A constatao de que
o tatu de nove bandas apresenta-se naturalmente infectado pelo P. brasiliensis nas reas endmicas abriu
novas possibilidades para estudos eco-epidemiolgicos
e tambm para uma melhor compreenso da sua interao patgeno-hospedeiro. Os tatus vm se mostrando
teis como animais sentinelas para demarcar reas
onde o P. brasiliensis ocorre mais freqentemente na
natureza, como em matas riprias midas e alteradas
pela ao humana. O freqente isolamento do fungo de
animais aparentemente sem sintomas da doena sugere
que este patgeno tenha uma biologia evolutiva associada a hospedeiros animais, apresentando caractersti103
104
E. Bagagli and S. M. G. Bosco
cas bem adaptadas, tanto para a vida saproftica como a

parastica. O estudo de patgenos em tatus representa
uma eficiente estratgia de baixo custo para monitorar
vrias importantes doenas infecciosas de potencial
zoontico.
Introduction
Armadillos are unique mammals in terms of pathogen
coexistence because their physiological and ecological
characteristics contribute to making them appropriate
hosts. Physiologically, they have lower body temperatures than most mammals and a weak immune system;
ecologically, they live literally immersed in soil and
organic matter, mainly in tropical and subtropical regions, under biotic and abiotic conditions that promote
multiple encounters with a diverse group of pathogens
and vectors (Taber 1945; Talmage and Buchanan 1954;
Storrs et al. 1974; Ulrich et al. 1976). Several different
pathogens that produce important zoonotic diseases
have been found frequently associated with armadillos and other xenarthrans (Acha and Szyfres 2003). For
example, armadillos are considered reservoirs for the
protozoan etiological agents of leishmaniasis (Lainson
et al. 1979; Lainson and Shaw 1989), and trypanosomiasis, or Chagas disease (Paige et al. 2002; Yeo et al. 2005).
Detection of these hemoflagellates was also reported in
the sloths Choloepus didactylus and Bradypus variegatus
(Lainson et al. 1981; Zeledon et al. 1975, respectively),
and the anteater Tamandua tetradactyla (Lainson et al.
1981). Additionally, the protozoan Toxoplasma gondii
was detected in the armadillos Dasypus novemcinctus
and Euphractus sexcinctus (Burridge et al. 1979; Silva
et al. 2006).
Other pathogens also occur. Concerning bacteria,
many efforts have concentrated on nine-banded armadillos, D. novemcinctus, and Mycobacterium leprae,
a subject thoroughly reviewed by Truman (this volume). Besides M. leprae, other important bacteria such
as Leptospira interrogans (serotype hardjo), Nocardia
brasiliensis, Escherichia coli, Salmonella sp., Enterobacter aerogenes, Acinetobacter hinshawii, and Streptococcus sp. have been isolated from armadillos (Gezuele
1972; Meyers et al. 1977; Lins and Lopes 1984; Diniz
et al. 1997). In their study, Diniz et al. (1997) also recovered parasitic nematodes, such as Ancylostoma sp.,
Strongyloides sp., Ascaris sp., and Trichuris sp., from
the gastrointestinal tracts of their animals. Viral agents
such as the causative agent of pox disease were detected
in two giant anteaters during an outbreak in a zoo. Both
animals had the severe form of the disease, which led

to death, and the zoonotic transmission of the virus
to a female attendant at the zoo was also documented
(Marennikova et al. 1977). Rabies virus has been reported in a nine-banded armadillo captured in an urban area in Texas, U.S.A. (Leffingwell and Neill 1989).
The aim of this chapter is not to review all the pathogens that have been found in armadillos or other xenarthrans, but to focus on the dimorphic pathogenic fungi,
especially Paracoccidioides brasiliensis. This particular
fungus is important because it causes life-threatening
fungal infections in humans. In animals, P. brasiliensis
has been found primarily in the nine-banded armadillo,
D. novemcinctus. We explore some possible biological
implications of the close association between armadillos and these fungi, and the use of armadillos as sentinel animals or appropriate tools to help solve some
eco-epidemiological questions about these pathogens.
Pathogenic fungi
The kingdom Fungi encompasses a great number of
species, estimated at more than 1.5 million, but the
vertebrate pathogenic groups are restricted to just a
few orders and families, probably as the result of evolutionary processes leading to adaptation for infection
(Bowman et al. 1992, 1996; Taylor 1995). Although most
pathogenic fungi have been frequently documented as
being associated with wild and domestic animal hosts,
the biological significance of these associations has not
been studied in detail, nor have they been exploited as
a means for monitoring these agents in nature.
Pathogenic fungi, with few exceptions, have both
saprobic and parasitic phases; some authors and public health authorities have considered the two phases
as disconnected steps. According to the famous medical mycologist John Willard Rippon, the infective form
of the fungus should be considered a blind alley because it is not contagious to other hosts or generally
of use in dissemination of the species (Rippon 1988).
Why would a free-living fungus without an obligate
host evolve virulence traits, such as morphological and
physiological features, to permit infection in animals?
Casadevall et al. (2003; see also Casadevall 2005) proposed the concept of ready-made virulence and dual
use of virulence factors. According to them, fungal
virulence may originate from selection imposed by organisms, such as amoeboid and nematode predators,
that phagocytize or feed on fungi in nature. Virulence
factors would be maintained because they offer dual
use for survival both in the external environment and

inside the host. This concept was applied mainly to species that have intracellular growth, such as Cryptococcus
neoformans and Histoplasma capsulatum (Casadevall
et al. 2003; Woods 2003; Casadevall 2005). Although it
emphasizes an evolutionary process for fungal pathogen origins, it does not attribute any significant role to
animal hosts as influences on this process.
The order Onygenales (Phylum Ascomycota) contains a disproportionate number of pathogens (Currah
1985; Untereiner et al. 2004). Here we find the dermatophytic fungi (Onygenales, Arthrodermataceae), which
are some of the most prevalent infectious agents of
humans and other animals throughout the world, and
which developed the ability to degrade keratin, a highly
insoluble protein in skin, nails, and hair. This ability is
a key innovation in the evolution of animal dermatopathogenicity (Leclerc et al. 1994; Scott and Untereiner
2004), and may have appeared in association with burrowing and denning animals (Weitzman and Summerbell 1995; Kaszubiak et al. 2004).
Dimorphic fungi such as Blastomyces dermatitidis,
Coccidioides immitis, Histoplasma capsulatum, and
Paracoccidioides brasiliensis are also referred to as true
fungal pathogens because they cause systemic or deepseated mycosis even in immunocompetent hosts. This
fungal group (Onygenales, Onygenaceae sensu lato)
presents several common mycological features, for example, dimorphism (filamentous and yeast forms in
the saprobic and pathogenic conditions, respectively),
arthroconidia (the airborne infective propagules), localized geographic distribution, and affinity for animal
wastes such as feces and uric acid (Currah 1985; Untereiner et al. 2004).
Molecular taxonomy originally placed P. brasiliensis
in the family Onygenaceae (Bialek et al. 2000; San-Blas
et al. 2000; Herr et al. 2001). More recently, a clade distinct from Onygenaceae sensu lato was proposed as a
new family (Onygenales, Ajellomycetaceae) to encompass the monophyletic group Ajellomyces (including
the anamorph genera Blastomyces, Emmonsia, and Histoplasma) and P. brasiliensis, because both characteristically associate with vertebrate hosts (Untereiner et al.
2004).
Molecular phylogenetic and phylogeographic studies have indicated the Onygenaceae (sensu lato) may
have originated in the Americas 320 million years
ago (mya) (Fisher et al. 2001; Kasuga et al. 1999, 2003).
Some of the most important members (Histoplasma
and Paracoccidioides genera) almost certainly evolved
105
in South America, perhaps not coincidentally in the

same area where Xenarthra were evolving, prior to the
connection of the Panamanian Isthmus.
The case of Paracoccidioides brasiliensis

Paracoccidioides brasiliensis is the etiological agent of
paracoccidioidomycosis (PCM), one of the most important systemic mycoses in Latin America. This fungus is thermodimorphic, that is, it grows as a yeastlike
structure in host tissue or when cultured at 3537C,
and as mycelium in the saprobic condition or when cultured at room temperature (1823C, figure 9.1). The
teleomorphic or sexual (meiospore) form is still unknown (Lacaz 1994) and its natural habitat has not been
clearly established. Several indirect indications suggest
the saprobic form may occur under restricted soil conditions, producing airborne infective propagules, such
as arthroconidia and aleuroconidia, that cause primary
infection in the lung (Restrepo 1985; Restrepo-Moreno
1994).
Early detection of PCM can be difficult because often there are no symptoms. Intradermal tests with the
so-called paracoccidioidins indicate high incidence in
endemic areas, with reactivity ranging from 2% to 60%
(Restrepo-Moreno 1994; Fava and Fava-Netto 1998).
For full-blown development of the disease, with uni- or
multifocal lesions, this rate decreases to ca. 13 cases per
100,000 inhabitants (Wanke and Londero 1994). PCM
is a life-threatening disease, representing the eighth
most common cause of death from chronic or recurrent types of infectious and parasitic diseases in Brazil
(Coutinho et al. 2002). Two distinct forms of PCM have
been observed: (a) an acute or juvenile form that occurs in young people of either sex and seems to develop
more rapidly once the primary infection has occurred,
and (b) a chronic or adult form that is observed mainly
in males and characterized by a prolonged period of
latency that, in some well-documented cases, can last as
long as 3040 years (Franco et al. 1994). Gender association in PCM seems related to the action of the female
hormone 17-beta-estradiol, which apparently inhibits
the transition from mycelial to yeast form (Salazar et
al. 1988; Aristizabal et al. 2002).
PCM has a significant social impact in most Latin
American countries, in part because it affects mainly
rural workers in their productive period of life. One
reason no measures or strategies to avoid new infection
have been promoted is because the ecological niche of
the pathogen is unknown, which in turn is due to the
106
Figure 9.1. P. brasiliensis dimorphism. Macroscopic aspect of the mycelial form, at room temperature. A. Microscopic aspect of
the hyphae, containing arthroconidia and aleuroconidia (A1, arrows, X400 magnification), and chlamidoconidia (A2, X1,000
magnification). B. Macroscopic aspect of yeast form, at 35C. Microscopic aspect of the multiple-budding yeast cells (B1,
X1,000 magnification). Slides stained with Lactophenol cotton-blue.
fact that the fungus has rarely been isolated from the
environment. Other reasons for the lack of preventive
measures include the prolonged latency period of the
disease and the fact that no outbreaks have been reported (Franco et al. 1994, 2000). PCM treatment is
frequently difficult and prolonged; patients often need
hospitalization, with increasingly high treatment costs.
Occurrence of sequelae is another important aspect of
PCM, frequently causing pulmonary dysfunction, serious dermatological deformations, or other disabilities.
Relapses, either by activation of endogenous quiescent forms or by new infections, also occur frequently
(Franco et al. 1994).
While P. brasiliensis can be cultured quite easily
from clinical samples, the same does not hold true
for suspected environmental samples. It is possible to
107
Table 9.1. Fungal Detection in Armadillos

Armadillo species
Sample size
and location
(number positive)
Pathogen
Culture
Dasypus novemcinctus
Par, Brazil
Par, Brazil
So Paulo, Brazil
So Paulo, Brazil
So Paulo, Brazil
Minas Gerais, Brazil
Minas Gerais, Brazil
Gois, Brazil
Caldas, Colombia
Argentina
Par, Brazil
Par, Brazil
Piau, Brazil
Florida, U.S.A.
Florida, U.S.A.
Dasypus hybridus
So Paulo, Brazil
Cabassous centralis
Caldas, Colombia
Fungal detection
Histopathology
PCR
Source
20 (4)
29 (18)
15 (10)
1 (1)
5 (2)
21 (1)
16 (4)
5 (2)
2 (1)
1 (1)
3 (1)
11 (1)
26 (3)
17 (2)
1 (1)
P. brasiliensis
P. brasiliensis
P. brasiliensis
P. brasiliensis
P. brasiliensis
P. brasiliensis
P. brasiliensis
P. brasiliensis
P. brasiliensis
H. capsulatum
H. capsulatum
H. capsulatum
C. immitis
S. schenckii
S. schenckii
1 (1)
P. brasiliensis
+ (ML,SP)
15
1 (1)
P. brasiliensis
+ (SP)
ND
16
+ (LI,SP)
+ (LI,SP)
ND
+ (LI,SP)
+ (LI,SP)
ND
+ (LI,ML,SP)
+ (ML)
ND
+ (SP,ML)
-
+ (FE,ML,SP)
-
-
+ (FE,LU,ML,SP)
ND
+ (LU)
ND
+ (LI,LU,SP)
+ (ML,SP)
ND
+ (SP)
-
ND
+ (ML)
ND
+ (LI,ML)
+ (LI,SP)
ND
ND
+ (LI,SP)
ND
ND
+ (LI,SP)
ND
ND
+ (LI,SP)
-
ND
+ (KI)
+ (KI)
ND
ND
+ (HR,KI,LI,LU,SK,SP)
ND
4
5
7, 14
15
15
11
12
10
9
3
1
6
13
2
8
Sources: (1) Arias et al. 1982, (2) Kaplan et al. 1982, (3) Bogado et al. 1983, (4) Naiff et al. 1986, (5) Naiff and Barret 1989, (6) Naif et al.
1996, (7) Bagagli et al. 1998, (8) Wenker et al. 1998, (9) Corredor et al. 1999, (10) Macedo et al. 1999, (11) Silva-Vergara and Martinez
1999, (12) Silva-Vergara et al. 2000, (13) Eullio et al. 2001, (14) Bagagli et al. 2003, (15) Bosco et al. 2005, (16) Corredor et al. 2005.
Notes: Abbreviations: ND = no data, + = positive, - = negative, FE = feces, HR = heart, KI = kidney, LI = liver, LU = lung, ML = mesenteric
lymph node, SK = skin, SP = spleen.
induce the fungal saprobic form in vitro by using artificial culture media or some natural substrates, such
as sterilized soil. Growth of the mycelial phase is relatively slow in comparison with other decomposer fungi
and it seems to have a restricted ability to compete or
survive in the presence of normal soil microbiota (Restrepo 1985; Lacaz 1994). The production of infective
propagula, such as arthro- and aleuroconidia, as well
as of resting spores (chlamydospores), has been demonstrated occasionally in the laboratory, though only in
some isolates and in very low amounts (Restrepo 1985;
Franco et al. 1989).
Armadillos and P. brasiliensis

Evidence of natural P. brasiliensis infection in some
wild and domestic animals was detected by means of
intradermal and serological tests (Costa, Diniz, and
Fava-Netto 1995; Costa, Diniz, Fava-Netto et al. 1995;
Ono et al. 2001). However, it was only from the viscera
of armadillos that the pathogen could be repeatedly
isolated, a fact that opened new avenues for study of
the fungus. The association between P. brasiliensis and
nine-banded armadillos was first observed in animals
obtained from the Amazon region (Naiff et al. 1986;

Naiff and Barret 1989; table 9.1). Similar results were
subsequently obtained in other areas of Brazil (Bagagli et al. 1998, 2003; Macedo et al. 1999; Silva-Vergara
and Martinez 1999; Silva-Vergara et al. 2000) and in
Colombia (Corredor et al. 1999; Restrepo et al. 2000;
table 9.1). In the latter, the fungus was also isolated from
the naked-tailed armadillo Cabassous centralis (Corredor et al. 2005). The frequency of P. brasiliensis infections in nine-banded armadillos is high. Data from the
published literature indicate nearly 100 animals have
been examined; of these, 37% harbored P. brasiliensis
in their organs (see table 9.1). In some Brazilian counties located in a hyperendemic PCM area, the fungus
has been isolated in 75100% of the evaluated animals
(Bagagli et al. 1998, 2003). P. brasiliensis isolates from
D. novemcinctus are indistinguishable from clinical
isolates in terms of antigen, virulence, and molecular
profiles, confirming that the same ecopathogenotypes
infect animals and humans (Sano, Defaveri et al. 1999;
Sano, Tanaka et al. 1999; Hebeler-Barbosa, Montenegro,
and Bagagli 2003; Hebeler-Barbosa, Morais et al. 2003;
Corredor et al. 2005).
Nine-banded armadillos are regularly consumed as
108
Figure 9.2. Necropsy of D. novemcinctus viscera exhibiting enlargement of mesenteric lymph nodes (A, arrows) that in slide
sections stained with Gomori-Methenamine Silver showed granuloma with the presence of P. brasiliensis (B, arrow). Picture by
Mrio Rubens Guimares Montenegro, UNESP, Brazil.
supplementary food by people in rural areas of Latin

America (see Noss et al. this volume), and contact with
these animals is a risk factor for PCM (Cadavid and
Restrepo 1993). Therefore, public health campaigns
have advocated avoiding direct contact with these
animals and/or their habitat. Nine-banded armadillos
have small home ranges (see McDonough and Loughry
this volume), and do not migrate, facts that make them
perfect sentinel animals for locating hot spots where
the fungus may occur in nature. Using this approach, it
has been demonstrated that P. brasiliensis occurs preferentially in a limited range of habitats, typified by the
presence of disturbed and humid vegetation located
close to water sources (Restrepo et al. 2000; Bagagli et
al. 2003). Although several attempts to isolate the fungus in its saprobic form from armadillo habitat have
been unsuccessful, using PCR we were able to amplify
P. brasiliensis DNA in soil samples collected inside armadillo burrows (Theodoro et al. 2005) and from feces
of naturally infected animals (Bosco et al. 2005).
Whether the relationship between P. brasiliensis
and armadillos is commensal or parasitic is not yet
completely understood. In studies of D. novemcinctus,
the fungus has been recovered from young and older
adult animals, of either sex, but with no visible signs
of PCM (Bagagli et al. 1998, 2003). However, the presence of granulomas with fungal elements, compatible
with P. brasiliensis, was observed in lymph nodes, liver,
and lungs of a few animals (figure 9.2); this might indicate a condition of active PCM disease in some of them
(Bagagli et al. 1998, 2003; Silva-Vergara and Martinez
1999; Silva-Vergara et al. 2000).
The pathogen has been repeatedly isolated, either by

inoculation of laboratory animals (hamsters) or macerated armadillo organs, or by direct culturing of mesenteric lymph node, spleen, or liver fragments. Direct
culture from fragments of lungs has given no positive
result mainly because of the presence of contaminant
flora that grow more quickly than the pathogen. In
some cases, a single animal provides positive culture
from different organs (lymph nodes, spleen, and liver),
clearly indicating that after infection (most probably by
airborne route) with the fungus propagula produced by
the saprobic mycelial form, the pathogen must transform to yeast in the lungs and disseminate systemically
into the armadillo host, in a manner similar to that
which may occur in humans. However, the frequent
isolation of the fungus from mesenteric lymph nodes,
and also its molecular detection in feces, might suggest
that armadillo infections occur through an alternative
route, such as the gastrointestinal tract (Bosco et al.
2005).
The armadillo group Dasypodidae and Ajellomycetaceae fungi are both ancient lineages that may have
been cohabiting in the same continental area for at least
1020 million years. Thus, one might ask whether the
armadillo-P. brasiliensis association is an ancient one or,
alternatively, if it has occurred only in recent times as
the result of human action. The latter scenario is plausible, as it seems to have happened with Mycobacterium
leprae, which probably originated in Africa and was introduced into America (and into armadillos) less than
500 years ago (Monot et al. 2005; Truman this volume).
Several features of P. brasiliensis seem more compatible
with a long evolutionary relationship with animal hosts.

These include: dimorphism, poor saprobic sporulation,
restricted occurrence in nature, and difficulties in isolation from the environment. Similarly, features of PCM
such as chronic forms, long latency, relapses, and male
association may also indicate an ancient relationship.
However, if armadillos are only recent hosts, then it is
very probable that P. brasiliensis had one or more ancient animal hosts (Bagagli et al. 2006).
Many important questions concerning the armadillo-P. brasiliensis relationship remain and should be
subjects for further studies. For example, do infected
armadillos eliminate viable fungus into the environment? If so, do they contribute to fungus dissemination
in nature and provide sources for human infection? Instead of being a blind alley, could armadillo infection
be a two-way street, that is, could these animals be important for the fungus life cycle in nature, contributing
not only to the anamorphic but also to the teleomorphic (sexual) phase? Sexual reproduction, despite being
the principal force for evolution, shows some peculiarities in fungi, mainly in the difficulties of encountering
compatible mating types in the immense vastness of
nature. Could armadillos facilitate sexual encounters
by carrying the fungus to a common and protected environment, for example, burrows or concentrations of
animal excreta?
There are some interesting examples in which fungal
species clearly take advantage of an animal host and
its products in order to accomplish dissemination and
sexual reproduction, for example, coprophilous fungi
associated with herbivorous animals. In some cases,
such as illustrated by Pilobolus crystalinus (phylum Zigomycota) and Ascobolus sp. (phylum Ascomycota),
the fungi have developed efficient mechanisms of phototropism and active ejection of their spores from animal feces in order to reach and attach to a new sprout of
grass. This grass will then be consumed and expelled by
the animal, thus increasing fungal chances of being in
the preferred habitat, finding partners, and developing
anamorph and teleomorph phases (Kendrich 2000).
Other species also use animals to facilitate reproduction. Species of the genus Onygena characteristically
engage in sexual reproduction in horns and/or feathers
(keratin-rich substrates) (Currah 1985). During sexual
reproduction in some Onygenales species, for example,
Ajellomyces capsulatus and Ajellomyces dermatitidis,
morphological ornaments, such as coiled hyphae and
helical appendages, are produced in the ascocarp. Some
109
authors consider these structures deterrents to avoid

arthropod predation, but they also may be a strategy
for dissemination by attachment to animal hair or skin
(Weitzman and Summerbell 1995; Kendrich 2000).
Molecular and mycological characterization of P.
brasiliensis isolates obtained from armadillos captured
in a well-defined geographic region, as well as from
clinical isolates from the same area, has provided some
information regarding the fungus sexual or clonal population structure, and the eventual increase in the rate
of sexual recombination within the animal reservoir.
The data so far suggest that armadillos harbor a wider
spectrum of genotypes than is found in clinical isolates
(Hebeler-Barbosa, Montenegro, and Bagagli 2003; Hebeler-Barbosa, Morais et al. 2003). It was also observed
that individual animals acquire multiple P. brasiliensis
infections, with distinct isolates belonging to different
genotypes (Sano, Defaveri et al. 1999; Sano, Tanaka et al.
1999). These observations support the hypothesis that
animal infection is critical for the fungus, for example,
by enhancing sexual reproduction, but the subject requires additional study.
Other dimorphic fungi in armadillos

The same group of researchers that first isolated P.
brasiliensis from armadillos in the Amazonian region
did so while searching for sylvatic animal reservoirs for
Leishmania spp. and Trypanosama spp. During their
investigations they isolated Histoplasma capsulatum in
nine-banded armadillos (1 positive from 11 evaluated
animals, table 9.1) and several other animal groups. Curiously, some of these other species were canopy dwellers, such as sloths (Choloepus didactylus, 1/7 positive)
and opossums (Didelphis marsupialis, 14/393 positive),
although pacas, Agouti paca, also tested positive (4/30;
Arias et al. 1982, Naiff et al. 1985, 1996). H. capsulatum
was also isolated from Dasypus hybridus in Argentina
(Bogado et al. 1983; table 9.1). Among the pathogenic
dimorphic fungi from the Onygenales, H. capsulatum
is the most widely distributed species, occurring in different continents but most highly concentrated in the
Americas (Rippon 1988). Molecular phylogeographic
studies with a great number of H. capsulatum isolates
indicate that this species most likely originated in South
America about 313 mya (Kasuga et al. 1999, 2003). This
fungus can be recovered repeatedly in its saprobic form,
usually in soil enriched with feces of bats and birds lo-
110
cated in protected environments such as cave interiors

and tree hollows (Rippon 1988).
Coccidioides immitis, considered one of the most virulent pathogenic fungi from the Onygenales order, occurs in some arid regions of the United States, but was
also detected recently in some arid areas of northeast
Brazil. Coccidioidomycosis in Brazil has been observed
mainly in armadillo hunters and their dogs; the fungus
has been isolated from both armadillo internal organs
and their burrows (Silva et al. 1997; Wanke et al. 1999;
Costa et al. 2001; Eullio et al. 2001; table 9.1). In contrast to P. brasiliensis and H. capsulatum, the origin and
dissemination center of C. immitis is presumed to be
North America. Molecular studies indicate this pathogen may have reached South America accompanying
the human migration that occurred 140,0009,000
years ago (Fisher et al. 2001).
Also in connection with the armadillo, we should
mention Sporothrix schenckii (Ophiostomataceae,
Ophiostomatales, Ascomycota), a cosmopolitan dimorphic fungus that causes sporotrichosis. Different from
the Onygenales dimorphic pathogens, where infection
occurs by the airborne route, sporotrichosis is acquired
normally by a traumatic route affecting mainly the cutaneous and subcutaneous tissues, with rare systemic
dissemination (Rippon 1988). Sporotrichosis has been
described in several wild and domestic animals, including cats that present a disseminate form of the disease that in turn may transmit the mycosis to humans
(Reed et al. 1993; Dustan et al. 1986; Fleury et al. 2001).
Disseminated systemic sporotrichosis causing animal
death was described in three nine-banded armadillos,
in two different reports (Kaplan et al. 1982; Wenker et
al. 1998; table 9.1). All three affected armadillos originated in Florida and were being maintained in captivity,
either for leprosy studies (Kaplan et al. 1982) or in a private Swiss zoo (Wenker et al. 1998). In Uruguay, close to
half of human sporotrichosis cases are associated with
hunting armadillos. The fungus has been recovered
from armadillo nests and burrows, but not from the
animals themselves (D. septemcinctus, Mackinnon et
al. 1969). We have experienced a case of zoonotic sporotrichosis by helping in the mycological diagnosis of a
southern Brazilian Army soldier (Jungle Infantry). He
developed the disease (subcutaneous form with lym-
phatic dissemination) after having been injured by an

armadillos claw while handling the animal, most probably D. novemcinctus. Finally, while attempting to isolate P. brasiliensis in soil from armadillo burrows (using
animal inoculation of hamsters and mice), we isolated
S. schenckii in a sample collected in the interior (1.5 meters deep) of a burrow; the animal occupant had been
previously collected and proved to be culture positive
for P. brasiliensis (Bosco et al. 1999).
Conclusions
To date, most pathogen studies have focused mainly
on nine-banded armadillos. An obvious goal for the
future is to broaden taxonomic coverage by sampling
other species of xenarthrans. However, even within the
armadillo-pathogen relationship, many questions remain. The following represent possible starting points
for future investigations:
1. Why have armadillos not been driven to extinction
by such frequent exposure to a gamut of different
pathogens?
2. Do armadillos play a key role in the evolution of
some important pathogens from South America, or
are they only circumstantial victims (like humans)?
3. Do armadillos have a special affinity for contraction
of human pathogens?
Obviously, there is still much to learn about armadillos
(and xenarthrans more generally) and their pathogens.
While interesting in their own right, studies of armadillos may have real value as an efficient low-cost strategy
for monitoring several important human infectious
zoonotic diseases.
To close on a more practical note, given the evidence
presented above, we would like to emphasize that educational campaigns should be stimulated to inhibit and/
or discourage the hunting of armadillos, especially in
Latin America.
Acknowledgments
We thank the State of So Paulo Research Foundation (Fapesp) for financial support (grant numbers
02/00466-5 and 06/03597-4).
10
Leprosy
Richard W. Truman
Resumen
Introduction
El armadillo Dasypus novemcinctus representa la nica

especie de la familia susceptible a la infeccin con Mycobacterium leprae, el agente etiolgico de la lepra.
Desde la dcada de 1970, estos armadillos han sido
seleccionados como huspedes para la propagacin
in vivo del bacilo de la lepra y su uso como modelos
translacionales para desarrollar reactivos diagnsticos
y vacunas contra la lepra. Algunos ejemplares silvestres
tambin sufren de infeccin natural con M. leprae y pueden representar un riesgo de infeccin por contacto
para algunos humanos. En este captulo se revisa el uso
de estos armadillos como modelos para las investigaciones sobre lepra y su importancia en la perpetuacin
de la lepra en humanos.
Leprosy is a slow, chronic, infectious disease caused

by Mycobacterium leprae. It primarily affects the peripheral nervous system and involves skin and other
tissue. Leprosy puts heavy burdens on both its victims
and the healthcare systems serving them. More than
515,000 new leprosy cases were detected globally in
2003 (World Health Organization [WHO] 2005), and
the United States had 130 new presentations (Pfeiffer
and Truman 2003). Besides humans, nine-banded armadillos (Dasypus novemcinctus) are the only natural
hosts of M. leprae. Armadillos are uniquely important
to leprosy research as both models of the infection, and
the principal source of leprosy bacilli for research and
diagnostic purposes (Truman and Sanchez 1993). In
addition, wild armadillos in the south-central United
States harbor a natural infection of M. leprae, which
may be a source of infection for some people (Truman
2005). I review here the use of armadillos as laboratory animals and disease models in leprosy research,
highlighting their unique contributions to the study of
leprosy, some of the challenges associated with the care
and use of armadillos in the laboratory, and the role
of wild armadillos in perpetuating human leprosy in
the Western Hemisphere. While the term armadillo
can be appropriately applied across many genera, the
armadillo of most importance to leprologists is Dasypus novemcinctus (known also as the nine-banded or
common long-nosed armadillo). There is only a single
report relating to leprosy in another armadillo species
(Balia et al. 1985), and, in this review, the term armadillo will refer only to D. novemcinctus.
Resumo
O tatu Dasypus novemcinctus unicamente suscetvel
a infeco pelo Mycobactacterium leprae, o agente etiolgico da lepra. Desde a dcada de 1970 esse tatu tem
sido o hospedeiro de escolha da propagao in vivo do
bacilo da lepra e eles vm sendo utilizados nos dias de
hoje como modelo de transio para o desenvolvimento de reagentes diagnsticos e vacinas anti-lepra. Alguns tatus selvagens tambm apresentam uma infeco
natural pelo M. leprae e podem constituir um risco de
infeco para humanos pelo contato com eles. Nesse
captulo eu revisei o uso desses tatus como modelos
para a pesquisa de hansenase e sua importncia na
perpetuao dessa doena entre os humanos.
111
112
R. W. Truman
Laboratory maintenance
Armadillos present many challenges for care in the
laboratory (see also Superina et al. this volume). They
do not reliably breed in captivity and must be obtained
from the wild for investigative purposes. Recently captured animals can suffer stress which may lead to a
number of medical and husbandry problems. Plasma
cortico-steroid concentrations among indoor housed
armadillos may equal or exceed those of free ranging
animals enduring a hard winter (Rideout et al. 1985).
Outdoor housing seems to reduce stress, and animals
kept individually indoors have fewer problems than
animals housed with others. Treatment of newly captured animals with steroids and prophylactic antibiotic
therapy greatly enhances the adjustment to captivity.
Births are not uncommon in the laboratory but occur primarily among animals that mated in the wild
and were brought into captivity while already pregnant.
Among some 700 females housed communally indoors
at our facility, we recorded only 14 successful matings.
Fecundity rates in outdoor pens can approach 25%, but
infant survival remains low (Job et al. 1984). Female
armadillos brought into captivity during their 45month diapause period (July to November, see Peppler
this volume) may lose their embryos or enter a state of
superdelayed implantation (Galbreath 1982). Storrs has
reported delays of up to 3 years between natural mating
and parturition (Storrs et al. 1988).
A variety of circumstances are employed for housing
armadillos, including enclosed kennels, runs, stalls, and
outdoor pens (Gilbert and Giacometti 1972; Burchfield
et al. 1976). The principal requirement is a sturdy unit
capable of withstanding this strong animals digging
and clawing. Armadillos are capable climbers and jumpers. Open-top enclosures must be at least 1.2 m high to
prevent escape. Hard-bottomed units require 2.55 cm
of paper or absorbent bedding to prevent foot problems,
and may benefit from the addition of cat litter. Softbottomed outdoor enclosures require barriers to a depth
of 1.8 m to prevent burrowing escapes. Armadillos are
good swimmers and are not contained by moats.
Armadillos of similar size and temperament can
be housed together. However, more vigorous animals
may show aggression toward the less able and co-housing armadillos with any open wounds can encourage a
licking behavior that may lead to cannibalism. Because
of these problems, we prefer to hold armadillos individually in standard rabbit cages fitted with injection
molded plastic inserts (Allentown Caging Equipment,
Allentown, NJ). Shredded paper is provided in a 20

36 cm plastic trash can that also serves as a sham burrow. The animals can be adapted to water bottles and
J-feeders, or given daily rations.
A number of diets have been described for armadillos (Meritt 1976a; Ramsey et al. 1981; Superina et al.
this volume). They are insectivorous to omnivorous in
nature; their natural diet comprises 75% insects, 15%
animal material, and 10% vegetation and dirt (Talmage
and Buchanan 1954). We find that a mixture of liver-flavored canned dog food, moistened dried cat chow, and
water with 1% mineral salts (Morton, Iofixt), blended
(1:1:1) and given as one cup per animal per day, is easy
to handle and sufficient such that most animals gain
weight. Armadillos also reportedly enjoy eggs, chicken,
and other meats, as well as some fruits such as bananas
and melons (Storrs 1971). It is important to supplement
the diet of captive animals with a source of vitamin K
to prevent severe anemia and clotting abnormalities.
Menadione equine supplements can be sprinkled over
the food for good effect.
Medical problems
The hard carapace makes armadillos difficult to examine or treat, and incessant efforts to escape make auscultation of the heart and lungs quite difficult (Wampler
1969). Injections and bleeding are best accomplished
through the saphenous or subclavian veins, as most
other vessels are obscured by the carapace or tough
skin. Toe-nail clips and Nobuto (Advantec, Dublin,
CA) strips can be used effectively for a variety of analyses. All disorders of armadillos respond to treatment
better when the animals are held in individual isolation
or in natural outdoor pens and burrows.
Armadillos require special care while adapting to
captivity. Entering animals are usually provided food
in a soupy slurry with water. Adults typically undergo
a period of anorexia that can be accompanied by dehydration. Fluids can be replaced by intravenous or intraperitoneal routes but they may not reverse the condition (Wampler 1969). The response is best when food
and water are taken orally. Sugar or other sweetener can
be added to the slurry to encourage consumption.
Minor wounds and injuries usually respond well
when washed under running water with Betadine soap
and treated topically daily with either hydrogen peroxide, Merthiolate, Furacin, or Granulex. Cracked, torn,
or punctured carapace wounds can result in severe
necrotizing infections and must be treated promptly
with systemic antibiotics. Benzathine penicillin or
Leprosy
other long-acting preparations that minimize the need

to repeatedly handle animals are most effective. Tail
injuries are especially serious and require immediate
attention. Amputation is occasionally necessary. Tail
problems occur most frequently when animals jump or
fall while trying to escape open-topped cages or runs.
We find that armadillos respond well to the most commonly used gas and injectable anesthetics, in dosages
appropriate for cats, and have had satisfactory experiences with ether, Halothane, Innovarvet, ketamine, and
ketamine combined with Rompun or Detomadine. We
use an intramuscular mixture of 0.1 mg/kg Detomidene
with 10 mg/kg ketamine to achieve a good working anesthesia, with analgesia lasting 2030 minutes without
the need of a reversal drug.
The principal maladies associated with housing armadillos are pododermatitis and bowel impaction. Both
conditions are rare among animals housed in outdoor
pens and present more commonly when armadillos are
maintained on artificial surfaces. Though pododermatitis can result in the loss of the nail or fulminate to gangrene involving the entire leg, such severe sequelae are
usually avoided with early topical treatment with zinc
oxide, crystal violet, furasol, or systemic antibiotics in
chronic lingering problems. Bowel impactions may occur among animals debilitated by captive stress or those
held in runs with excess bedding material that can be
kicked into their food. The symptoms are typical, for
example, anorexia and lethargy. However, they can be
difficult to note among co-housed animals and daily
observation of eating behavior and feces production is
recommended. For impacted animals that continue to
eat, 20 cc of molasses added to the food is often productive. For those that stop eating, enemas with water or
mineral oil are occasionally effective.
Wild animals may harbor a number of infectious
agents that can complicate their use in the laboratory.
Armadillos live in the same general environment as
opossums, skunks, and raccoons but are comparatively
free of parasites (but see Bagagli and Bosco this volume). Their heavy carapace and scant body hair contribute to a paucity of external parasites; internally, low
body temperature probably serves as a highly selective
host factor. Cestodes and acanthocephala are rare (Talmage and Buchanan 1954) other than as encysted larval
forms on mesentery membranes. Nematodes (Talmage
and Buchanan 1954) and schistosomes (Krotoski et
al. 1984) are common, as are sarcocysts (Lindsay et al.
1996). Armadillos occasionally carry leptospira (Roth
et al. 1964) and Trypanosoma cruzi (Yaeger 1988). More
113
significantly, though, many appear to be naturally infected with M. leprae.
Propagation of M. leprae
M. leprae is not cultivable on artificial laboratory media and, prior to 1960, could be obtained only from
infected human tissue (Kirchheimer 1982). However,
because human lesions occur mostly in the extremities, Shepard (1960) reasoned that cool body temperature might favor the growth of M. leprae. He inoculated
suspensions of leprosy bacilli into the footpads of mice
and the localized infection that developed over a period
of 6 months was the first successful culture of M. leprae
outside of humans. Expanding on Shepards observation that cool temperatures favored the growth of M.
leprae, Kirchheimer and Storrs began experimentally
infecting armadillos in 1968. Armadillos were chosen
because even under normal conditions an armadillos
core temperature is in the range of 3235C (Storrs
1971). By 1971 Kirchheimer and Storrs had shown that
armadillos were remarkably susceptible to M. leprae
and that huge quantities of leprosy bacilli could be
made available through them (Kirchheimer and Storrs
1971). The M. leprae made available through armadillos
has been a boon to leprosy research and provided the
resources necessary to develop specific diagnostic antigens (Caldwell et al. 1979; Hunter et al. 1982; Spencer et
al. 2002) and prototype vaccines (Rees 1983; Crawford
1992; Job et al. 1993; WHO 1995; Gupte et al. 1998).
Armadillos are susceptible to infection with M. leprae by a number of routes, including intravenous, percutaneous, and respiratory instillations (Truman and
Sanchez 1993). As few as 103 M. leprae are sufficient to
establish infection (Job et al. 1985). However, incubation
periods are shortest when high doses of bacilli (0.14.0
109) are delivered intravenously (Prabhakaran et al.
1984). These bacilli are taken up in reticuloendothelial
tissues where they slowly proliferate and disseminate
to other parts of the body. Intermittent low-level bacteremia leads to a generalized dissemination of bacilli
in the late stages of the infection. No organ system is
spared, but cooler body regions such as ears, nose,
tongue, footpads, bronchi, and lungs tend to exhibit
greater involvement; adrenal glands, bladder, heart,
intestine, kidneys, ovaries, and testes are involved less
commonly (Job et al. 1985, 1992).
The specific immunological factors underlying the
armadillos susceptibility to leprosy remain unclear.
Armadillos exhibit the full spectrum of immunological
114
R. W. Truman
Table 10.1. Representative Clinical Values for Normal Armadillos and Animals Experimentally Infected with M. Leprae Manifesting
Fully Disseminated Late Stage Disease

Chemistryc

Hematologyd

Test
Fully Disseminateda
Normalb
ALP IFCC (IU/dL2)

ALT (IU/dL2)
BUN (mg/dL2)
Creatinine (mg/dL2)
Glucose (mg/dL2)
LDH-L (IU/dL2)
86 13.7
9.7 5.5
24 7.6
0.95 0.13
78 19
3394 1220
64 22.6
6.4 3.2
24 6.5
0.78 0.19
81 21
740 45
Red blood cells ( 106/ml)

Hemoglobin (g/dL2)
Hematocrit (%)
Mean cell volume (mu3)
Mean cell hemoglobin (pg)
5.1 0.8
9.8 1.5
29 5.6
58.1 2.2
19.2 0.7
6.9 0.3
13 0.7
40 3
61 2
19.5 0.5
Notes: Abbrevations: ALP-IFCC = alkaline phosphotase determined by International Federation of Clinical Chemists procedure, ALT =
alanine aminotransferase, LDH-L = lactate dehydrogenase using l-lactate reaction.
a. Means SD from sacrifice blood samples of 7 animals that yielded >109 M. leprae/gram of liver tissue.
b. Means SD from 45 naive armadillos held in captivity for 60 days.
c. Chemistries determined on Ciba-Corning auto-analyzer with reagents and procedures according to manufacturers recommendations.
d. Performed on Coulter JT-3 gated for human cell types.
responses to M. leprae ranging from Tuberculoid (TT)

to Lepromatous (LL) and can be classified according to
the Ridley-Jopling scale with Lepromin (Job et al. 1983,
1985). The majority of armadillos appear to be multibacillary types. The specific response and time course
of experimental leprosy exhibits individual variation.
Incubation periods may be as short as 10 months or as
long as 4 years (Job et al. 1985). M. leprae infection in
armadillos produces few gross symptoms, and it is not
possible to distinguish normal from M. leprae-infected
animals by their outward appearance (Binford et al.
1976, 1977). M. leprae can be demonstrated in the skin,
nodules, lymph nodes, blood, or other organ tissues
with direct smears, histopathological exam, and PCR
(Job et al. 1985, 1991; Walsh et al. 1986). Multibacillary
armadillos also have a strong antibody response to M.
leprae. Among laboratory-infected animals, IgM antibodies to the M. leprae-specific phenolic-glycolipid-1
antigen (PGL-1; Hunter et al. 1982) arise in about a third
of the time required for bacilli to become detectable in
skin scrapings and ear biopsies (Truman, Morales et
al. 1986; Truman et al. 1990). The timing of their appearance and general level is highly correlated with the
bacterial load in the animals tissues (Truman, Morales
et al. 1986). First detection is generally associated with
a 1+ BI (Bacteriological Index, approximating 10 4 bacilli/gram) in some reticuloendothelial tissue (Job et al.
1990, 1991). Levels of IgM antibodies to PGL-1 increase
with increasing bacterial load in the animal and per-
sist over the course of the disease (Truman, Morales et

al. 1986). Most animals develop heavy infections, with
approximately 1012 recoverable bacilli in the liver and
spleen within 1824 months (Job et al. 1985; Truman
and Sanchez 1993). A number of abnormalities associated with a severe hypochromic microcytic anemia
appear in the hemogram (table 10.1).
Leprosy in wild armadillos

A leprosy-like disease was first reported among wild armadillos in 1975 (Walsh et al. 1975), and by 1983 the sylvan agent had been confirmed to be identical to the M.
leprae infecting humans (Smith et al. 1983). Armadillos
today are recognized as a large natural reservoir of M.
leprae. Numerous studies have shown that M. leprae
is highly prevalent among wild armadillos in parts of
Louisiana and Texas, but rare or absent in other locales
(Truman 2005). The probable origin of the infection,
its geographic range, and the risks it might present to
humans have been subjects of considerable speculation.
Some editorials suggested that sylvan leprosy was a new
agent that could be useful in modeling human leprosy
(Weiser 1975). Others hailed the report as revealing a
new zoonosis and a threat to public health. However,
DNA homology studies showed that the sylvan agent
actually was M. leprae and not something new (Smith et
al. 1983). Skinsnes (1976) speculated that sylvan leprosy
was newly evolved and that it might have originated
Leprosy
through the escape or improper disposal of experimentally infected animals at one of the research centers
working with armadillos. The accusation polarized the
community and trivialized armadillo studies for years.
The environmental contamination claim was perpetuated by several groups, and is frequently alluded
to even today by uninformed individuals (Kirchheimer
1977; Kirchheimer and Sanchez 1978a; Kirchheimer
1979; Walsh et al. 1986).
A number of surveys were launched to determine
the geographic range of armadillo leprosy by sampling
for systemic acid-fast bacilli in ear tissues. Walsh et al.
(1977) found 50 M. leprae-infected armadillos from
among 459 animals sampled at 11 different locations in
Louisiana and 1 infected animal in Texas. With such a
wide geographical distribution it seemed unlikely that
sylvan leprosy might be a recent phenomenon. However, no evidence of the infection was found among
armadillos in Florida, and only one M. leprae-infected
armadillo (out of 218 sampled) was found in Mississippi (Fox et al. 1977; Walsh et al. 1977, 1986). Kirchheimer examined more than 400 armadillos from various locations in central Texas, Louisiana, and Florida
before confirming existence of the disease in a single
Louisiana animal (Kirchheimer 1977; Kirchheimer and
Sanchez 1978a,b; Walsh et al. 1986). Smith showed that
leprosy was highly prevalent (21/451) among armadillos in counties along the Texas coastline and extended
toward Mexico (Smith et al. 1978, 1983). However, only
one (1/96) infected armadillo was found in Mexico
(Amezcua et al. 1984). Eventually, one other infected armadillo was found in Argentina (Martinez et al. 1984).
Otherwise, none (0/536) of several armadillo species
examined in either Colombia (Muoz-Rivas 1978) or
Paraguay (Innami 1978) appeared to be infected. The
apparent geographical distribution of the infection
fueled continued speculation that armadillos had acquired leprosy in the United States, perhaps through
natural mechanisms (Myers et al. 1978), or from contamination of their environment.
The environmental contamination hypothesis was
disproved finally in 1985. With the advent of serological screening methods to detect antibodies to the PGL-1
antigen, we examined sera taken from wild armadillos
in years predating their use in leprosy research to determine the time frame in which armadillos might have
acquired M. leprae (Truman, Shannon et al. 1986). Collected by Roth (Roth et al. 1964) in 19601964 as part
of a survey for leptospires in Louisiana wildlife, the sera
had been stored frozen at Louisiana State University for
115
25 years prior to our study. We found that 17/182 sera

reacted specifically with PGL-1, indicating that M. leprae must have been enzootic among armadillos at least
by 1961. Armadillo leprosy could not have originated
through any possible accidental contamination of the
environment on the part of leprosy researchers, who
only started working with armadillos in 1968 (Truman,
Shannon et al. 1986). Therefore, it must have evolved
by natural means and armadillos could have acquired
the infection in any number of different locations or
repeated the event many times over the years.
Geographic distribution of armadillo leprosy

Armadillos are not native to the United States but began slowly expanding their range north from Mexico
around 1880. They first entered Louisiana in 1926 and
had crossed the state to the Mississippi River by 1957
(Humphrey 1974). Today armadillos are found from
Argentina to Colorado and through the southeastern
United States to Florida (Taulmann and Robbins 1996).
They are present on many Caribbean islands and likely
were carried there aboard trading ships. Several reports
relate private citizens introducing armadillos into the
state of Florida beginning in about 1922 (Humphrey
1974). These animals established a second U.S. population, which also slowly expanded its range and merged
with the main U.S. group around 1985 (Smith and
Doughty 1984; Taulmann and Robbins 1996).
Seeking mainly to confirm the presence or absence
of disease in distant locales, early workers sampled
small numbers of armadillos at random locations and
often reported disparate prevalence rates. The size and
structure of animal populations vary in different environments, and can influence the observable prevalence
of disease. Louisiana is a lowland area at the mouth of
the Mississippi River. Among its most prominent features are the alluvial swamps and bottoms associated
with the Mississippi, Red, and Atchafalaya rivers, and
the generally similar low prairies and marshes that
form the coastal margin and extend through southern
Texas. This region juxtaposes mixed pine and deciduous forests that are easily identified in geologic soil
maps and satellite images of the region (Kniffen and
Hilliard 1988). A representative drawing of the region
is shown in figure 10.1.
There is a large reservoir of M. leprae among wild
armadillos in this lowland region. By 1986, based on
samples from some 3,500 animals, investigators had
described at least 136 armadillos with histopathologi-
116
R. W. Truman
Figure 10.1. Map of the southeastern United States by county indicating areas of incidence of and sampling for leprosy. Insert
shows general location in country. Dark line indicates the northern range of armadillos in the United States. Grey counties are
part of the alluvial and coastal lowlands. Stippled counties have had native-born residents presenting with leprosy. Numbers identify general locations where armadillos have been surveyed for leprosy. All locations are approximate: 1 = Carville,
Louisiana (LA); 2 = New Iberia, LA; 3 = Picayune, Mississippi (MS) and area near Kentwood, LA; 4 = Woodville, MS and Feliciana
parish, LA; 5 = Leesville, LA; 6 = Houston, Texas (TX); 7 = Corpus Christi, TX and Welder Wildlife Refuge; 8 = Luckenbach, TX; 9 =
area of College Station and Palestine, TX; 10 = Lawton, Oklahoma, Wichita Mountains National Wildlife Refuge; 11 = Clarksville, Arkansas (AR); 12 = Desha County, AR; 13 = Tallahassee, Florida and St. Marks National Wildlife Refuge; a, b, c = Louisiana
locations for systematic survey of armadillos infected with M. leprae, a = Tensas River National Wildlife Refuge, b = East Atchafalaya Management Area, c = Lacassine National Wildlife Refuge.
cally detectable M. leprae in their ear tissues, mainly

in the lowland areas of Louisiana and Texas (Smith et
al. 1983; Job et al. 1986; Walsh et al. 1986). A systematic survey among 565 animals spanning 4 locations
between north Louisiana and Corpus Christi, Texas,
confirmed the average 3.8% histopathological prevalence rate reported by others, and showed that about
16% of the armadillos populating the area also had detectable IgM antibodies to PGL-1 (Truman et al. 1991).
Repeated observations at the same location where
Roth (Roth et al. 1964) had taken armadillos during
his 19601964 survey showed a steady maintenance of
the infection from year to year, and confirmed that the
prevalence rate had not changed significantly over the
ensuing 30 years (Truman et al. 1991). M. leprae is intensely transmitted among armadillos in this region. In
one location, the incidence density estimate based on
PGL-1 IgM sero-conversion was 3.5 cases/1,000 animal
days (Paige et al. 2002). The antibody-positive animals
were almost entirely in the adult cohort. Using plasma

progesterone concentrations to age stratify the populations showed that the age-adjusted antibody prevalence
rate remained reliably consistent between different locations, and nearly a third of all the adult armadillos
surveyed appeared to harbor M. leprae (Truman et al.
1991). Therefore, M. leprae infection is really quite common among armadillos in these lowland areas; steady
maintenance of high rates of disease over a very large
geographic area suggests that they have harbored M.
leprae for many generations.
Armadillos are more difficult to acquire outside the
lowland areas, and most surveys in non-lowland regions
have had to rely on samples from road-killed animals
(Loughry and McDonough 1996) or commercial trappers. However, numerous results show that M. leprae
infection is significantly less common among armadillos in those locales. No infected armadillos have been
reported from among nearly 1,000 animals examined in
Leprosy
117
Table 10.2. Locations and Features of Wild Armadillos Examined for Naturally Acquired Leprosy

Number
Country
State
City
sampled
Argentina
Corrientes
Grenada
NA
United States
Arkansas

Florida

Louisiana

Mississippi

Oklahoma

Texas

Mercedes
St. Georges
Clarksville
Desha County
Sarasota
St. Marks NWR
Tallahassee
Kentwood
Deritter
Felecianas
Atchafalayac
Tensas NWRc
Lacassine NWRc
Woodville
Lawton
Luckenbach
Corpus Christi
83
25
102
42
26
37
142
145
13
135
550
77
78
54
67
86
35
Number
PGL-1 IgM
positivea
Number
histopathological
positiveb
2
0
0
9
0
0
0
9
1
3
85
18
16
1
0
0
6
0
0
0
ND
0
0
0
2
1
0
16
5
1
0
0
0
2
Notes: Armadillos were acquired by local trappers at various locations and examined for evidence of infection with M. leprae. Abbreviations: NA = not applicable, ND = not done, NWR = National Wildlife Refuge.
a. Elisa positive for IgM antibodies for PGL-1.
b. Histopathological examination of ear tissues detected acid-fast bacilli in dermal nerves.
c. Values presented for comparison purposes; prevalence previously reported in Truman (2005).
Florida (Kirchheimer and Storrs 1978a; Truman, Shannon et al. 1986; Walsh et al. 1986), and, similarly, we
found no evidence of the infection (0/67) among armadillos taken near the other end of their range around
Lawton, Oklahoma (table 10.2). In Texas, one infected
armadillo was reported from near College Station (Anderson 1978) and Walsh found one (1/61) infected animal among a group taken near Palestine, Texas (Walsh
et al. 1977). Otherwise, no evidence for leprosy has been
reported among 427 armadillos sampled >100 km from
the Texas coastal margin (Kirchheimer 1977; Wilson
et al. 1984; Walsh et al. 1986; Clark et al. 1987). While
many of these animals were examined only histopathologically, which is problematic because this method has
a lower likelihood of detecting the disease, at least one
group of animals from near Luckenbach, Texas, was
screened serologically and still revealed no evidence of
infection in 86 animals (table 10.2). A large survey of
ear tissue from 853 armadillos in Alabama, Arkansas,
Florida, Georgia, and Mississippi also failed to find any
histopathological evidence for M. leprae infection in
those areas (Howerth et al. 1990), nor did Walsh for armadillos taken near Picayune, Mississippi (0/178, Walsh
et al. 1977). One armadillo (1/40) with histopathologically detectable bacilli in its ear tissues was found near
Natchez, Mississippi (Fox et al. 1977), and we captured
one (1/54) PGL-1 IgM positive armadillo near Woodville, Mississippi (table 10.2). In addition, we found serological evidence for the infection (9/42) among armadillos in the eastern Arkansas lowlands around Desha
County, but not in armadillos from western Arkansas
near Clarksville (0/67, table 10.2). M. leprae infection is
found among Louisiana armadillos outside the lowland
areas, but the prevalence appears to be less. We captured one (1/12) PGL-1 IgM positive armadillo near Deritter, three (3/135) in the Feleciana parishes, and nine
(9/145) near Kentwood, Louisiana (table 10.2). Similar
low prevalence rates of disease may be found elsewhere.
A recent study along the Alabama/Mississippi border
found M. leprae infection in two (2/61) armadillos from
near Columbus, Mississippi, and one (1/70) from near
Jackson, Alabama (unpublished observations).
The distribution of the infection among armadillos
does not seem to be influenced by local variations in
susceptibility or disease type. Laboratory studies with
armadillos from all these locations show that they are
susceptible to experimental infection with M. leprae
and tend to react similarly to Lepromin-A. The lowland
areas appear to favor both the number of armadillos
and a high prevalence of infection. Animal density may
play some role in the observed prevalence of disease,
but measures indexing animal crowding may be more
118
R. W. Truman
reliable and systematic studies outside the lowland

region are needed to address the issue (Truman et al.
1991).
There have been few contemporary studies on armadillos from outside the United States. Pea (unpublished data) found 2/83 PGL-1 IgM positive armadillos
in northern Argentina (table 10.2), but was unable to
demonstrate presence of M. leprae in the tissues examined. We found no evidence for the infection (0/25)
among armadillos on the island of Grenada in the Caribbean (table 10.2). However, Deps recently reported
5/14 Brazilian armadillos sampled near Vitoria to be
positive by PCR for M. leprae in their blood (Deps et
al. 2002, 2004). Her results await confirmation, but armadillo leprosy is likely to be more common in other
countries than is currently recognized.
Clearly, armadillos in parts of Louisiana and Texas
are a large natural reservoir of M. leprae. Between the
two states, there are more than 60,000 square miles of
lowland habitats where we find a high prevalence of sylvan leprosy (figure 10.1). If armadillos utilize even 1% of
that space at a density of just 12 animals per acre, then
there are likely more than 100,000 M. leprae-infected
armadillos in Louisiana and Texas. The size of this reservoir alone suggests that armadillos may contribute
to some cases of human infection, but the impact they
have had on human health in the region has been difficult to discern.
Association with humans

Most cases of leprosy in the United States arise among
immigrants from endemic countries or nationals who
may have acquired their disease while living abroad.
However, there is persistent autochthonous transmission of the infection in Texas and Louisiana that gives
rise to 3040 cases each year among native citizens
(Enna et al. 1979; Joseph et al. 1985; Neill et al. 1985).
The counties of residence of those native-born cases,
as derived from the National Hansens Disease Registry
(available from the National Hansens Disease Program
[NHDP], Baton Rouge, Louisiana), show a clustering
within areas bordering the western crescent of the Gulf
of Mexico (figure 10.1).
Published case reports relate development of leprosy
among at least 13 individuals from Louisiana or Texas
who had no known exposure to M. leprae, other than
perhaps contact with the organism through armadillos. Though casual exposure also was implicated, these
people generally reported extensive direct contact with
armadillos, such as handling the animals, or preparing

and consuming their flesh (Freiberger and Fudenberg
1981; Lumpkin et al. 1983, 1984; West et al. 1988). However, in an early case-control study with 19 native-born
patients in Louisiana, Filice et al. (1977) found no association between contact with armadillos and the presence of leprosy in humans: both cases and controls reported no difference in the nature or frequency of their
contact with the animals.
Exposure to armadillos is quite common in this
region. About half of all the leprosy cases attending a
Houston clinic acknowledged some direct or indirect
exposure to the animals (Bruce et al. 2000). In another
case-control study among Mexican-born patients attending health clinics in Los Angeles, Thomas et al.
(1987) found a significantly increased risk for leprosy
among individuals who reported a history of contact
with armadillos in Mexico. Unfortunately, the geographical distribution and prevalence of leprosy among
Mexican armadillos has not been described, and it is
uncertain if the risk attributed in this study relates to
contact with M. leprae through armadillos or other factors.
Definitive conclusions about zoonotic transmission
of leprosy may come only with establishment of suitable molecular strain typing systems. A recent analysis based on single nucleotide polymorphisms (SNPs)
predicts that leprosy was carried to North America by
European immigrants and African slaves. Armadillos
too are relatively recent arrivals to the area. Based on
SNP data, the strain type of M. leprae seen among armadillos matches that of human settlers to the region
(Monot et al. 2005) and indicates that the animals must
have acquired M. leprae from humans at some point in
time. Both humans and armadillos have established a
persistent focus of infection in the U.S. region lining the
western Gulf of Mexico (figure 10.1). Deciphering the
factors that underlie the relationship between humans
and armadillos to M. leprae in this region could provide
significant new insights into leprosy transmission.
Conclusions
The nine-banded armadillo is uniquely important to
leprosy research as a biomedical model of the infection
and source of leprosy bacilli. The M. leprae made available through armadillos has been a boon to leprosy research and provided the resources necessary to develop
specific diagnostic antigens and vaccines for leprosy.
Armadillos are the only immunologically intact animal
Leprosy
that regularly develops lepromatous-leprosy. The specific factors that underlie their unique susceptibility remain important scientific questions and they likely will
be used in leprosy studies for many years to come.
Armadillos do not require specialized housing but
can be incorporated into many existing animal facilities. Their long life span (up to 12 years in the laboratory), cool body temperature, and unusual reproductive cycle can recommend them for a variety of studies
(e.g., Boily this volume). The main factors limiting their
use have been a paucity of specific biochemical or immunological reagents for use with armadillos, and our
inability to breed the animals in captivity. With the current availability of the total armadillo genome sequence
(Chang and Adams this volume), a vast array of biochemical and immunological reagents is now readily
available and will likely facilitate more extensive use of
armadillos in research studies.
Nine-banded armadillos in parts of the southern
United States are known to harbor M. leprae, but it is
unclear where or when the animals might first have acquired it. However, the infection appears to have evolved
by natural means and armadillos today support intense
transmission of M. leprae in the wild. The disease is
119
most common among armadillos in lowland habitats

and may be rare or absent elsewhere. Few studies have
addressed its occurrence among armadillos outside the
United States. Animal densities and local topographical
or environmental circumstances may influence detection of the disease in some locales. However, it seems
unlikely that the infection would be confined wholly to
a single nidus in the United States, so the issue merits
investigation elsewhere in the armadillo range.
Leprosy remains rare in the United States, yet exposure to armadillos is common. The impact that infected armadillos have on human health is difficult to
discern. Colonists, slaves, and armadillos were all relatively recent arrivals to this region, but both humans
and animals now show similarities in the geographical
distribution of their infections. Understanding the ecological relationship of humans and armadillos with M.
leprae in this region may come only with establishment
of more suitable molecular strain-typing methods. Additional studies with these animals could benefit our
basic understanding of leprosy transmission. Presently,
armadillos should be viewed as a potential source of
infection for our citizens.
11
The spleen of the armadillo
Lessons of organ adaptation
Emma B. Casanave and Elena J. Galndez
of the armadillo family, while D. hybridus belongs to

another branch, the Dasypodini (Delsuc et al. 2004).
These four species have been the primary subjects of
the studies we review here.
Like any species, armadillos have endured many
climatic, geologic, geographic, and ecological changes
through time and have adapted to them. Such adaptations include a flexible carapace, relatively low body
temperature (Casanave and Affanni 1994), and a number of interesting physiological features (Garca Samartino et al. 1987; Casanave and Polini 1999; Polini et al.
1999; Hernndez et al. 2001; Maldonado et al. 2002).
One particularly notable set of adaptations involves the
plasticity of their organ systems. It is in this context that
we focus on the spleen. In spite of the phylogenetic position of armadillos and their importance as biomedical
models (Maldonado and Casanave 1996; Bermdez et
al. 2004; Casanave et al. 2005, 2006), there are few studies about their physiopathology and immunology, and
most of these involve responses to leprous and parasitic diseases (Silva Vergara and Martnez 1999; Scollard 2000; Fernandes et al. 2004; Bagagli and Bosco
this volume; Truman this volume). However, as we
shall describe, information on spleen morphology and
function is broadening our understanding of armadillo
biology.
The spleen is an amazing organ that reflects the natural history of different vertebrate groups (Zapata and
Cooper 1990). It exists, as an extramural organ, from
Chondrichthyes to mammals (Galndez and Aggio
1998) and its location in the middle of systemic circulation contributes to monitoring many physiological parameters (Weiss 1989). Furthermore, as a good example
of a mesoderm-derived organ, it may adopt different
functions according to external or internal require-
Resumen
Los armadillos, miembros de un antiguo orden de
mamferos, presentan un bazo adaptable que es producto de una mezcla de caractersticas morfolgicas y
fisiolgicas relictuales y modernas. Esto resulta en un
rgano de almacenamiento que exhibe, adems, un
compartimiento inmunocompetente desarrollado y
tejido hemopoytico activo. La estructura esplnica se
discute en relacin con otros grupos de mamferos y en
un contexto biolgico.
Resumo
Tatus, membros de uma antiga ordem dos mamferos,
apresentam um bao adaptvel que o produto de uma
mistura de caractersticas morfolgicas e fisiolgicas
antigas e modernas, resultando em um rgo de armazenamento com um compartimento imunocompetente
bem desenvolvido e um tecido hematopoitico ativo.
Esta estrutura complexa discutida em relao com
outros grupos mamferos e em um contexto biolgico.
Introduction
Armadillos, family Dasypodidae, are conspicuous
members of Xenarthra, the earliest order of American
Eutheria (Delsuc et al. 2001; Delsuc and Douzery this
volume). Argentina is a very important reservoir of armadillos and in the southeast of Buenos Aires province
four species coexist: Chaetophractus villosus, Chaetophractus vellerosus, Zaedyus pichiy, and Dasypus hybridus (Casanave et al. 2003). Cha. villosus, Cha. vellerosus,
and Z. pichiy are closely related and together with the
genus Euphractus constitute the Euphractini branch
120
121
Table 11.1. Morphometric Data for the Spleens of Four Species of Armadillos
Species
Splenic index
Capsulo trabecular system
Mean
SD
CT (m)
Number of trabeculae
Cha. vellerosus
Cha. villosus
D. hybridus
Z. pichiy
0.502
0.373
0.081
0.288
0.104
0.043
nc
0.026
5
26
3
9
3060
3060
4070
4065
Abundant
Moderate
Very abundant
Abundant
Reference
4
1
2
3
Sources: (1) Galndez et al. 1997, (2) Galndez et al. 2000, (3) Galndez et al. 2003, (4) Galndez et al. 2006.
Notes: Splenic index = spleen weight/body weight 100; CT = capsule thickness; nc = not calculated.
ments, including a variety of health conditions (Freedman and Saunders 1985; Romanovsky and Petersen
2003).
The spleen of armadillos

The armadillo spleen shows a mixture of structural and
physiological features that make it an exceptional example of a malleable organ; some of these characteristics are: (1) it is a storage or accumulation spleen
(types III or IV, Hartwig and Hartwig 1985); (2) it is
hemopoietically active in normal adults, at least in
some species (Oria 1941; Weiss and Wislocki 1956; Cuba
Caparo 1979; Galndez et al. 1997, 2000, 2003, 2006); (3)
it has well-developed immunological tissue (Purtilo et
al. 1975; Diniz et al. 1997; Scollard 2000; Cheadle et al.
2001); and (4) it has important hemocatheretic activity
(Weiss and Wislocki 1956; Galndez et al. 1997, 2000,
2003, 2006).
Anatomically, the armadillo spleen shows the typical

segregation between pulps and an extensive capsulotrabecular system (figure 11.1).
Interestingly, armadillos possess one of the smallest spleens among mammals (table 11.1). In fact, their
spleens are smaller than those of other species, including
some marine mammals (Cabanac et al. 1999), that are
also exposed to low O2 pressure. This suggests no direct
relationship between fossorial lifestyle and spleen size.
Armadillos may be able to survive the hypoxic environment of burrows and still have small spleens because of
their special management of low O2 pressures (Affanni
et al. 1987; Casanave and Affanni 1995; Casanave et al.
1995).
The white pulp, the immunocompetent tissue of the
spleen, does not show marked interspecific differences
(Purtilo et al. 1975; Cuba Caparo 1979; Galndez et al.
1997, 2000, 2003, 2006). In the four species we have
studied, Cha. vellerosus shows the largest and best-or-
Figure 11.1. Low magnification view

of the splenic parenchyma of Zaedyus
pichiy. Abbreviations: wp = white
pulp, vt = vascular trabeculae.
122
E. B. Casanave and E. J. Galndez
Figure 11.2. Transmission

electron microscopy image
of an interdigitant-like cell
from the marginal zone of
Chaetophractus vellerosus.
Arrow: plasmocyte, arrow
heads: plasmoblasts, asterisk: cytoplasmic projections
of the cell.
ganized nodules, while D. hybridus has the least defined

structure (Galndez et al. 2000, 2006). Nonetheless, all
have periarterial sheaths, lymphocyte cords, and nodules with germinal centers, plus plenty of amorphous
material positive to the periodic acid Schiff reaction
(PAS +).
Not only is spleen structure similar among armadillos, it is consistent with other Metatheria and Eutheria (Hayes 1968a; Coutinho et al. 1995; Agar 1996). For
example, within the white pulp, transmission electron
microscopy reveals cells that are morphologically similar to follicular dendritic cells, as well as cells with the
structural characteristics of interdigitant cells (figure
11.2). The presence of follicular-like cells is common in
other mammals and vertebrates (Gallego et al. 1995; Zapata and Cooper 1990) and has been used as evidence
that a mechanism for the presentation of antigens was
conserved in vertebrates (Cisternas and Armati 2000).
Thus, the splenic construction exhibited by armadillos
appears to be a basal condition in vertebrates, in agreement with the proposal that structural phenotypes are
conserved among evolutionary clades (Carleton 1996).
Other aspects of armadillo spleen anatomy are more
variable. The transitional zone between pulps (figure
11.3) is not a marginal zone as in most modern mammals (Schmidt et al. 1993; Aichele et al. 2003). Rather, it
is an intermediate zone, defined as a network of vascular spaces and a mixture of cells, as seen in species such
as the platypus (Tanaka 1986, 1990) and insectivores
(Tanaka 1987, 1990). This region is occupied by ellipsoidal sheaths and a resident population of Perls-positive macrophages. The Schweigger-Seidel sheaths vary
in number and organization between species: in D. hybridus ellipsoids are less abundant and less structurally
defined; Z. pichiy shows the opposite (Galndez et al.
2000, 2003). Ellipsoids have been considered the site
for splenic clearance (Blue and Weiss 1981). However,
the interspecific differences we have documented may
not be directly related to clearance functions, but may
instead reflect the complexity of the intermediate zone
in each species.
In neonatal armadillos the spleen shows the typical
characteristics of the basic spleen model. However,
a subcapsular ring of hemopoietically active tissue is
present, and there is no clearly organized white pulp or
trabecular system (Galndez et al. 1997). This shows not
only that splenic organization can be altered evolutionarily (as with the transitional zone between pulps), but
also that interactions with the environment can reorganize the tissues for better performance.
Development of and changes in organization of
white pulp are most likely a direct response to exposure
to antigens. What might cause changes in other splenic
tissues? Microanatomically, the red pulp is a mixture
of tissues that can be easily remodelled according to
needs. Thus, the question becomes: if armadillos are
born with the basic structure, what promotes the dramatic changes from this original pattern in the adult?
Figure 11.3. Transmission electron microscopy image of the

transitional zone of Chaetophractus vellerosus. Abbreviations: wp = white pulp, mz = intermediate zone, rp = red
pulp.
123
Unfortunately, we cannot answer that question at present.

Table 11.2 summarizes what we do know about the
red pulp for our four species of armadillos. Because of
its organization, red pulp plays an important role in the
management of blood flow, O2 supply, thermal regulation, and clearance of parasites (Tablin et al. 2002). Evidence of modification is illustrated by the presence of
contractile cells interposed between the vascular beds
(figure 11.4) and the development of a system of closed
circulation (figure 11.5) as a complement to the open
vascular spaces to allow for slower blood flow.
A final noteworthy feature of armadillo spleens concerns the production of blood components. The hemopoietic prerequisite for most adult mammals is to
confine myeloid tissue to the bone marrow. Some animals such as the Prototheria (Tanaka 1986), Insectivora
(Tanaka 1987), some Xenarthra (Oria 1941), and several
Rodentia (Aggio 1987) show myeloid noncompensatory activity in healthy adult spleens, in addition to an
active hemopoietic bone marrow. Not only do armadillos show hemopoietic activity in the bone marrow and
spleen, but, in at least one case, activity is present inside
the dermal bones of the carapace too (Weiss and Wislocky 1956), albeit only seasonally. This nontraditional
location in mesoderm-derived structures supports a
nomadic model of hemopoiesis and makes armadillos an excellent experimental system for the study of
phenomena such as nidation and commitment of the
hemopoietic stem cell.
Table 11.2. Comparative Characteristics of the Red Pulp of Armadillo Spleens

Characteristic
Cha. vellerosus
Cha. villosus
Smooth muscle cells

or myofibroblasts?
Not present
Scarce
Vascular spaces

Splenic sinus?
Yes
Yes

Open vascular spaces limited by
Yes
Yes

reticular cells and/or macrophages?

Other features?
None
None

Hemopoiesis
Yes
Yes

Madurative proliferative compartment

of myeloid lines?

Reference
4
1
2
Note: Numbered references are the same as those in Table 11.1.
D. hybridus
Z. pichiy
Not present
Abundant
No
No
No
Yes
Postcapillary
venules with a very
thick basal membrane
Direct anastomosis
between small vessels
and medium-sized veins
Erythroid and
thrombocytoid lines
only
3
Yes
Figure 11.4. Scanning electron microscopy image of the red pulp of Zaedyus pichiy. Reticular
cells define vascular spaces.
Figure 11.5. Transmission electron microscopy image of the sinus wall of the spleen of Chaetophractus vellerosus and the typical basal membrane conformation of splenic sinuses (arrows).
Evolution of the armadillo spleen

As stated above, armadillo spleens can be classified as
type III or IV according to Hartwig and Hartwig (1985)
or, more precisely, as storage spleens. Boily (2002) hypothesized that the low metabolic rate of armadillos
(McNab 1980) evolved partly because a low risk of predation (due to the carapace) made selection for aerobic capacity less important than the need for energy
conservation. In this context a storage spleen (even
a small one) such as seen in armadillos might be expected. All armadillos are fossorial (although the extent
of this varies widely) and the environment of burrows is
clearly hypoxic. Thus, a storage spleen clearly would be
adaptive. In the species we have studied, spleen structure seems related to burrow use. Cha. vellerosus has
a well-defined splenic organization suited for storage.
This species spends considerable time in small, sloped
burrows (Greegor 1985; Abba and Cassini this volume).
In contrast, D. hybridus has a simpler splenic structure,
even more so than other members of its genus (e.g., D.
novemcinctus Hayes 1968b, 1970). The absence of a welldefined storage structure may explain why this species
seems to spend, at least in our region, more time on the
surface (Casanave et al. 2003).
The preceding scenario may explain facets of spleen
evolution within armadillos, but broader questions re-
125
main. For example, a question we have yet to answer is

why the more recently derived Artiodactyla and Perissodactyla have the same splenic structure as armadillos
(Hartwig and Hartwig 1985), which are presumed to be
near the base of the placental tree (Delsuc and Douzery
this volume).
Final remarks
Armadillos, as ancient mammals, show a mixture of
physiological and morphological characteristics that
depict paradoxes between phylogeny and adaptation.
This is amply illustrated in the spleen, which is clearly
a combination of relict and modern features. We conclude by urging further study and interpretation of this
important organ, including all its functions, in the hope
that these analyses may illuminate many aspects of the
biology and phylogeny of xenarthrans.
Acknowledgments
This study was supported by grants from the Secretara
General de Ciencia y Tecnologa, Universidad Nacional del Sur, SGCyT-UNS, Project 24/B086 and 24/B122,
and Agencia Nacional de Promocin Cientfica y Tecnolgica, ANPCYT, PICTR 074/02.
12
The use of armadillo clones from
the genus Dasypus as experimental models to
investigate the source of physiological variability
Patrice Boily
August Krogh, a Nobel Prize laureate who conducted

pioneering physiological studies in the late nineteenth
to early twentieth centuries, owed much of his success
to his recognition of the importance of choosing the
right animal model for a given study. Accordingly, the
notion that for many problems in biology there is an organism on which it can be most conveniently studied is
usually referred to as the August Krogh principle. This
principle continues to profoundly influence selection
of study organisms for a wide range of research areas,
from the genetic engineering of thermal adaptations
in prokaryotes (Bennett 2003) to the use of naturally
existing variation to investigate the mechanisms of animal development (Burggren 1999).
The selection of the most convenient organism
depends on a number of factors. First and foremost,
it should exhibit traits that make it particularly well
suited for the specific study. For field studies, it should
be abundant in the wild and easy to observe or capture.
For laboratory studies, it should be easy and inexpensive to acquire (from commercial suppliers or wild populations) and to maintain in captivity. In addition to
these practical considerations, ethical concerns must be
weighed, such as the threatened or endangered status of
species and the possibility of using lower organisms,
cell cultures, or even computer simulations to acquire
the same information.
Compared with small rodents, which are by far the
most widely used mammalian research models, armadillos are neither easy nor inexpensive to care for in
captivity (Superina et al. this volume). However, many
species of armadillos are abundant, relatively easy to
capture, and, most importantly, equipped with highly
unusual traits that make them especially well suited
for physiological studies. Consider the following ex-
Resumen
Los armadillos del gnero Dasypus tienen la propiedad exclusiva entre los vertebrados de producir continuamente camadas de hermanos monocigticos. Esta
produccin natural de clones puede ser un importante
recurso para muchos campos de investigacin. En este
captulo se discute el uso de Dasypus como un modelo
experimental para dividir el componente gentico de
la variabilidad intraespecfica en distintos rasgos fisiolgicos.
Resumo
Os tatus do gnero Dasypus possuem a propriedade mpar entre os vertebrados de produzirem continuamente
proles de gmeos monozigticos. Essa produo natural de clones genticos pode ser uma fonte de alto valor
para muitos campos de pesquisa. Esse captulo discute
o uso de Dasypus como modelo experimental na diviso do componente gentico da variabilidade intraespecfica em distintos aspectos fisiolgicos.
Introduction
For a large number of problems there will be some animal
of choice or a few such animals on which it can be most
conveniently studied. . . . we used to say as a laboratory joke
that this animal had been created expressly for the purposes
of respiration physiology. I have no doubt that there is quite
a number of animals which are similarly created for special
physiological purposes, but I am afraid that most of them are
unknown to the men for whom they were created and we
must apply to the zoologists to find them and lay our hands
on them.
August Krogh (1929)
126
The use of armadillo clones from the genus Dasypus as experimental models
amples: first, armadillos of the genus Dasypus are the

only mammals, other than humans, that can be systemically infected with Mycobacterium leprae (mice can
be infected, but only on their footpads). Accordingly,
Dasypus species, especially D. novemcinctus, have been
used extensively in Hansens disease (leprosy) research
(see Truman this volume). Second, armadillos have a
semirigid carapace that provides protection against
predators but limits the capacity to increase lung tidal
volume. Thus, they are excellent models for investigating the effects of body armor on the evolution of physiological traits (Lovegrove 2001) and the mechanical
consequences of a stiff thorax on breathing patterns
(Frappell et al. 1998). Third, armadillos spend considerable time, even while giving birth, in burrows where
air quality can be poor. They are therefore good models
for investigating the effects of hypoxia and hypercapnia on breathing patterns and metabolic rates (Boggs
et al. 1998; McNab 1979). Fourth, armadillos have relatively low and variable metabolic rates and body temperatures, and relatively high thermal conductance.
As a consequence, they represent an ideal system for
investigating the selective forces that led to the evolution of endothermy (Boily 2002) as well as the potential
effects of physiological traits on the global distribution
of mammals (McNab 1980). Finally, as do a handful of
other mammals (Lovegrove et al. 1991; Klir et al. 1990),
some armadillos (at least D. novemcinctus) exhibit a
cold-induced fever: their body temperature undergoes
a controlled increase, by up to 3.5C, during cold exposure (Johansen 1961; Mercer and Hammel 1989). D.
novemcinctus can therefore be a good model for investigating the costs and benefits of maintaining a variable
or stable body temperature during cold exposure (Boily
and Knight 2004).
Clones
With the exception of M. leprae infection, the traits described above apply to most armadillo species, and in
fact to many other mammals as well. However, armadillos of the genus Dasypus have a reproductive trait that
is unique among vertebrates: the production of clonal
siblings. Indeed, although the clonal nature of siblings
has been genetically confirmed only in D. novemcinctus
(Prodhl et al. 1996; see also Prodhl et al. this volume),
females of all Dasypus species are presumed to produce
genetically identical siblings that range in number from
two in D. kappleri to as many as twelve in D. hybridus
(Hayssen et al. 1993).
127
Clonal siblings are valuable as a research tool because they potentially allow partitioning of the genetic
and environmental components of phenotypic variability; that is, resolving the question of relative contributions of nature vs. nurture to the traits exhibited by
individuals. Partitioning the genetic component of phenotypic traits is important for addressing fundamental
and applied biological problems such as the selective
forces that led to the evolution of adaptations (Garland
and Carter 1994) or the degree of genetic contribution
to an individuals predisposition to disease (Bouchard
et al. 1997).
Genetic heritability of traits can be tested using parent-offspring correlations or artificial selection experiments with captive animal colonies. These methods,
to be done effectively, require expensive and labor-intensive long-term studies, especially when higher vertebrates with long generation times are used (Garland
and Carter 1994). Because clonal siblings have identical
genotypes, a potentially simpler method to assess the
genetic heritability of traits is to compare phenotypic
variation within and among litters, an experimental approach that is often used in humans (Bouchard et al.
1997). Such studies are limited by the rarity of twins and
the ethical limits on what can be done experimentally
with human subjects.
Although somatic cell nuclear transfer technology
has allowed the artificial production of clones in a wide
range of mammalian species during the last decade,
this method is still highly inefficient, plagued with
phenotypic abnormalities (Li et al. 2003) and inapplicable to natural populations. The regular and natural
production of clonal siblings by Dasypus therefore has
the potential to be a highly valuable research resource
for both basic and applied biological studies. For example, for basic studies, Dasypus species are ideal models in which to partition the genetic and environmental factors producing phenotypic variation in natural
populations. For applied studies, especially biomedical,
Dasypus species have the potential to be the most easily
acquired source of mammalian clones. There are, however, fundamental and technical considerations that
may limit the usefulness of Dasypus as a model.
Maternal and environmental effects

One of the major hurdles to overcome in genetic heritability studies is to properly account for maternal and
environmental effects. Indeed, clonal siblings may have
similar phenotypic traits independent of genetic ef-
128
P. Boily
fects because they were exposed to nearly identical in

utero and postpartum environmental conditions. This
potential problem was recognized more than 130 years
ago, close to a century before the discovery of DNA,
when Galton (1875) discussed the use of twins to investigate the nature of inherited traits in humans. A
typical recent example of the problem is the study by
Bagatto et al. (2000), who used wild-caught pregnant
D. novemcinctus females acclimated to captive conditions to investigate the genetic component of variability
in physiological traits. The use of offspring from pregnant females acclimated to standardized laboratory
environmental conditions minimized potentially confounding maternal and environmental effects, but did
not eliminate them. They were not eliminated because
the possibility remained that some of the lower withinlitter variability in physiological traits compared with
that between litters resulted from maternal effects; that
is, some females may have been in better condition
than others, resulting in higher quality and quantity of
nutrient delivery in utero or through milk, with the result that some litters were in consistently better condition and with higher metabolic rates (and by extension
higher ventilation and heart rates) than others.
To properly account for maternal and environmental effects, the most robust studies using human twins
to assess genetic heritability of traits (Bouchard et al.
1997) compare the variability in phenotypic traits of
monozygotic (identical) twins raised together (identical genotypes, similar in utero and postpartum environments) with that of monozygotic twins raised apart,
usually as a result of adoptions (identical genotypes,
similar in utero environment, but different postpartum
environment), and of dizygotic (fraternal) twins (different genotype, similar in utero and postpartum environments).
This approach could, to some extent, also be used
with Dasypus. Newborns of D. novemcinctus can be
raised successfully by bottle-feeding (F. M. Knight personal communication), thus eliminating potential differences in milk quality and quantity. Although Dasypus females are not known to ever give birth to fraternal
offspring within a given litter, comparing different litters
produced by the same females is similar to comparing
monozygotic and fraternal twins from different mothers, assuming that females do not vary their maternal
investment with age. Unfortunately, captive breeding
of Dasypus species is difficult and has been successful
only when the animals were kept in semicaptive conditions (Ferrari et al. 1997; see also Truman this volume
and Superina et al. this volume), conditions that require

outdoor space and human and financial resources that
are not available to all laboratories. Combined with this
technical hurdle is the fact that Dasypus females give
birth to one litter per year, making such an experimental approach potentially long and arduous.
Field sampling experimental approach

In areas where a Dasypus species is relatively abundant
and easy to capture, it is possible to conduct experiments on a large number of wild, free-ranging individuals held captive for a very short time, just long enough
to perform the desired measurements. Microsatellite
genetic markers can then be used to identify clones
within the sampled individuals, and comparison of
variability within and between clonal siblings can be
used to estimate the genetic contribution to this variability. By restricting the sampling to adult, fully grown
individuals, any in utero and postpartum confounding
effects are presumed to be diluted by the prolonged
exposure of each individual to different environmental conditions. In other words, because all individuals
within the population are presumably exposed to similar, and similarly variable, environmental conditions,
a lower phenotypic variability between clonal siblings
compared with the general population should reflect
genetic effects.
This approach has been used to investigate whether
the genetic uniformity of D. novemcinctus clonal littermates translates into a high level of phenotypic similarity in morphological traits, and whether the level of
phenotypic variability changes with age (Loughry and
McDonough 2002). A major limitation of this field sampling method is that it relies on chance to capture and
measure enough clonal siblings. Indeed, this approach
was also attempted by Boily et al. (2000) to investigate
the genetic contribution to variation in basal and maximal metabolic rates in D. novemcinctus, but the number
of clones sampled was too small to allow valid statistical
analyses. An additional limitation of this method is that
it does not completely eliminate environmental effects.
Low levels of within-litter phenotypic variability could
reflect habitat or diet similarities if littermates spend
most of their time in similar microhabitats and feed on
similar prey. This potential limitation is not merely hypothetical; as Prodhl et al. (1996) demonstrated, within
a population of D. novemcinctus in Florida, USA, clonal
juveniles (<1 yr old) were captured on average within
55 m of each other, and clonal adults, although more
The use of armadillo clones from the genus Dasypus as experimental models
dispersed, were still captured on average within 500 m

of each other.
Conclusion: overall usefulness of Dasypus as

models
Clones are widely recognized as highly valuable research resources for a variety of fundamental and applied biological studies. Since the widely publicized
birth of Dolly, the sheep cloned from adult somatic
cells in 1996, scientists have developed artificial methods to clone a wide variety of mammalian species. This
artificial approach is, however, still plagued with low
efficiencies and developmental abnormalities (Li et al.
2003). Although artificial cloning of mammals may
129
eventually become an effective way to obtain genetically identical research mammals, it is unlikely ever to
be applicable to the study of wild, natural populations.
Armadillos of the genus Dasypus are the only known
vertebrates to naturally and regularly produce clonal
siblings. Accordingly, as long as the potential pitfalls
and drawbacks associated with using Dasypus species are recognized and dealt with, these animals can
be highly valuable research subjects, especially for research on wild, natural populations. Considering that,
so far, only two (D. novemcintus and D. hybridus) of
the seven Dasypus species have been extensively studied and used as models, Dasypus species are an existing
natural resource whose usefulness has yet to be tapped
to its full potential.
13
The physiology of two- and three-toed sloths
Desmond Gilmore, Denia Fittipaldi Duarte, and Carlos Peres da Costa
perodo gestacional. O envolvimento dos dois dura um

ano ou mais. A maturidade sexual alcanada aos trs
ou quatro anos de idade. A capacidade das preguias
sobreviverem sob uma dieta insuficiente, em relao a
outros folvoros, combinada a um baixo dispndio energtico, pela permanncia em repouso ou dormindo
durante dois teros do dia, so apenas alguns fatores
importantes que garantem sua sobrevivncia.
Resumen
Los perezosos vivientes son mamferos que se encuentran, en distintas densidades, en las selvas de Amrica
tropical. Presentan numerosas adaptaciones para protegerse de los depredadores, incluyendo un excelente camuflaje y la necesidad de descender al suelo de la selva
a orinar y defecar slo una vez a la semana. Muchos
aspectos de la fisiologa de los perezosos son extremadamente especializados, incluyendo el funcionamiento
de los msculos, la actividad motora, cardiovascular,
respiratoria, digestiva y el sistema nervioso central.
Slo la madre se encarga del cuidado de la nica cra
que nace luego de un largo perodo de gestacin y ambos permanecen vinculados durante un ao o ms. La
madurez sexual se adquiere entre los tres o cuatro aos
de edad. La capacidad de los perezosos de sobrevivir
con una dieta insuficiente con relacin a otros folvoros, combinada con un bajo gasto energtico por descansar y dormir durante ms de dos tercios del da, son
slo algunos de los importantes factores que garantizan
su supervivencia.
Introduction
Present-day sloths inhabit comparatively narrow niches
in the environments of tropical Central and South
America. The two-toed sloths are represented by Choloepus didactylus and C. hoffmanni and the three-toed
by Bradypus tridactylus, B. variegatus, the endangered
B. torquatus, and the recently discovered B. pygmaeus
(Aguiar and Fonseca this volume; Chiarello this volume). This last species is a small endemic found only
on the islands of Bocas del Toro off western Panama
and, at present, little is known about it (Anderson and
Handley 2001, 2002).
The ecology of sloths is, in many respects, a reflection
of their physiology. Unlike most mammals, the body
temperature of sloths varies according to the surroundings, from 32.7 to 35.5 C (slightly higher in two-toed
sloths), and the animals are therefore unable to tolerate
temperate latitudes (McNab 1985). Body temperature
falls throughout the cooler hours at night, during wet
weather, and when the animals are inactive. Such labile
temperatures help to conserve energy. Goffart (1971)
argued that sloths may resist cold poorly because their
relatively small muscle mass is unable to produce sufficient heat to maintain constant body temperature.
Shortly after daybreak sloths show minimal activ-
Resumo
As preguias so mamferos encontrados nas florestas
tropicais da Amrica. Para se protegerem de predadores
desenvolveram algumas adaptaes, as quais incluem
excelente camuflagem, e a descida das rvores para o
cho, para urinar ou defecar, apenas uma vez por semana. Apresentam aspectos fisiolgicos altamente especializados na funo muscular e na atividade motora
e nos sistemas cardiovascular, respiratrio, digestivo e
nervoso central. Somente a me se envolve na criao
do filhote que, geralmente, nico e nascido aps longo
130
ity and can be found high in the tree canopy warming

themselves in the early morning sunshine. However,
this is also the usual hunting time for the Harpy eagle,
one of the few natural predators of the sloth (see Izor
1985). Other predators include the anaconda, jaguar,
margay cat, and probably the ocelot (Goffart 1971).
Nevertheless, sloths tend to survive well because of
their excellent camouflage and the fact that they move
little and then only slowly. Feces and urine are passed
only every four to six days at habitual sites at the bases
of trees. This retention of waste products reduces the
need to descend regularly to defecate and urinate and
so minimizes exposure to predators.
The remainder of this chapter is an overview of various aspects of sloth anatomy and physiology. As such, it
builds on the earlier efforts of Goffart (1971), who drew
together much of the literature on sloths published over
the past century or so, and also on the more recent updates on sloth biology provided by Gilmore et al. (2000,
2001). For each topic, we attempt to describe our current understanding of both mechanistic and functional
aspects.
Hair
Sloths possess an outer coat of long coarse hair that
provides their distinctive color and has extremely good
thermal-insulating properties (Goffart 1971). Color
patterns vary between the sexes and geographically in
Bradypus (Goffart 1971; Chiarello this volume). Interestingly, the abdominal hair parts in the middle, flowing
outward, which permits efficient runoff of water from
the abdomen when the animal is hanging upside down.
Through the use of scanning electron microscopy it was
found that the intricate cuticular scales in the two-toed
sloths hair are interrupted by grooves (Wujek and Cocuzza 1986). Freeze-fractured sections indicate that the
cuticle is not continuous around the shaft. In Bradypus,
on the other hand, the hairs are more complex: scattered shallow air pockets occur beneath the cuticle and
the cuticle does not extend unbroken along the shaft.
Wujek and Cocuzza (1986) argued that these differences are another justification for placing the two genera in separate families.
A particularly fascinating feature of the outer coat of
most sloths is the presence during certain times of the
year of symbiotic algae representing several phyla (Goffart 1971, Thompson 1972). During long rainy periods
the coat becomes distinctly green as a result of their
presence. It has been suggested that they give camou-
131
flage to the animals while procuring shelter for themselves (Aiello 1985). Interestingly, sloths of the Atlantic
forest rarely show this greenish coloration, perhaps because of less rainfall and lower temperatures there relative to Central American and Amazon regions (A. G.
Chiarello personal communication). In two-toed sloths
the algae lie longitudinally along grooves in the hairs,
but are found in short lateral lines in three-toed sloths.
They do not appear in the hair of juveniles until the
young are approximately two months old (Aiello 1985).
Moths of the subfamily Chrysauginae spend their lives
as adults in the fur of sloths, particularly in the threetoed species (Waage and Best 1985). It appears these
moths may receive some protection from avian predators and possibly find nutrients in secretions of the
sloths skin and/or the algae present on the fur (Wolda
1985). Larval stages of the moths live in and feed on the
sloths dung.
Brain and EEG activity

A detailed description of the anatomy of the sloth brain
has been provided by Goffart (1971); it is slightly larger
in Choloepus than in Bradypus. The outer surface is
characterized by a number of sulci of fairly constant
shape. There is enormous development of the frontal
sinuses, which extend dorsally over the anterior section
of the brain and possess openings into the nasal cavity. The olfactory bulbs are extremely well developed,
whereas those regions of the cortex concerned with
touch are not. The cerebral hemispheres do not extend
over the cerebellum, which is considerably larger in
Choloepus than in Bradypus. An investigation into sensory and motor representation in the cerebral cortex
of Bradypus by Saraiva and Magalhes-Castro (1975)
indicated that the area devoted to the representation of
the forelimb, and especially the forepaw, is considerably
larger than that for remaining body parts. The cortical
area for the forelimb is followed in decreasing order of
magnitude by the projections of the hindlimb, trunk,
and face. Visual and auditory projections occupy only
restricted areas on the caudal dorsolateral surface of the
neocortex.
In many animals the anterior lobe of the cerebellum
exerts a strong tonic inhibitory influence over brain
stem mechanisms regulating posture and muscle tone.
Its removal in the cat and dog results in a marked lack
of coordination (e.g., oscillatory head and body tremor
and truncal ataxia) that is followed by a phase of stabilized deficit during which significant compensation
132
D. Gilmore, D. F. Duarte, and C. Peres da Costa
takes place (Murphy and OLeary 1973). These authors

examined hanging and climbing ability in C. hoffmanni
and B. tridactylus after cerebellectomy and observed
no differences at all from intact animals. Murphy and
OLeary (1973) argued that because sloths move so
slowly, minimal equilibratory and altitudinal controls
are required. Thus, one does not see the typical effects
associated with cerebellectomy.
EEG recordings by Galvo de Moura Filho et al.
(1983) were utilized to investigate various types of behavior in B. variegatus. They found that sloths were
awake only about 30% of the time. Behavior was categorized as awake-exploring, awake-fixating, and awakealert (A1), or awake but relaxed and drowsy (A2). Light
sleep constituted 56% and deep sleep about 10% of the
recordings. Light sleep was subdivided into S1 and S2,
deep sleep into S3 and S4, and paradoxical sleep was
categorized as S5. Paradoxical sleep was found to occupy about 11% of total sleep time. Both sleep and
awake EEG patterns were considered comparable to
those seen in other mammals with a more differentiated cortical cytoarchitecture. For example, sloths lack
a well-developed, relatively pure, high amplitude alpha
rhythm, just as do most other mammals except primates. Low-voltage activity was seen in this frequency
range, mixed with other frequencies during A2 and S1.
True K-complexes were lacking, although vertex sharp
waves were prominent in S2 and S3. The sleep spindles
(67/sec) in sloths were of a lower frequency than those
in humans, cats, guinea pigs, and opossums, but only a
little less than those recorded in several species of rodents.
Hearing
Sloths can be aroused by sound, but only when motivated, for example when a separated young is calling
for its mother (Goffart 1971). Adult sloths rarely vocalize, except in extreme danger, when they make a sound
like a long, shrill, penetrating whistle (Montgomery
and Sunquist 1974). The onomatopoetic native names
for the sloth of ai-ai in Tupi or a-a in Guran are
based on these calls (Britton 1941). In some locations,
such as the Atlantic forest of Brazil, local inhabitants report that sloths call at certain times of the year to attract
other sloths (A. G. Chiarello personal communication).
However, while it remains possible that this occurs on
occasion, such calling has not been observed in more
than 1,300 hours of field time (A. G. Chiarello personal

communication).
The sloth ear appears designed for the detection
of low-frequency sounds. Ramprashad et al. (1985)
showed that the cochlea of two-toed sloths lacks a welldeveloped secondary spiral osseous lamina (a characteristic feature of high-frequency hearing in mammals)
within the basal coil. In Choloepus this structure is
present only at the round window. Ramprashad et al.
(1985) also found that the density of bipolar ganglion
cells in the cochlea falls from 1,300 cells/mm2 at the
base to 800 cells/mm2 3 mm from the basal end. The
density then increases gradually to a maximum of 1,600
cells/mm2 13.5 mm from the basal end. This maximum
density occurs within the region associated with lowfrequency vibration. Sloths had the smallest number of
bipolar ganglion cells among a wide number of mammals studied. Total number of ganglion cells is related
to the frequency range of hearing in mammals, with
those mammals utilizing high frequencies having the
greatest number of ganglion cells (Ramprashad et al.
1985).
Consistent with the above, the hearing range of
sloths has been reported as 0.330 KHz, although in
both two- and three-toed sloths the best frequencies
for activating the adult ear were 28 KHz (Suga 1967).
Peterson and Heaton (1968) demonstrated that the peripheral auditory function for C. hoffmanni was optimized for 0.53 KHz. Even so, they concluded that peripheral auditory functioning in this species, as well as
in tamanduas and armadillos, was less efficient than in
guinea pigs for the entire range of frequencies tested.
Vocalizations of young sloths may reflect constraints
due to the design of the adult cochlea, as well as ecological influences. Young three-toed sloths, when separated from their mother, will emit an almost pure-toned
whistle that rises from 1.9 to 2.6 KHz and continues for
two or three seconds (Montgomery and Sunquist 1974).
The distress call of sympatric young two-toed sloths is
a lower-pitched bleat, with most of the energy concentrated in three bands below 1.4 Khz (Montgomery and
Sunquist 1974). While the overall frequency range of
the calls is undoubtedly shaped by the auditory tuning of adult ears, differences between the species may
reflect selection for reduction of interspecific ambiguity
and/or optimization of sound propagation in the forest
canopy, where most adult activity occurs (Montgomery
and Sunquist 1974).
Vision
It has been argued that visual acuity is a requirement
for successful adaptation to life in the forest canopy.
However, because sloths have no ciliary muscles, they
cannot accommodate for near vision and are thus extremely shortsighted (see Goffart 1971). The animals are
also handicapped in having a very convex cornea and a
thick lens, allowing little power of discrimination (Goffart 1971). According to Beebe (1926), the eyesight of the
young is better than that of adults. Sloth eyes are very
mobile and it appears that they can be partly retracted
when the lids are tightly closed. The pupils are small and
the irises are of a deep tawny color. The eyes are conspicuous because they are circled by black hair. Twotoed sloths have been reported to wink frequently and
slowly, often with just one eye at a time. When stressed
or in an aggressive situation they may look pop-eyed,
as the eyes appear to rise from their orbits almost as if
they are being forced forward (Meritt 1985a).
Piggins and Muntz (1985) utilized ophthalmoscopy
to determine pupil size and shape, examine the fundus,
and estimate the size of the visual fields and position
of the optic axes in B. tridactylus. Their findings suggest that the eyes of these sloths are primarily adapted
for vision at low light intensities. They estimated the
included angle between the diverging optic axis of each
eye to be 50, horizontal monocular and binocular visual fields to be 70, and binocular overlap, 35.
Goffart (1971) reported the eyes of Bradypus to be
some 30 mm apart (4547 mm in Choloepus), relatively
frontal-looking, of moderate size, and with a wide and
convex cornea. The eyeballs were 1012.5 mm in diameter. Piggins and Muntz (1985) examined the eyes of B.
tridactylus in more detail. The eyes were nearly circular
in anterior-posterior section, with overall dimensions
of 8.4 mm axially and 8.3 mm equatorially. The radius
of the thick rounded crystalline lens was 2.35 mm at
its anterior and 1.84 mm at its posterior surface. The
equatorial:axial length ratio of 1.36 was less than the
1.70 reported for C. hoffmanni by Goffart (1971). Piggins
and Muntz (1985) suggested that, because of this higher
ratio, the lens of Choloepus was probably less powerful
than that of Bradypus. Furthermore, the lens was not as
thick and appeared less round.
The retinal histology of Bradypus is similar to that
reported by Goffart (1971) for Choloepus, that is, many
slender rods, few ganglion cells, and cones rare or
133
absent. Consistent with the evidence of Piggins and

Muntz (1985) described above, the distribution of photoreceptors indicates that sloths are better adapted for
nocturnal vision. However, because the pupil is highly
contractile, sloths may be quite capable of seeing well in
daylight too (B. L. S. Andrade da Costa personal communication).
In the three-toed sloth, in a region deep in the inferior temporal retina located 4.3 0.53 mm from the optic disc, ganglion cell density gradually increases from
about 500 cells/mm2 at the periphery to a peak of 1,700
cells/mm2 at the center (Andrade da Costa et al. 1987).
This corresponds to 40.56 of visual angle, considering
a retinal magnification factor of 0.106 mm/degree. The
region of maximum cell density (area centralis) constitutes only a weak retinal specialization, with a maximum resolution power estimated as 2.18 cycles/degree
of visual angle. The inferior nasal and superior temporal retinal areas contain intermediate cell densities and
retinal cell size is also intermediate in comparison to
the area centralis and superior nasal retina (Andrade
da Costa et al. 1987). The unusual position of the area
centralis and visual streak in the three-toed sloth retina
has been explained by the ability to rotate the head 180
while climbing upside down along horizontal branches
(Mendel 1985b). This would enable the animal, by obtaining an image upon the inferior temporal region, to
clearly visualize the branch directly above its head for
accurate claw placement.
In two-toed sloths, retinal cell density increases
approximately 10-fold, from 100 cells/mm2 at the periphery to a peak of 1,000 cells/mm2 in a central region
along the temporal horizontal meridian. Interestingly,
in contrast to the situation in the three-toed species, no
significant difference has been found in the distribution of ganglion cell sizes throughout the retina (Andrade da Costa et al. 1989). In C. didactylus, eye and
retinal dimensions were found to be 40% greater than
in Bradypus (Andrade da Costa 1987). The axial length
of the eye recorded in three specimens averaged 13.75
mm (Andrade Costa et al. 1989), which, according to
the relationship proposed by Hughes (1977) for several
mammals, would correspond to a retinal magnification
of 0.15 mm/degree. Thus, in spite of the lower cell density in the area centralis of Choloepus, the maximum
resolution power of 2.3 cycles/degree of visual angle
is comparable to that of Bradypus (Andrade da Costa
1987).
134
The gastrointestinal system and feeding and

digestion
Sloths are folivorous and feed mainly at night, although
their activity patterns may vary considerably, even
within the same species, with some individuals being
almost completely nocturnal and others mainly diurnal
(Chiarello 1998a, this volume). Three-toed sloths have
a more restricted diet than two-toed species, although
recent work indicates a wider variety of tree species are
utilized than was previously believed (Queiroz 1995;
Chiarello this volume). Sloths are extremely adept at
manipulating leaf-bearing branches, drawing them
toward the mouth and releasing the stem back to its
original position after eating the leaves (A. G. Chiarello
personal communication). The large soft tongue is pale
and flesh colored and densely covered with sharp backward-directed spines that are minute and transparent
(Beebe 1926). Goffart (1971) reports that the grinding
of food occurs in opposite directions in Choloepus and
Bradypus.
Sloths possess several unique structural features associated with feeding in an inverted position. For example, a large thick tongue and lips are required for food
manipulation and ingestion (Naples 1986). In all sloths,
but especially in Choloepus, the hyoid and its associated cartilages are large, fused and ossified, and apparently function as a rigid and stable base for the muscles
of the tongue and throat (Naples 1986). Moreover, the
masticatory and tongue musculature is reversed and
the palate is extremely complex, with tubercles of soft
tissue present medial to the maxillary teeth and separated by grooves that channel particles of food toward a
larger, central groove (Naples 1986). Food can be either
returned to the tooth rows for further mastication, or
passed along a posterior deepening groove to form a
bolus before being swallowed.
The stomach in sloths is composed of partially
separated sacs and three smaller chambers connecting
it with the pylorus (Goffart 1971). The stomach when
full may account for almost one third of body weight.
The crescent-shaped fore-stomach, which arises from
the inferior outer aspect of the stomach, is reported to
be 1520 cm long in Bradypus, but only about 3 cm in
Choloepus (Britton 1941). Sloths are foregut fermenters,
unlike most folivorous mammals. Digestion is characterized by the slow passage and low fermentation rate
of a large volume of digesta in the fore-stomach. Work
by Foley et al. (1995) on B. tridactylus indicated a mean
retention time of ca. 150 hours, the longest recorded for

any mammal. Most of the retention of digesta (>70%)
occurs in the stomach, but a substantial proportion is as
feces in the rectum, where considerable water absorption occurs (Goffart 1971).
The concentration of total short-chain fatty acids in
the stomach is hardly different from that in other foregut fermenters, but the rate of fermentation is considerably lower, for example, 612 mmol/l/h (Foley et al.
1995). Tree foliage is rich in fiber, and although protein
content is not high, more is present than in fruit (Milton 1979). However, fruit does provide ready energy
in the form of nonstructural carbohydrates. A diet of
leaves alone is a very poor source of energy, as leaves
possess low caloric density. Energy from leaves is further reduced because there is a limit on the maximal
daily bulk that can be processed. Sloths thus may be
living near the limits of their energy requirements. The
ability to retain a very large volume of digesta in the
stomach and intestines, coupled with a particularly low
rate of fermentation, may allow sloths to exist on such
a diet (see Foley et al. 1995). While slow fermentation is
undoubtedly partly the result of the lignified nature of
the foliage eaten, Foley et al. (1995) suggested it might
also be due to the low and variable body temperature
of sloths. Regardless, although this pattern of digestion
can yield energy only very slowly, Nagy and Montgomery (1980) argued it is sufficient because sloths expend
so little energy. This ability to exist on food of extremely
poor nutritional content may enable sloths to colonize
environments that cannot support other folivores with
higher metabolic rates.
Metabolism and motor activity

Sloths have a metabolic rate only 4045% of what might
be expected for their body weight (McNab 1982). Even
correcting for the contents of the stomach and intestines and the volume of urine in the bladder, the resting
metabolism in two- and three-toed sloths is still 34%
and 51% respectively, below that of similar-sized mammals (McNab 1982). However, as sloths sleep or rest
for up to 20 hours per day, they probably only perform
about 10% of the work of a similarly sized mammal
(Goffart 1971). McNab (1985) suggested that low basal
metabolic rates in folivorous mammals might reduce
absorption of the many toxic substances, including alkaloids, phenols, and terpines, often found in leaves.
He also postulated that, because in arboreal mammals
muscle mass makes up a relatively small proportion of

total body mass, they consequently have a larger proportion of tissue with a low metabolic rate.
McNab (1978) estimated the mean energy expenditure for Bradypus to be 95.5 kcal/day. However, Montgomery and Sunquist (1975) calculated that the daily
intake of food by Bradypus provided just 54.4 kcal.
McNab (1978) argued that the discrepancies between
these estimates may simply reflect higher rates of decay
in the feces than assumed by Montgomery and Sunquist (1975).
Consistent with McNabs (1985) observation regarding arboreal mammals, sloths have a reduced muscle
mass in comparison to other mammals of similar size.
Several studies employing electrophysiology, biochemistry, and histology have examined neuromuscular responses in sloths. Much of this work was carried out
by Goffart and his colleagues on C. hoffmanni and this,
along with the results of earlier studies, is summarized
by him in great detail (Goffart 1971). To recap briefly
here, sloth skeletal muscles are all of the red variety, although the organization of the filaments within the fibril is the same as that seen in other vertebrates. Goffart
(1971) found contraction times of muscles in two-toed
sloths to be 46 times slower, and half-relaxation times
914 times longer, than those of corresponding muscles
in the cat. He related these findings to those of Barny
et al. (1967), who found that ATPase activity and actinbinding ability were extremely low in sloths compared
to cats. Similarly, Silva et al. (1975) showed that the
gastrocnemius internus and tibialis anticus muscles in
B. variegatus had very slow contraction times relative
to other mammals, being 34 times slower than in the
cat.
Work by Silla and Stephens (1976) in B. variegatus
demonstrated that cells within the diaphragm strongly
resembled those in the skeletal muscle of other mammals, being multinucleate with the nuclei lying in the
subsarcolemmal space. Sarcomere length averaged 3.0
m, much longer than recorded for frog semitendinosus
or cat heart papillary muscle. However, the diaphragm
muscle was very similar to that of frog semitendinosus in regard to sarcomere pattern and its components.
Contractility of the diaphragm in Bradypus was slower
than in Choloepus, the sartorious of the frog, and the
trachealis of the dog, although faster than mammalian
smooth muscle.
Toole and Bullock (1973) studied neuromuscular
responses in C. hoffmanni and B. variegatus and con-
135
cluded that, while muscular mechanics most certainly

contribute to the slow movement of these animals, they
are not the sole rate-limiting factor. They demonstrated
that the rapidity of spontaneous movement and the
stretch reflex were markedly slowed in sloths when rectal and muscle temperature were reduced, but did not
change significantly when these were increased. Toole
and Bullock (1973) also reported that needle electromyography of the latissimus and pectoral muscles failed to
produce prolonged postinsertion potentials as usually
seen in myotonic muscle. Muscle twitch contraction
averaged 120 ms and relaxation half-time 108 ms. Interestingly, the height of muscle contraction and speed
of relaxation were not increased, as might be expected,
by injections of calcium gluconate, quinine, quinidine,
or procaine into the brachial artery. Although ethylene
gallamine tetra-acetic acid caused a loss of responsitivity to nerve and direct muscle stimulation, these effects
were reversed when calcium ions were added to the
perfusates. Toole and Bullock (1973) thought it unlikely
that the limitation in the responses of sloth muscles, in
comparison with those of other mammals such as the
rabbit, was in electomechanical coupling or related to
calcium. They speculated that sloth muscle might be
slow because tropic material from the motoneuron is
unavailable during development to convert newborn
red to adult white or mixed muscle.
Cardiovascular system
A detailed update on blood chemistry in sloths was
provided by Gilmore et al. (2000). The heart in sloths is
small compared to animals of similar weight, although
larger in Choloepus than Bradypus. The atria are large;
the shape of the atrial valves is variable and the orientation of the ventricular papillary muscles is irregular
(Goffart 1971). Nearly all the research into cardiac function in sloths has been carried out on B. variegatus and
a summary of selected results is provided in table 13.1.
A comparative study of myocardial mechanical
characteristics by Cingolani et al. (1979) showed that
the time to peak tension was significantly higher in B.
variegatus (444 33.7 ms) than in the cat (327 21.5
ms) and the rat (119 9.1 ms). The derivative of the force
of contraction in relation to time (df/dt maximum),
which gives an idea of how fast the force changes and is
an index of the force of contraction, also differed in the
three species (sloth: 2.9 0.7 g/s, cat: 5.2 0.6 g/s, rat:
136
Table 13.1. Summary of Cardiac Function and Arterial Blood Pressure in Sloths
A. Cardiac function
Unanesthetized
Anesthetized
Reference
0.240 0.033
0.93 0.13
0.248 0.039
0.97 0.16
3
3
Range or mean SD
Reference
60110
89.7 9.7
70130
94.1 13.3
91.0 4.5
76.4 20.8
79.8 22.3
83.3 10.8
96.5 8.6
84.5 9.9
65.5 5.9
74.1 5.6
1
3
1
3
4
5
5
6
6
6
6
6
48108
C. Blood pressure (mm Hg)
Systolic
Diastolic
Unanesthetized
Unanesthetized
Seated in chair
In light
In dark
177 31
121 22
129.2
125 21
120 21
128 23
86 17
89.2 13.6
88 15
85 17
Cardiac output (l/min)

Cardiac index (l/M2/min)
B. Heart rate
Bradypusunanesthetized
Bradypusunanesthetized
Choloepusunanesthetized
B. variegatusanesthetized
B. variegatusanesthetized
B. variegatussitting
B. variegatussuspended
B. variegatus-before atropine
B. variegatusafter atropine
B. variegatusbefore propranolol
B. variegatusafter propranolol
B. variegatusafter double blockade
with atropine and propranolol
C. didactylusanesthetized
Mean Arterial
Heart Rate
144 25
89 10
98 18
73 16
102.5 17.0
87.4 12.6
100 17
97 17
Reference
3
8,9
7
9
9
Sources: (1) Britton 1941, (2) Bush and Gilroy 1979, (3) Cabral et al. 1980, (4) Oliveira et al. 1980, (5) Duarte et al. 1982, (6) Duarte et al.
1983, (7) Duarte et al. 1987, (8) Silva 1996, (9) Duarte et al. 2003.
Note: All measurements in A are from B. tridactylus and in C from B. variegatus. Except where indicated, all animals were unanesthetized
for the measurements in B and C.
6.4 0.8 g/s). Not only was the contractile process slow
in the sloth, but the relaxation process was also.
Biotelemetry was used to monitor heart rate changes
during a 24-hour period in three-toed sloths allowed
to roam freely in a laboratory (Duarte et al. 2004). The
data revealed a biphasic pattern, with lower values observed during daylight, probably related to a decrease
in motor activity at this time. The highest heart rate was
observed at 19:00 h (Silva 1996). A study by Duarte et al.
(1983) on the autonomic component of heart rate in B.
variegatus revealed that atropine raised heart rate significantly whereas propranolol slowed it (table 13.1B).
After double blockade (in either order), heart rate was
lower than control values, indicating a slight predominance of sympathetic tone at rest. The fact that atropine
elevated heart rate by a mean of only 16% suggests that
there may be an upper limit of around 110 bpm for sloth
heart rate.
The ECG pattern in B. variegatus reveals the presence of a sinus rhythm. For a mean heart rate of 81 18
bpm the duration (in seconds) of the P wave was found
to be 0.05 0.02, the PR segment 0.07 0.02, the QRS

complex 0.07 0.02, the ST segment 0.11 0.005, and
the T wave 0.15 0.05 (Silva et al. 2005). These values
are near those seen in other mammals of the same size.
In the frontal plane the mean P axis was +34 68 (100 to +45), the mean QRS axis was -35 63 (-100
to +60) and the T axis was -23 68 (- 90 to +75).
This suggests that, as sloths have a large abdominal cavity and the alimentary contents accumulate for several
days before elimination as feces, when seated the abdominal mass would tend to displace the diaphragm
upward, causing the heart to shift and produce a left
axis deviation (Silva et al. 2005).
Sloth hearts are more difficult to fibrillate than those
of cats, and, once fibrillation is induced, they recover
more quickly (Oliveira et al. 1980). In B. variegatus the
average ventricular fibrillation threshold current was
2.3 0.4 mA/g and, when fibrillation was induced, 40%
of sloths showed a spontaneous reversal. This ventricular threshold current was four times that recorded in
young dogs (Valentinuzzi et al. 1984).
Measurements of arterial pressures recorded in sloths

are listed in table 13.1C. Experiments carried out on B.
variegatus, cannulated under local anesthesia, reported
arterial pressure fell substantially after the administration of pentobarbitone, suggesting that endogenous
catecholamines might promote the apparent hypertension seen in unanesthetized sloths (Oliveira et al. 1980).
Studies on unanesthetized sloths of the same species revealed mean arterial blood pressure was within the human range and not significantly greater than that seen
in relaxed unanesthetized laboratory animals (Duarte
et al. 1987). However, handling or the mere approach of
a person produced a rise in sloth blood pressure, and
in some cases a hypertensive bout, suggesting extreme
lability in the control of blood pressure.
Recent studies using biotelemetry in free-moving
sloths in the laboratory indicate that blood pressure
levels, like heart rate, show circadian variation. Spontaneous variation in blood pressure levels occurred over
a period of 24 hours without any relation to external
stimuli (Duarte et al. 2003). Blood pressure levels were
significantly higher in the light than the dark period
and acrophase occurred between 16:00 and 17:45 h. This
circadian variation is similar to that seen in cats (Brown
et al. 1997), dogs (Mishina et al. 1999), and marmosets
(Schnell and Wood 1993). During the experiment, behavior was recorded as resting (behavioral sleep and
wake-alert), eating, and moving. During behavioral
sleep mean arterial blood pressure (in mm Hg) decreased from 110 19 to 90 19, but both feeding and
moving about caused it to rise significantly, from 96
17 to 119 17 at 17:00 h, and from 97 19 to 105 12
at 15:00 h, respectively. These results demonstrate the
existence in sloths of a biphasic circadian fluctuation
in blood pressure, with higher levels during light and
synchronized mainly with feeding behavior (Duarte et
al. 2003). Not surprisingly, circadian rhythms of activity have also been confirmed for free-ranging sloths observed in the wild (see Chiarello 1998b, this volume).
Studying cardiovascular responses to locomotor activity and feeding in unrestrained B. variegatus, Duarte
et al. (2004) found significant increases in systolic and
diastolic blood pressures and heart rate of 8%, 7%, and
14% respectively for locomotor activity and 27%, 21%,
and 14% respectively for feeding. During eating there
was abundant salivation and elevated blood pressure
levels persisted. These results suggest that the proportionally greater increase in blood pressure than in heart
rate during feeding is due to a rise in peripheral vascular resistance as part of an overall response to feed-
137
ing. The fact that blood pressure and heart rate levels
remained elevated during feeding suggests that baroreceptors have a minimal impact in buffering blood pressure during this behavior.
The limbs of sloths possess a rete mirabile (described in detail by Goffart 1971). This consists of a
branching of the main artery and its accompanying
vein into bundles of small arteries (about 50) intermingled with parallel thin-walled veins, all enclosed within
a common fibrous sheath. There are numerous anastamoses between individual veins in the bundle, but few
between the arteries. Such vascular bundles run the
whole length of the limb. It would appear that circulation through the rete can take place along alternative
pathways, giving preference at times to a direct flow
through the main arterial vessel and on other occasions
to passage through the high resistance smaller arterial
branches. The plexiform arrangement of the veins, in
which the blood flows in a counter-current fashion to
that in the arteries, may provide a cooling mechanism.
It could also reduce thermal loss from the arteries at the
extremities and thereby conserve heat. Sloths lack the
ability to generate heat by shivering. The temperature
gradient along the rete can be 0.2C at normal environmental temperatures, which is 10 times more than in
the human (Scholander and Krog 1957).
It has also been suggested that the rete mirabile may
act as a blood reservoir or facilitate the return of blood
to the heart (see Goffart 1971). This could be extremely
important for sloths, in which there are long periods
of sustained rather than intermittent muscular contractions, and would reduce any risk of deep vein thrombosis. Sloths are distinguished from all other mammals by
the ability of the skeletal muscle to maintain the animals
clinging to a support, or suspended in a hanging position, for an extraordinary length of time. Davies (1947),
dealing with similar issues in the slow loris (Nycticebus
tardigradus), suggested that the rete maintains the supply of nutrients to the muscles during the typical slowly
developing and sustained tetanic contractions that take
place. The rete would thus provide multiple arterial inputs to the muscle at regular intervals along its entire
length and act to maintain normal blood flow under
the peculiar and difficult gravitational situation of being upside down.
Lungs and respiration

Sloths have an elongated thorax that is more broad
than deep (Goffart 1971). The unusually delicate lungs
138
Table 13.2. Respiratory Rates Measured under Different Conditions in Sloths

Measurement condition
Choloepus
Excited or agitated
Awake and active
Inactive
Anesthetized
Bradypus
Awake
Anesthetized
Impedance technique
Breaths per min
Reference
1030
73 (range: 6578)
46 (range: 3852)
1848
Britton (1941)
Meritt (1985a)
Meritt (1985a)
Bush and Gilroy (1979)
5.81 0.81
3.0 0.8
8.6 5.3 (range: 4.920.6)
Santos et al. (1998b)

Felix et al. (1987)
Santos et al. (1988a)
of both two- and three-toed sloths are not clearly subdivided into lobes as in most mammals. In the fetal sloth,
however, there is an upper and lower lobe in both lungs
separated by loose connective tissue. Wislocki (1928a)
proposed that the postnatal disappearance of the lobular division might be because of the lack of lobar slippage, due to the constant shape of the rib cage and
diaphragm. The ribs and diaphragm remain essentially
the same in both prone and supine positions, which is
unusual among mammals (Hoffman and Ritman 1985).
In Choloepus the trachea is extremely short and divides
immediately upon entering the thorax into the bronchi, which enter the lungs (Goffart 1971). In contrast, in
Bradypus the length of the trachea is 2528 cm and is
probably correlated with the enormous range of movement possible in the head and neck, with the thoracic
loop providing slack that can be taken up when necessary (Goffart 1971).
Oxygen utilization in the air breathed is only 2% and
the carbon dioxide component in the alveoli is thus
correspondingly low (less than 3%, Goffart 1971). Interestingly, sloths can recover after having their breathing
arrested for more than 20 minutes; this is accompanied
by a depression in the resting metabolic rate. In such a
situation, the heart rate falls to less than one-third of
its normal resting value and lactate accumulates in the
tissues, but is lost during recovery (Beebe 1926; Britton
1941; Goffart 1971).
Comparative data on respiratory physiology are rare,
with the exceptions of respiratory rate and blood gases.
The limited data available indicate respiratory rate is
considerably higher in Choloepus, although this rate
may depend on the animals physiological state (table
13.2). Blood gases, measured at 37C by Bush and Gilroy (1979) in anesthetized C. didactylus, were: pCO2 =
33.2 2.0 mm Hg, pO2 = 89.9 18.8 mm Hg; pH was
7.40 0.03. Pedrosa et al. (2002) obtained the following measures for unanesthetized B. variegatus at rest,
adjusted to body temperature (32C): pCO2 = 35.2 5.3

mm Hg, pO2 = 77.5 8.2 mm Hg; pH was 7.42 0.05.
Most other studies of respiratory function have been
carried out on B. variegatus and are summarized in tables 13.3 and 13.4. Aside from these, other notable findings include the work of Santos et al. (1998a), who used
an impedance technique to study respiratory function
in B. variegatus, and found breathing to be slowest at
11:30 h and most rapid at 02:30 h, when there was an
associated increase in the animals motor activity. In B.
variegatus anesthetized with sodium pentobarbital, but
breathing spontaneously, tidal volume was 57 14.4 ml,
inspiration time was 3.8 0.4 s and expiratory time was
15.9 0.7 s for a respiratory rate of 3.0 0.8 breaths per
minute (Flix et al. 1987). The same authors recorded
respiratory minute volume as 181.3 27.9 ml and there
was a long expiratory pause of 13.1 6.8 s. In the lab,
free-roaming B. variegatus doubled their breathing
rate while feeding (from 5 to 10 breaths per minute),
without any significant changes in arterial pH, PCO2,
or PO2 levels (Duarte et al. 2004). It appears that this
increase in breathing rate was due to the act of mastication. When Flix et al. (1988) compared active versus
passive mechanical properties of the respiratory system
in anesthetized B. variegatus, active elasticity and active
resistance were 13% and 56% higher, respectively, than
their passive counterparts.
The endocrine organs, liver, spleen, and kidneys

Gilmore and da Costa (1995) provided a brief update on
the endocrine system in sloths. The adrenals (averaging
245 g combined weight in Bradypus) are situated high
up in the abdominal cavity on either side of the aorta
and well separated from the kidneys, which occupy
a pelvic position. As is normal in mammals, they are
comparatively much larger in the fetus because of an
expanded cortical region. Adrenaline and noradrena-
139
Table 13.3. Pulmonary Volumes and Capacities in Anesthetized Sloths (B. Variegatus)
Measure
Tidal volume
Inspiratory volume
Inspiratory capacity
Vital capacity
Functional residual capacity
Mean SD
13.5 3.4 ml
24.2 6.8 ml
66.1 6.6 ml
75.8 6.8 ml
33.9 6.6 ml
Source: Aguiar et al. (1988a).
Table 13.4. Features of the Respiratory System of B. Variegatus

Mechanical properties

Passive elasticity of the lung and thoracic wall

Measure
Mean SD
Resistance
Active elasticity
0.33 0.10 cm H2O/l

0.09 0.02 cm H2O/l
Respiratory system
Pulmonary system
Thoracic wall
0.183 0.030 cm H2O/ml

0.045 0.002 cm H2O/ml
0.134 0.024 cm H2O/ml
Sources: Aguiar et al. (1988b), Flix et al. (1988).
line are stored in specific cells grouped in islets in the

medulla of the adrenal.
In Bradypus the thyroid consists of two independent
lobes and lacks a capsule; the follicles are irregularly
distributed within large areas of fat (Goffart 1971). In
both genera of sloths the thyroid is comparable in size
to that of other mammals, but thyroxine secretion is
low. As the sensitivity of the peripheral tissue to thyroxine is also not particularly high, these factors could contribute to the low metabolic rate seen in these animals.
However, sloths have no circulating thyroxine-binding
globulin (Toole 1972), which probably indicates that
thyroid function is adequate.
In the three-toed sloth, immunohistochemical studies by Mota et al. (1992) revealed typical islets of Langerhans in the pancreas, as well as endocrine cells scattered
about singly or in small groups in the exocrine tissue.
Most of the pancreatic endocrine cells are immunoreactive for insulin and glucagon, while only a few are
immunoreactive for somatostatin and serotonin (see
Goffart 1971).
In contrast to earlier studies that reported the sloth
liver to be very small relative to that of other similarly
sized mammals, Goffart (1971) concluded that it may be
only marginally so. Interestingly, developmental studies
(see Goffart 1971) suggest the liver in adult sloths has
undergone rotation through an angle of 135 clockwise;
this has not been documented in any other mammal.
In three-toed sloths, the liver has two lobes, but there
are three in Choloepus. Bradypus does not possess a gall

bladder; Choloepus does, but it is small, thin walled,
and contains little muscle. The bile duct enters the upper part of the duodenum well above the pancreas in
Choloepus, but is much lower and distal to the pancreatic duct in Bradypus.
The spleen of Bradypus is irregularly tubular in
shape and consists of one single mass. In contrast, in
Choloepus the spleen is roughly triangular and markedly flattened; it is intimately fused with the adjacent
pancreas. See the chapter by Casanave and Galndez
(this volume) for further discussion of this tissue in armadillos.
Sloth kidneys are situated low in the abdominal cavity; their anatomy and physiology are covered in some
detail by Goffart (1971). The large bladder, reaching 120
mm in diameter, may contain close to a liter of urine
when full. In young sloths the bladder can occupy
half the peritoneal cavity (Beebe 1926). Meritt (1985a)
stated that the urine of Choloepus is clear and yellow
to brown, with a specific gravity of 1.0201.031 and a
pH of 7.09.0. In B. variegatus, Nagy and Montgomery (1980) found that water influx and efflux rates were
almost equal, averaging around 39 ml/kg/day. Water
intake from milk by young sloths amounted to about
15 ml/kg/day. Lactating sloths allocated around 5% of
their water intake and 11% of their energy intake to milk
production (Nagy and Montgomery 1980). This same
study estimated water production during oxidation of
140
food to be only 4.2 ml per day, leaving a daily deficit

of more than 34 ml. Remaining water input apparently
came from ingested leaves rather than drinking rainwater, as there are no recorded accounts of sloths drinking in their natural habitat. However, Meritt (1985a)
reported average daily water consumption of captive
C. hoffmanni as 170 ml per animal; measured amounts
of excreted urine averaged 808 ml (range = 3401600
ml). The presence of cystic calculi, composed of calcium, magnesium, ammonium, and phosphorus, was
reported in captive two-toed sloths by Rappaport and
Hochman (1988). They speculated that the stones were
a result of the sloths diet, which consisted of a mixture of fresh fruit and vegetables along with a protein
supplement.
Reproductive behavior and breeding

Goffart (1971) provided a very detailed description of
the internal anatomy of the reproductive system in
sloths, summarizing earlier findings. It is normally extremely difficult to determine the sex of a sloth from
an examination of its external genitalia (Lara-Ruiz and
Chiarello 2005). These are small and inconspicuous
and close to the rectal orifice, but separated from it by
a small amount of skin. The male genitalia are anterior
to the rectal orifice and appear in the mature adult as a
small penis about the diameter of a human finger. At the
end of this penis is a very small round opening. LaraRuiz and Chiarello (2005) have recorded the length of
the penis as 1.52.0 cm in mature B. torquatus. The penis consists only of a small pair of corpora cavernosa;
no true corpus spongiosum is present. In the female, the
external genitalia are closer to the rectal opening and
the clitoris is smaller than the penis with a slit along
its length.
Sloths are slow to mature. Tirler (1966) reported that
three-toed sloths may take up to six years to reach maturity, but this is undoubtedly an overestimate. It is almost
certainly closer to three or four years, as reported for B.
torquatus by Lara-Ruiz and Chiarello (2005). In captivity, female two-toed sloths do not usually breed until
three years old and males only reach sexual maturity
when four to five years of age (Eisenberg and Maliniak
1985). However, Meritt (1985a) recorded a case of sexual
maturity in a female two-toed sloth (C. hoffmanni) at
two years of age; it died at the age of 30 months with
a well-formed fetus in utero. In French Guiana, Richard-Hansen and Taube (1997) estimated that female C.
didactylus reach maturity at the end of their third year
and males during their third year. In the same locality,

C. hoffmanni females appear to become mature at the
beginning of their third year and males during it. Captive two-toed sloths continued to produce young until
25 years old and males remained sexually active up to
21 years (Eisenberg and Maliniak 1985).
Little is known about the onset of estrus, courtship,
and mating in Bradypus, but it appears to be direct and
brief, with most of the initiative coming from the male
(Goffart 1971). Similarly, not much is known about levels of reproductive hormones in sloths and their alterations during the estrous cycle. Gilmore et al. (1991) recorded circulating estradiol and progesterone levels of
49 pmol/l and 38 nmol/l respectively in a nonpregnant
adult female in Recife. Prolactin levels ranged from 2.28
to 5.47 ng/ml in eight male sloths, were 1.89 ng/ml in a
pregnant female with a small fetus, and 4.04 ng/ml in
a female carrying a near-term fetus (Gilmore and da
Costa 1995). Very recently Mhlbaeur et al. (2006) used
a radioimmunoassay to examine fecal levels of estradiol
and progesterone metabolites in four B. variegatus from
Recife. A close correlation was found between plasma
and fecal levels of these gonadal steroids and ovulation
was detected in two animals. The findings demonstrate
that this technique can be used to monitor reproduction in sloths in a noninvasive manner. Moreover, this
study and the earlier ones in Recife by Gilmore et al.
(1991, 1994) indicate that follicular development leading to ovulation is not seasonal, at least for this species
in this locality.
Vent to vent copulation has been observed in B. tridactylus by Richard-Hansen and Taube (1997). It lasted
only a few minutes while the male, secured by his forelimbs, held the female, who was hanging from a branch,
with his hind limbs. Five minutes later mating occurred
again, but this time with the male on the females back.
This copulation was interrupted by the female, who
withdrew after two minutes. However, the male followed her and eventually they mated again vent to vent
for one and a half minutes. The whole session from initial approach to final separation lasted 25 minutes. In
another incident, a captive male B. torquatus was observed by Pinder (1993) erecting its genitalia while in
the presence of another male and a subadult female;
genital length was only 8 mm! Behavioral changes are
associated with estrus in two-toed sloths (Meritt 1985a).
Two female C. hoffmanni at Lincoln Park Zoo, Chicago,
were observed rubbing against each other in the anogenital region. Female sloths were also seen to solicit
a male by attempting to rub their anogenital region
against his, but this elicited little response. When mating took place it was vent to vent.
There is still some disagreement as to whether or
not there is a breeding season for three-toed sloths.
Early workers were of the opinion that mating might
occur at any time, but more often during the dry season
(March and April). Pregnancy then proceeded during
the wet season. More recent work by Taube et al. (2001)
in French Guiana suggests that breeding is definitely
seasonal in B. tridactylus, with births occurring at the
beginning of the dry season. In B. variegatus there appears to be a definite, but somewhat irregular, breeding
season that varies according to climatic variations in
different locations and between years. In the relatively
temperate climate of So Paulo, a B. variegatus birth was
recorded in August (Goffart 1971); one also occurred
during the same month in Recife. However, another
birth was recorded in Recife during November (Gilmore and da Costa 1995), and two others were recorded
in the Mamirau region of Amazonas in mid-July and
mid-February (Queiroz 1995). Breeding in B. torquatus
is considered aseasonal (Pinder 1993). Newborn individuals, along with their mothers, have been captured
in July and two juveniles near weaning in August and
December; another juvenile of intermediate weight was
caught in November. In Santa Teresa, Brazil, Lara-Ruiz
and Chiarello (2005) noted that newborn sloths were
more commonly seen at the end of the wet season (FebruaryApril), indicating that gestation takes place during that period. Taube et al. (2001) have proposed that,
because three-toed sloths eat a highly specialized diet
and are dependent on seasonal variation in abundance
of their preferred food, the breeding season is limited
so as to cover the high energy demands of pregnancy
and lactation. Moreover, seasonal reproduction, associated with high population density, would maximize
the opportunity for sexual encounters in these solitary
animals.
Work in Costa Rica on B. variegatus by Toyama et al.
(1990) suggested that males are fertile throughout the
year. However, other research in Pernambuco by Gilmore et al. (1991, 1994) on the same species indicated that,
although breeding may occur throughout the year, only
a minority of males are fertile and sexually active at any
one time. In the majority of nine male sloths examined
in Recife, no active spermatogenesis was taking place
and testosterone levels were very low, in the range of
0.61.35 nmol/l. However, these rose more than 20-fold
after a bolus injection of 5 g of GnRH was administered.
141
There is little evidence of seasonality in the breeding of the two-toed C. didactylus, as births have been
recorded in every month except April, September, and
November at the National Zoological Park of Washington, D.C. (Goffart 1971). In French Guiana, pregnant females of the same species were captured in every month
except June and September by Taube et al. (2001), who
even found one sloth to be both pregnant and lactating. Breeding may be slightly seasonal in C. hoffmanni
under natural conditions, with pregnancy proceeding
throughout the rainy season and births taking place at
the beginning of the dry season (Taube et al. 2001).
Detailed information on placentation in sloths has
been provided by Benirschke and Powell (1985) and
Benirschke (this volume). The length of gestation is
uncertain in three-toed sloths, but is thought to be 69
months in B. variegatus, with a one-year interval between births. More recent estimates favor a gestation
length of 6 months, which is much less than that in
two-toed sloths, where gestation may equal or exceed
10 months (Taube et al. 2001). In Bristol Zoo, England,
intervals between successive births in C. didactylus
were 395, 810, and 910 days (Partridge 1991), while in
French Guiana intervals between births were 16 months
or more (Taube et al. 2001). In captivity, female C. hoffmanni whose offspring died within two weeks of birth
were able to produce a new young within 14 months
(Eisenberg and Maliniak 1985). If, however, the young
survived, the average interval to the next birth was 26
months, considerably longer than that recorded in the
wild for the same species (Taube et al. 2001).
Young C. didactylus begin to eat solid food when
only a few weeks old, progressively switching from milk
to a solid diet. Weaning length is variable, with lactation lasting 35 months. In French Guiana, young B.
tridactylus become independent of the mother as early
as 5 months of age, but B. variegatus may become independent two or three months later (Taube et al. 2001).
At this point the mother will move to a new home
range, leaving the former one to her young. E. Taube
(personal communication) believes such behavior is
advantageous because inexperienced young do not suffer the costs of dispersal, which would probably entail a
strenuous search for a new home range containing suitable food. In B. torquatus, young become independent
between 8 and 11 months of age (Lara-Ruiz and Chiarello 2005); the mother then abandons the young and
travels up to several hundred meters away. The two are
unlikely to encounter each other again, as their home
ranges will overlap little, if at all (Lara Ruiz and Chi-
142
arello 2005, A. G. Chiarello personal communication).

Taube et al. (2001) reported that young C. didactylus
become independent at around 12 months old, compared to about 9 months in C. hoffmanni. The duration of lactation is unknown in the latter species. The
mothers tendency to carry her offspring decreases at
7 months and at 9 months she may begin rejecting the
young sloth, although a behavioral association can be
maintained through to the next birth; in captivity this
association is virtually unlimited. It has been suggested
that if the association between mother and offspring
lasts less than 8 months the young is unlikely to survive.
Sloths are reported to give birth to their young on the
ground or in a hanging position (Goffart 1971). Most of
the information about birth and growth of the young
has been obtained from observations of two-toed sloths
in zoos. Examination of female C. hoffmanni that had
recently given birth revealed that the genitalia were not
obviously swollen, torn, or distended (Meritt 1985a).
Females invariably give birth to a single young (weighing 350450 g). Young are born in a mature state and
capable of coordinated movement, with eyes open, hair
grown, teeth erupted, and claws well developed. Female
sloths have two pectoral mammary glands. Suckling by
the young stimulates the flow of milk. Newborn captive
two-toed sloths suckle for 615 minutes at a time on
several occasions each day (Meritt 1985a). Grabowski
et al. (2002) recently examined the constituents of milk
drawn from C. didactylus. In comparison to cows milk,
sloth milk was higher in fat (6.9%), protein (6.1%), and
all minerals and vitamins involved in ossification, dentition, and energy status.
For several days after birth, mothers spend a considerable amount of time licking the newborns head
and face and cleaning the anogenital region. Licking
by the mother stimulates the young to pass urine and
feces, which are then consumed by the mother (Meritt
1985a). According to Britton (1941), two-toed sloths are
very solicitous of their young and guard them jealously,
but three-toed sloths show only slight regard for their
offspring. Males of all species ignore the young altogether. Generally, two-toed sloth mothers with newborn young will permit other females to approach and
lick them. However, some mothers react aggressively
to the approach of others and drape their limbs over
the offsprings body to shield it. Solid food is first taken
by young two-toed sloths when just a few weeks old
and play behavior begins about the same time. Juveniles commence exploring away from their mothers at
this same age; as time elapses the explorations become
longer and the distances away from the mother farther.
If unable to return to their mother, young sloths call out
immediately, with bleats lasting 3090 s (see above).
Mothers respond by moving toward the sound.
In contrast to adults, young sloths are nimble in
their movements. Baby sloths cling to the mothers abdominal hair with an extremely powerful grip and are
very skilled at avoiding being crushed when the mother
is climbing (Tirler 1966). If water has to be crossed, the
young climb onto the mothers back. Young two-toed
sloths first adopt clinging in an upside-down position when 2025 days old and may be seen sleeping in
an adult position at 50 days of age (Meritt 1985a). By
5 months the young sloths are eating solid food away
from the mother and by 6 months defecate in the adult
manner. Although they are capable of an independent
existence earlier, they frequently continue to be carried
by the mother until 8 months old (Meritt 1985a). There
is no sudden break in the relationship between the two,
instead it changes gradually from one of dependence
to one of continued association that can last for up to
2 years.
Concluding remarks
While a great deal of data has been collected on the
biology of sloths over the last two centuries, there are
still some large gaps in our knowledge. Currently, only
limited information exists about certain aspects of sloth
physiology, especially those related to endocrinology
and reproduction. In particular, little is known about
reproductive development, fertility, courtship, and mating in three-toed sloths. Such data are clearly essential
from a conservation standpoint because of population
declines due to loss of habitat from continued deforestation (Aguiar and Fonseca this volume). It is also vital
that, again in the interests of conservation, noninvasive
approaches such as those described by Chiarello (this
volume), Duarte et al. (2003, 2004), Lara-Ruiz and Chiarello (2005), and Mhlbauer et al. (2006) be employed
as much as possible. With increasing miniaturization of
devices to track, film, and record body functions, one
can expect to see some interesting developments over
the next few decades.
14
Sperm evolution in dasypodids
Pablo D. Cetica and Mara Susana Merani
particulares caractersticas estructurales de los tractos

reproductivos femeninos tienen relacin con la morfologa espermtica en cada gnero de Dasypodidae.
Resumen
El espermatozoide es una clula extremadamente especializada, cuya morfologa parece haber sido modelada para fertilizar al ovocito. Los espermatozoides de
los euterios presentan un alto grado de uniformidad
en la forma y dimensiones. La mayora de las especies
tienen una cabeza oval y un largo promedio de 70 m,
patrn propuesto como el tipo bsico ancestral. Las
formas derivadas han sido descriptas principalmente
en varios grupos de roedores. Sin embargo, nosotros
hemos informado una importante diversificacin evolutiva en los espermatozoides de dasipdidos. Basados
en el anlisis de la forma y dimensiones de los espermatozoides, se identificaron cuatro grupos espermticos
diferentes. El patrn espermtico 1, representado en el
gnero Dasypus, presenta el tpico patrn espermtico
ancestral de mamferos. Los restantes tipos espermticos son derivados. Los patrones espermticos 3 (Priodontes, Cabassous) y 4 (Chaetophractus, Zaedyus,
Euphractus) presentan una cabeza muy peculiar, con
forma de pala o cuchara respectivamente. Estas cabezas
son muy grandes y extremadamente finas, con medidas de largo y ancho tpicamente mayores que aquellas
de otros mamferos. El patrn 2 (Tolypeutes) es nico
comparado con los otros; la cabeza espermtica tiene
forma de cuchara pero tambin presenta una particular extensin laterocaudal que le otorga una asimetra
notable y caracterstica. Esta diversificacin evolutiva
de los espermatozoides de dasipdidos implica fuertes
presiones selectivas que afectaron la morfologa de las
clulas espermticas en los diferentes gneros de armadillos. Observamos una asociacin entre la morfologa
de los tractos genitales femeninos y los grupos esper
mticos previamente descriptos. Concluimos que las
Resumo
O espermatozide uma clula extremamente especializada e cuja morfologia parece ter sido modelada para
fertilizar o ovcito. Nos espermatozides de eutrios
observa-se um alto grau de uniformidade na forma e
dimenses espermticas. A maioria das espcies apresenta uma cabea oval e um comprimento mdio de 70
m, padro proposto como o tipo bsico ancestral para
os espermatozides. As formas derivadas foram apenas
descritas, principalmente em vrios grupos de roedores.
Mesmo assim, ns tambm revelamos uma importante
diversificao evolutiva nos espermatozides de dasipoddeos. Baseados na anlise da forma e dimenses
dos espermatozides podemos classific-los em quatro
grupos espermticos diferentes. O padro espermtico
1, representado pelo gnero Dasypus, apresenta o tpico
padro espermtico ancestral dos mamferos. Os tipos
espermticos restantes so derivados. Os padres espermticos 3 (Priodontes, Cabassous) e 4 (Chaetophractus,
Zaedyus, Euphractus) apresentam uma cabea muito
peculiar, com forma de p e colher, respectivamente.
Essas cabeas so muito grandes e extremamente finas,
com medidas de comprimento e largura tipicamente
maiores que aquelas de outros mamferos. O padro 2
(Tolypeutes) nico comparado aos outros; a cabea
espermtica possui forma de colher, porm tambm
apresenta uma particular extenso latero-caudal que
fornece uma assimetria notvel e caracterstica. Assim
sendo, estabelece-se uma importante diversificao
evolutiva nos esparmatozides de dasipoddeos. Uma
143
144
P. D. Cetica and M. S. Merani
forte presso seletiva deve ter afetado a morfologia das

clulas espermticas dos diferentes gneros de tatus.
Ns tambm observamos uma associao entre os diferentes tratos genitais femininos de dasipoddeos e os
grupos espermticos previamente descritos. Destas observaes pde-se assumir que caractersticas estruturais particulares dos tratos reprodutivos femininos possuem distintas relaes com a morfologia espermtica
em cada gnero de Dasypodidae.
Sperm morphology
The concept of evolutionary change is generally associated with the evolution of species but can also be applied to systems, organs, and cells. Spermatozoa are a
clear example of this: their shape and size have changed
significantly in accord with characteristics of the reproductive systems found in different species.
The spermatozoon, or male gamete, is an extremely
specialized cell and its morphology would seem to have
been modeled for the function it must carry out: fertilization of the oocyte, or female gamete. In marine
invertebrates with sexual reproduction and external
fertilization, sperm arose as the cells in charge of moving in the water to look for the oocytes released there by
females. To be able to accomplish this, sperm adapted
the organelles and conventional proteins that all cells
possess and distributed them into a great globular body
with a mobile tail. This simple morphology, adapted to
the aquatic environment, has been maintained without
major changes in those animals that carry out external
fertilization. With the advent of internal fertilization,
which probably evolved as a means of increasing the
likelihood of fertilization, structures of sperm evolved
to allow adaptation to morphological complexities of
the female reproductive tract and to moving through
more viscous organic fluids (Cetica and Merani 2004).
Females from different species show a wide diversity of
internal environments and genitalia, leading to a wide
array of sperm shapes (Sivinski 1984).
The shape and dimensions of sperm are extremely
variable, as has been observed in comparative studies
between diverse animal species (Jamieson et al. 1995).
Undoubtedly, this important evolutionary differentiation has occurred because of the intense selection pressure sperm experience in carrying out their reproductive function. Over the last decades, various researchers
have started using sperm morphology as a tool in phylogenetic studies and for analyzing the evolution of sexual reproduction and anisogamy. In particular, sperm
shape and dimensions have been used in these sorts of

studies because they show a certain constancy within
species and also allow for the collection of data without
the necessity of obtaining ejaculates from live animals
(Bell 1978; Afzelius 1979; Parker 1982; Cummins and
Woodall 1985; Ward 1998). Comparative analyses of
sperm morphology have been developed mainly in different groups of invertebrates (Austin and Bishop 1958;
Franzen 1977; Afzelius 1979; Baccetti 1979, 1985, 1986,
1987; Jamieson 1987; Jamieson et al. 1995), whereas
studies on sperm evolution in vertebrates are scarce
(Roldan et al. 1992a; Gomendio and Roldan 1994; Jamieson et al. 1995).
Studies on shape and dimensions of mammalian
sperm are confined to little more than 5% of the species that compose the taxon (Cummins and Woodall
1985; Roldan et al. 1992a). Based on these, it has been
observed that the mammalian sperm cell is organized
into two basic structures: a head containing the nucleus
and acrosome, whose function is to take the paternal
genetic material to the oocyte, and an enlarged tail, in
charge of giving the cell motility (Bedford and Hoskins
1990; Linderman and Kanous 1997). In monotremes,
spermatozoa have maintained the vermiform structure
described for reptiles and birds. In contrast, marsupial
and eutherian spermatozoa generally have oval heads
with tails of variable length (Roldan et al. 1992a; Cetica
and Merani 2004).
In most eutherian mammals studied to date, a high
degree of uniformity is observed in sperm shape and
dimensions. Most species present an oval head and an
average length of 70 m (Cummins and Woodall 1985;
Roldan et al. 1992a; Gomendio and Roldan 1994). This
pattern has been proposed as the basic ancestral shape
of eutherian spermatozoa, and coincides with that proposed in other metazoa (Franzen 1977; Roldan et al.
1992a; Gomendio and Roldan 1994).
Derived spermatic shapes in eutherians have been
described mainly in various groups of rodents and in
some groups of prosimians, where the heads exhibit
more complex shapes and the tail is elongated. Examples of various head morphologies include a hook
shape, presence of caudal extensions, increase in head
size at the expense of the acrosome, and lateral folding
of the apical ridge. In general, it has been observed that
changes in head shape are associated with an increase
in tail length (Roldan et al. 1992a,b). Sperm midpiece
and flagellum lengths are positively associated with
both head length and area (Cummins and Woodall
1985; Gage 1998).
In some analyses, spermatic morphology has been

used as a taxonomic tool; some mammalian species
that were not easy to distinguish by their external phenotype proved to have different spermatozoa (Roldan
et al. 1985; Gordon and Watson 1986; Rouse and Robson 1986; Harding et al. 1987; Prez Zapata et al. 1987;
Breed et al. 1988).
Sperm shape and dimensions in dasypodids

In armadillos, sperm morphology and morphometry
have been studied in 7 of the 8 (possibly 9; Gardner
2005) genera and 10 of the 21 extant species. Data on
the shapes and dimensions of armadillo spermatozoa
were obtained using optical microscopy, transmission
electron microscopy, and scanning electron microscopy (Altman and Dittmer 1962; Nagy and Edmonds
1973b; Heath et al. 1987; Cetica et al. 1993, 1997, 1998).
In our own studies, sperm morphology was evaluated
on a slide using a coverslip and phase contrast microscopy. Size was measured on photographs of smears
stained with Giemsa or silver nitrate and confirmed using a micrometer eyepiece. From transmission electron
microscopy samples, the average sperm nucleus width
was obtained and thus the nucleus volume calculated
Figure 14.1. Frontal plane of a Dasypus septemcinctus spermatozoon. Silver-nitrate staining (X1,126)
145
by multiplying width by the corresponding area (Cetica

et al. 1993, 1997).
Sperm shape and dimensions in species of the genus
Dasypus (D. hybridus, D. septemcinctus, and D. novemcinctus) are very similar among themselves. Sperm
heads are relatively small, oval in the frontal view and
with an ensiform profile (figures 14.1 and 14.2). Sperm
head length and width varies, at 6.68.1 m and 3.84.5
m, respectively, with the acrosomes occupying approximately 5060% of head length. Total sperm lengths of
the three species vary between 63.1 and 68.2 m (figure
14.3).
Sperm shapes in the genera Chaetophractus (Cha.
vellerosus and Cha. villosus), Zaedyus (Z. pichiy), and
Euphractus (E. sexcinctus) are also quite similar to one
another. Sperm heads are very large, being very long
and wide in frontal view, but in profile they are extremely thin. Maximum widths usually occur close to
the equatorial region (figure 14.4). These sperm present
a central concavity, making them similar to a spoon and
allowing them to be piled one on another (figure 14.5).
Heads are paddle-shaped and end caudally truncated.
Length and width of sperm heads are variable, 12.314.5
m and 10.012.7 m respectively, with acrosomes occupying approximately 7073% of head length. Total
sperm length in the four species varies between 71.0
and 89.8 m (figure 14.3).
Sperm shapes in the genera Priodontes (P. maximus)
and Cabassous (Ca. unicinctus) are very similar and, to
a certain extent, share many characteristics with the
genera described previously. Sperm heads are also very
large, the frontal view being long and wide, with an ex-
Figure 14.2. Dasypus septemcinctus sperm head. Scanning

electron micrograph (X10,000).
146
Figure 14.3. Sperm dimensions of the four dasypodid sperm groups (see text for definitions
of each group). Cross = average sperm size of each group, segment = range of sperm sizes
of each group.
tremely thin profile. They present a central concavity,

but their maximum widths are close to the apical region
of the head, giving them a shovel-like shape. Heads
are paddle-shaped and end caudally truncated (figure
14.6). Sperm head length and width varies, at 13.218.0
m and 11.616.0 m respectively, with the acrosomes
occupying approximately 7475% of total head length.
Total sperm length varies between 79.9 and 98.0 m
(figure 14.3).
Sperm shape in the genus Tolypeutes (T. matacus)
presents unique characteristics that distinguish it from
the other groups described so far. The sperm head is

paddle-shaped, of an intermediate size and relatively
thin profile. It presents a central concavity and a characteristic laterocaudal extension in the postacrosomal
region (figures 14.7 and 14.8). Transmission electron
microscopy studies have revealed that this particular
extension contains part of the nuclear material in its interior. Sperm head length and width is 10.9 m and 7.6
m, respectively, with the acrosome occupying approximately 72% of total head length. Total sperm length is
81.4 m (figure 14.3). Remarkably, the sperm tail in this
Figure 14.5. Concave side of a Euphractus sexcinctus sperm

head. Scanning electron micrograph (X9,000).
Figure 14.4. Frontal plane of Chaetophractus vellerosus spermatozoa. Silver-nitrate staining (X1,126).
Figure 14.7. Frontal plane of a Tolypeutes matacus spermatozoon. Silver-nitrate staining (X1,126).
Figure 14.6. Frontal plane of a Priodontes maximus spermatozoon. Silver-nitrate staining (X2,880).
Figure 14.8. Convex side of a Tolypeutes matacus sperm

head. Scanning electron micrograph (X10,000).
148
Figure 14.9. Sagittal section of a Chaetophractus villosus sperm head and

transverse section of a middle piece of
the sperm tail, illustrating the extreme
slimness of the sperm head. Transmission
electron micrograph (X33,700).
species has an oblique insertion at the base of the head,

unlike any other of the spermatozoa studied, where tail
insertion is central.
Sperm patterns in dasypodids

Based on the foregoing morphological analyses, armadillo sperm have been classified into four different spermatic groups. Sperm in group 1, represented by species
from the genus Dasypus, differ markedly from the other
three (Cetica et al. 1998). The pattern in group 1 coincides with that described for the majority of eutherian
mammals (Cummins and Woodall 1985; Roldan et al.
1992a; Gomendio and Roldan, 1994) and also with that
proposed as the ancestral mammalian type (Roldan et
al. 1992a; Gomendio and Roldan 1994).
Sperm patterns for group 3 (Priodontes, Cabassous)
and group 4 (Chaetophractus, Zaedyus, Euphractus) are
very similar, save for a small difference in head shape.
Both groups present a very unique head, in the shape
of a shovel or spoon, which is quite uncommon among
mammals. The heads of these spermatozoa are very
large, with length and width measurements typically
greater than those of other mammals (Cummins and
Woodall 1985; Cetica et al. 1998). In other eutherian
species that have sperm with large heads (e.g., guinea
pigs, chinchillas, shrews, and ground squirrels), this
increase in size depends mainly on an increase in acrosome volume (Fawcett 1970). However, in these armadillos, a thin acrosome covers a large surface of the
nucleus, but does not significantly increase the size
of the sperm head. Therefore, to compensate for this
large area, heads have become extremely thin so as to

maintain a relatively constant nuclear volume (Cetica
et al. 1993, 1997; see figure 14.9). In this way the C-value
(DNA content of the cell), which is known to be very
stable among mammals (Hamerton 1968; Esnault and
Nicolle 1976), remains unchanged. Thus, although average nuclear thickness of sperm from groups 3 and 4 is
very low (0.090.11 m), total volume varies between
10.5 and 16.6 m, similar to that reported for the genus
Dasypus (Cetica et al. 1997; see also figure 14.3). On the
other hand, the majority of mammals with long, and
generally large, sperm heads also have long sperm tails,
as has been described in many rodents and marsupials
(Cummins and Woodall 1985; Roldan et al. 1992a; Gomendio and Roldan 1994; Gage 1998). In sperm from
groups 3 and 4, long and large heads are not accompanied by a marked elongation of the tail, making the
relative length of the heads with respect to total sperm
length one of the largest among those reported for eutherians (16.117.3%; Cummins and Woodall 1985; Cetica et al. 1993, 1997; Gage 1998; see also figure 14.3).
Sperm from group 2 (Tolypeutes) are unique compared to other dasypodids. The sperm head is shaped
like a spoon, in concordance with groups 3 and 4, but
also presents a peculiar laterocaudal extension, giving
it a notable and characteristic asymmetry (Cetica et al.
1997). This extension is very uncommon among eutherian spermatozoa except in some Ctenomys species
(Rodentia) where a postacrosomic process has been
described (Feito and Gallardo 1982a,b). Also unusual
is the oblique tail insertion, although it has been observed in some marsupial sperm (Dalcq 1973). Head
149
Figure 14.10. Proposed evolutionary trend for spermatozoa in dasypodids. The length of the segments
does not represent actual evolutionary distances.
length, width, and thickness are intermediate between

those of group 1 and groups 3 and 4, but tail length and
total sperm length are similar to those in groups 3 and
4 (Cetica et al. 1998).
To summarize, contrary to the majority of eutherian
taxa, sperm morphology is not constant in the Dasypodidae. Sperm from group 1 would be considered the
most ancestral type and would be clearly separate from
the other groups, all of which would be considered derived forms. The sperm from group 2 would be farthest
from groups 3 and 4, which would have differentiated
more recently from one another (Cetica et al. 1998; see
figure 14.10).
Sperm selection in dasypodids

Based on the above we can conclude that there has been
an important evolutionary diversification of dasypodid
spermatozoa, implying strong selective pressures that
affected the morphology of these cells in different armadillo genera (Cetica et al. 1998). Although there is no
conclusive explanation concerning adaptive modifications of sperm morphology in mammals generally or
in armadillos specifically, it has been proposed that two
principal selective forces have favored the evolution of

sperm shape: sperm competition and female selection
(Gomendio and Roldan 1991; Roldan et al. 1992a; Birkhead 1995; Bedford 1999).
Although some studies have suggested that mammalian sperm competition influences sperm shape and
number (Parker 1984; Gomendio and Roldan 1991; Gomendio and Roldan 1994; Anderson and Dixson 2002),
others do not seem to support this hypothesis (Cassinello et al. 1998; Bedford 1999).
Alternatively, it has been proposed that morphological and functional characteristics of the female reproductive tract may be involved in the evolution of
spermatozoa (Smith and Yanagimachi 1990; Roldan et
al. 1992a; Gomendio and Roldan 1994). In animals with
internal fertilization, females present a wide range of
internal environments and genitalia and this could lead
to the evolution of a wide variety of sperm morphologies. Functional, metabolic, and morphological changes
in spermatozoa might have evolved as a consequence of
the need to survive and advance through the female reproductive tract (Sivinski 1984). Although it has been
proposed that sperm shape and dimensions could be
related to different female genital tract morphologies
150
and/or to characteristics of the oviduct and uterine fluids (Smith and Yanagimachi 1990; Roldan et al. 1992a;
Gomendio and Roldan 1994), only a few studies in Coleoptera and Passeriformes have examined these potential relationships (Dybas and Dybas 1981; Briskie and
Montgomerie 1992).
We recently carried out a macroscopic and histological study on the female reproductive tracts of various dasypodids, establishing the main similarities and
differences between them (Cetica et al. 2005). Genital
tract morphology allowed classification of armadillos
into 3 different groups, based on features of ovaries,
uteri, and lower portions of the genital tracts. We observed an association between these groups and the
sperm groupings described above. Group 1 of female
reproductive tracts consisted solely of individuals from
the genus Dasypus, just as with sperm type 1. Similarly,
female group 2 occurred only in the genus Tolypeutes,
just as with sperm type 2. Lastly, female group 3 (Chaetophractus, Zaedyus, Cabassous) was associated with
the closely related sperm types 3 and 4. From these
observations we conclude that particular structural
characteristics of female genital tracts have distinct relationships with sperm morphology in each genus of
Dasypodidae. In armadillos, female reproductive tracts
could represent different structural and functional barriers for spermatozoa that would induce changes in
sperm morphology, suggesting a process of coevolution
between female reproductive biology and spermatozoa
(Cetica et al. 2005).
Future studies directed toward describing the movements of armadillo spermatozoa in fluids of different
densities and viscosities will contribute toward under-
standing other possible selective factors involved in

sperm evolution. Sperm movements have been studied
mainly in sperm possessing the ancestral sperm morphology or in others with hook-shaped heads (Wooley
1979; Cooper 1984). These studies indicate that sperm
motility is affected by subtle changes in the shape of
the head (Dresdner and Katz 1981). Sperm often move
in an undulatory fashion. Armadillo spermatozoa with
large spoon-shaped heads and/or with caudal cephalic
extensions may be unable to locomote effectively in this
way because of hydrodynamic considerations. Because
of their head shape, sperm types 2, 3, and 4 are able to
pile on top of one another. It is possible that this behavior has importance for understanding locomotor adaptations in these groups and constitutes a reproductive
strategy in these species, but, at present, the functional
significance of this behavior remains unknown (Heath
et al. 1987; Cetica et al. 1997).
Final considerations
Sperm morphology can be considered one important
phenotypic trait for the study of phylogenetic distances
among species. For most eutherian mammals sperm
shape and dimensions remain relatively constant between different taxa (except for rodents). However,
these characteristics vary dramatically between different genera of armadillos. In this chapter we have
documented an important evolutionary diversification
in dasypodid sperm morphology. The adaptive significance of these changes must now be identified through
studies of reproductive morphology, physiology, and
behavior in each species.
15
Reproductive biology of the nine-banded armadillo
Richard D. Peppler
oocytes in the concavity of the ovary and that ovulation always occurred there. This area is now referred to
as the ovulation fossa. It takes about five to seven days
for the ovum to pass through the oviduct. If the ovum
is fertilized, a period of delayed implantation follows
usually lasting about 3.5 months, although it can last
much longer (Storrs et al. 1988). The blastocyst during
this time remains in the fundic cavity of the uterus and
receives its nourishment from fluids secreted by glands
in the uterine lining. Shortly after implantation in November, the single blastocyst divides to produce four
embryos attached to the uterus via a common placenta
(figure 15.1). Typically, the four genetically identical offspring are born in the month of April following a 4.5month gestational period. Enders (2002, this volume)
has reported on how the single blastocyst forms these
four offspring. Normally the young born in one year are
reproductively mature by the following summer.
The male reproductive cycle is not as unique as that
of the female. Once males reach reproductive maturity, they are active throughout the year. However, the
spermatogenic cycle is one of the shortest in mammals
(slightly more than 8 days) and it takes only about 33
days for mature spermatozoa to develop.
Resumen
En este captulo se presentan parmetros reproductivos de machos y hembras de Dasypus novemcinctus.
Se reportan estudios de histologa gonadal, hormonales (progresterona, estradiol, testosterona, FSH, LH) y
niveles de plasma en individuos inmaduros y maduros.
Se discuten los efectos de la ovariectoma bilateral o
unilateral y la subsiguiente administracin de clomifeno.
Resumo
Os parmetros reprodutivos das fmeas e machos do
Tatu-galinha (Dasypus novemcinctus) so apresentados
neste captulo. Estudos histolgicos das gnadas e nveis
plasmticos hormonais (progesterona, estradiol, testosterona, FSH, LH) so reportados para animais imaturos
e adultos. Discutem-se tambm, para as fmeas de tatus, os efeitos da ovatiectomia bilateral ou unilateral e
as conseqncias da administrao clomifnica.
Introduction
The reproductive cycle of the female nine-banded armadillo (Dasypus novemcinctus) is very unique. Most
of the pioneering work on this topic has been the result
of investigations by Newman (1912), Hamlett (1932),
Enders (1966), and Peppler and Stone (1980a,b). In the
United States (e.g., Louisiana, Florida, and Texas) the
cycle begins with a single oocyte being shed spontaneously in mid-July and the subsequent formation of
a single corpus luteum. Unlike most mammals, where
the oocyte is shed at random locations along the ovarian surface, Enders (1960c) reported a concentration of
Experimental methods
All the information that follows was derived from study
of animals in the United States, near the northern limit
of the geographic range for D. novemcinctus (Taulman
and Robbins 1996). A colony of armadillos, both immature and mature, was maintained for almost three
years in order to characterize different reproductive parameters in females and males (Peppler and Stone 1976,
1980a,b, 1981; Peppler 1979, 1985; Peppler and Canale
151
152
R. D. Peppler
Figure 15.1. Placenta

from female armadillo
with four young, each
attached by separate
umbilical cord.
Three animals of each sex were killed at 3-month intervals from birth to 27 months of age. Follicular size was
analyzed by measuring two diameters at right angles to
each other in the section containing the oocyte nucleolus. Normal follicles >202 m were counted and categorized into 15 different size groups (every 77 m). Female and male armadillos born in captivity had blood
samples taken monthly from 3 months of age until 27
months of age in order to correlate plasma levels of
progesterone (for females) and testosterone (for males)
with histological changes in the gonads.
antibody. Separation of the free from the bound radioactive hormone was accomplished using a dextrancoated charcoal incubation. Recovery was monitored
by adding known amounts of the labeled hormone to
the samples prior to extraction.
During the middle of each month, 36 mature female armadillos were killed and both ovaries removed
for quantitative histological evaluation. Ovaries were
inspected visually for a corpus luteum (CL) and the
presence or absence of a CL was confirmed microscopically. Follicular size was calculated using the same
method described for immature females. Normal and
atretic follicles >358 m were counted. The lower limit
of follicular size (358 m) was selected because the antrum begins to appear in follicles of this size (Newman
1912).
Mature animals
Circadian rhythm
For a period of three years, female and male armadillos

were captured from the wild and maintained in pens,
under conditions of controlled lighting and temperature, for variable periods of time. Once a week the animals were anesthetized with ketamine hydrochloride
(20 mg/kg) IM and 23 ml of blood was obtained in
a heparinized syringe from a superficial femoral venipuncture. The plasma was removed and frozen until analyzed by radioimmunoassay for progesterone,
estradiol, FSH, LH, and testosterone. The procedure
consisted of a simple ether extraction of the hormone
followed by incubation with a specific antihormone
In four animals that had been in captivity for more

than a year, blood was collected every four hours for 24
hours in mid-July to examine whether diurnal variation in progesterone or testosterone concentrations occurred.
1980; Peppler et al. 1986). The experimental methods

are described in detail in those articles but are abbreviated for the reader as follows:
Immature animals
Results
Immature females
Peppler (1985) and Peppler et al. (1986) described the
pattern of follicular development from birth to 27
months of age in detail. At birth, the ovary contains

only primordial follicles. Secondary follicles >202 m
first appear at 3 months of age. Each ovary contains 16
26 of these sized follicles from 3 to 27 months of age,
although there is a 30% decrease in the number of follicles at 15 months compared with 12 or 18 months. This
decrease is caused by the atresia of some follicles near
the time of ovulation (months of July or August). Normally an ovary contains 46 atretic follicles throughout
the year, except during July and August when the number is doubled (Peppler and Canale 1980). The number
of follicles >358 m in the ovaries of immature females
15 months old (6.9 0.9) is the same as in ovaries taken
from mature females in July and August (6.2 1.3 and
5.8 1.8, respectively). This finding corroborates the
accuracy of follicular counts reported in two separate
studies of animals 15 months of age (Peppler and Canale 1980; Peppler et al. 1986), whether using the lower
follicular size of 202 m or higher of 358 m.
Peppler and Canale (1980) determined that follicles
have to exceed 978 m to be ovulated. Ovaries of some
females may contain some of these large-sized follicles
by 9 months of age; however, other females may not
exhibit these follicles until 27 months old. Even when
these large follicles are present, nests of primordial follicles are still present up to 18 months of age. There are
fewer primordial follicles present at 15 months than at
153
12 months. Similarly, very few primordial follicles are

present at 21 months compared to 18 months. A single
corpus luteum appears in the ovary in 67% of the animals after 15 months of age, and in 100% after 18 and 21
months. In summary, these histological studies indicate
that the earliest an animal born in MarchApril of one
year ovulates is August of the following year, if then.
This conclusion is supported by the observation that
plasma progesterone concentrations do not reach adult
levels until after 15 months of age (Peppler et al. 1986).
In the first year of life, the level is below 5 ng/ml and a
concentration indicative of a functional corpus luteum
(>10 ng/ml) does not occur until 1720 months of age
(Peppler 1985). Collectively, these data substantiate the
report of Enders (1966) that animals born in the spring
are still subadults by the end of the breeding season that
fall.
Immature males
The plasma testosterone concentration in males 30300
days of age is shown in figure 15.2. Plasma testosterone
levels reached the level found in mature animals at 300
days of age and thereafter. In mature males, monthly
concentrations of plasma testosterone varied in the
range of 914 ng/ml, with no seasonal variation (Peppler and Stone 1981). Thus, a male armadillo born in the
Figure 15.2. Ontogeny of plasma testosterone concentration (ng/ml SE) in male armadillos (n = 3), 30300 days of age.
154
R. D. Peppler
month of April of one year appears to be mature by the

next March, in time for the annual breeding season.
Mature females
The reproductive system of the female armadillo is
similar to that of other mammalian species. The ovaries, oviducts, simplex uterus and vagina are within the
pelvis. Unlike some species, accessory adrenal tissue
is usually present in the connective tissue of the mesovarium portion of the broad ligament (Enders and
Buchanan 1959).
Follicular development
Peppler (1979), Peppler and Canale (1980), and Peppler (1985) have described follicular development in
the adult ovary. The total number of follicles >358 m
in diameter remains constant from January to June
(range: 12.818.9) but is 50% less (5.87.0) during the
remaining months of the year, except in October and
November (Peppler 1979, 1999). The number of largersized follicles, >978 m (i.e., those most likely to be
ovulated), is greatest in April, May, June, and October
(2.74.1 follicles/ovary) than during any of the other
months (ranging from 0 in December to 1.6 in March).
One of these follicles will go on to be shed from the
ovulation fossa in July or August and the other follicles
will undergo atresia. During each month of the year
the ovary contains 46 atretic follicles; in July and Au-
gust the number increases to 1012. The percentage of

atretic follicles per total number of follicles >358 m
increases significantly during these two months (65%
versus 20%).
Plasma progesterone and estradiol
An increase in plasma progesterone levels to 10 ng/ml
occurs in June or July (Peppler and Stone 1976; Peppler
and Stone 1980). This increase correlates with the presence of a corpus luteum (n = 30). An early morphological study indicated that luteal cells appear to be inactive
until implantation, at which time secretory droplets are
seen (Hamlett 1935). However, Talmage and Buchanan
(1954) indicated that the corpus luteum is functional
and secretory as soon as it is formed. Similarly, Talmage
et al. (1954) stated that the corpus luteum is active during both delayed implantation and gestation. Their
conclusion was derived from the bioassay of peripheral
serum for progesterone and the total oxygen consumption of luteal and ovarian tissue. Light and electron microscopic studies by Enders (1966) also showed that the
corpus luteum is functional during delayed implantation.
Data on the plasma progesterone levels of female armadillos reveal some variation. In a group of nonpregnant animals (n = 7), plasma progesterone decreased
after three weeks and remained at 56 ng/ml thereafter
from July to May (figure 15.3). Figure 15.4 shows data
Figure 15.3. Annual plasma progesterone levels (ng/ml SE) in nonpregnant adult female armadillos (n = 37).
155
Figure 15.4. Annual plasma progesterone levels (ng/ml SE) in a second group (n = 69) of nonpregnant adult female armadillos.
from a different group of females (n = 9), in which

progesterone increased during November, December,
and January, then returned to baseline for the rest of
the year (FebruaryMay). In three pregnant animals
(figure 15.5), the level of progesterone during the 3.5month delayed implantation period (AugustOctober)
increased to 13 ng/ml and increased again, to 20 ng/
ml, for the entire 4.5-month gestational period (NovemberMarch). Progesterone content of the corpus
luteum does not differ between animals with or without blastocysts, but total progesterone content is lower
during gestation (Labhsetwar and Enders 1968).
Unlike many species that exhibit a decrease in plasma
progesterone around the time of delivery, progesterone
levels in the armadillo remain elevated until almost two
months following parturition. It has been suggested that
estrogen levels may be low during pregnancy. This may
not be the case if the developing follicle is the source of
this hormone. The ovaries of pregnant animals contain
50% more follicles >358 m than nonpregnant animals
(23.7 versus 15.3) during the months of February and
March (Peppler and Canale 1980). Plasma progesterone and estradiol were monitored in 4 nonpregnant
animals from June through December (figure 15.6). Estradiol levels paralleled the levels of progesterone from
July through August and then increased in September.
There was an increase in atretic follicles in July and August, which may have been responsible for the change
in hormone levels at this time.
Daily changes in plasma progesterone were examined by sampling 4 animals over a 24-hour period in
July (figure 15.7). Values ranged between 9.4 and 11.4
ng/ml. Fluctuations were not apparent within the sampling period nor when compared with levels from 365
days previously or 60 days following the sampling period (values were 4.1 and 6.6 ng/ml, respectively).
Plasma gonadotropins
Measurements of FSH plasma levels in 10 adult females
during July and August averaged 1.9 ng/ml, with a range
of 02.5 ng/ml. For LH, the average was 7.9 ng/ml, with
a range of 4.611.5 ng/ml. Because this is the first report
on the levels of FSH and LH in the female armadillo,
interpretation of these results is not feasible until levels throughout the year and during different reproductive states (i.e., increased follicular development, pregnancy) have been measured.
Bilateral/unilateral ovariectomy
Buchanan et al. (1956) demonstrated that implantation can follow bilateral ovariectomy, suggesting that
some nonovarian tissue may assume the function of
Figure 15.5. Plasma progesterone levels (ng/ml SE) in pregnant adult female armadillos during the periods of delayed
implantation and gestation (n = 3).
Figure 15.6. Plasma progesterone (ng/ml SE) and estradiol (pg/ml SE) levels in nonpregnant adult female armadillos (n = 4).
157
Figure 15.7. Daily changes in plasma progesterone concentration (ng/ml SE) among female armadillos sampled over a 24-h
period in July (n = 4).
Table 15.1. Plasma Progesterone Levels in Female Armadillos

Following Unilateral Ovariectomy (ULO)
Day
Plasma Level (ng/ml)
-90
-60
-30
ULO
+7
+14
+21
9.5 2.3
9.3 2.4
6.7 2.0
7.2 2.2
7.7 2.1
7.2 2.3
5.6 2.1
Note: Data are from 3 females. Values reported as means SE.
Thirty to ninety days prior to the surgery the level fluctuated between 6.7 and 9.5 ng/ml. Between 7 and 21
days postsurgery, the level was between 5.6 and 7.7 ng/
ml (the differences were not significant, p >0.05 with
Students t-test). Such would suggest plasma progesterone levels are maintained by a nonovarian source, such
as the adrenal or accessory adrenal gland. Nakakura et
al. (1982) have suggested that captive female armadillos
have a high progesterone concentration attributable to
adrenal production.
Clomiphene injection
maintaining the uterus. If bilateral ovariectomy is done

during the middle of delayed implantation (September
or October), implantation occurs about a month earlier than anticipated. If the operation is delayed, implantation occurs at the normal time but embryos are
resorbed or aborted. Resorption of embryos also occurs if the corpus luteum is removed during unilateral
ovariectomy in early gestation, but not if the corpus luteumcontaining ovary is removed after the first third
of the gestational period (Buchanan et al. 1956; Enders
and Buchanan 1959). Collectively, these findings support the hypothesis that implantation following ovariectomy may be due to some nonovarian tissue producing progesterone to maintain the uterus.
Table 15.1 shows the effect of unilateral ovariectomy
in mid-July on the circulating level of progesterone.
The use of the armadillo as an animal model has been

slow to develop because of lack of information about
its reproductive cycle and failure to breed this species
in captivity (Anderson and Benirschke 1966; Superina
et al. this volume; Truman this volume). Reproduction
in captivity might be facilitated by administration of
agents that promote ovulation. The armadillo ovary will
respond to exogenous and endogenous gonadotropins
(DAddamio et al. 1977; Peppler and Stone 1980b). Figure 15.8 shows the level of plasma progesterone in seven
female armadillos treated with clomiphene citrate (a
nonsteroid estrogen analog used to stimulate ovulation). Six days postinjection, levels of plasma progesterone increased to those indicative of ovulation; corpora
lutea appeared in the ovaries 427 days postinjection.
DAddamio et al. (1977) reported cyclic changes in
158
R. D. Peppler
Figure 15.8. Plasma progesterone levels (ng/ml SE) of adult female armadillos in response to injection with clomiphene
citrate (n = 7).
urogenital smears after gonadotropin administration.

In a similar fashion, urogenital smears changed from
a diestrous pattern prior to clomiphene injection to a
proestrous and subsequently estrous pattern 48 days
after injection.
Mature males
Testis structure and spermatogenesis
The male reproductive system is similar to that of other
mammalian species except that the testes descend only
to the entrance of the superficial ring of the inguinal
canal. Thus, there is no scrotum. The intra-abdominal
location of the gonads has no effect on spermatogenesis
or plasma testosterone levels (as it might in some species) because the armadillo has a low body temperature
(3035C). The structural and developmental changes
of the maturing germ cells within the armadillo testis,
and the morphology of the epididymis, are similar to
those of most mammals (Nagy and Edmonds 1973a,b;
Edmonds and Nagy 1974; Weaker 1977a,b; Torres et
al. 1981). The seminiferous epithelium exhibits cellular associations of mixed composition; most cross
sections of the tubules demonstrate only one type of
these associations. The spermatogenic cycle is divided
into 8 or 10 stages. These stages are based upon (1) nuclei morphology of the spermatids and their position
and arrangement in the seminiferous epithelium; (2)
the presence of secondary spermatocytes; and (3) the
release of spermatozoa into the lumen of the tubule.

These stages occur within the mature testis throughout
the year even though females ovulate only in mid-July.
The duration of each spermatogenic cycle (8.15 days)
is one of the shortest when compared with other species (Torres et al. 1981). The formation of spermatozoa
extends over four cycles, so the entire spermatogenic
process lasts 32.6 days. The interstitial tissue of the testis contains clusters of Leydig cells, connective tissue
elements, blood vessels, and a network of lymphatic sinusoids. There are no changes in the morphology of the
interstitial tissue on a seasonal basis. Overall, testicular
histology is identical in free-ranging and captive males
throughout the year.
Prostate gland
The anatomy and microscopy of the prostate gland
(Weaker 1980) and related ducts in the armadillo is
similar to that of the human with the exception that
some of the glands ducts empty into the vas deferens.
In addition, the ducts of the seminal vesicle and vas
deferens open individually into the prostatic urethra
and do not join before each terminates there. As in the
testis, the histology of the prostate gland does not vary
throughout the year.
Plasma testosterone
Physiological findings confirm these morphological
studies. Testosterone concentrations in weekly plasma
159
Figure 15.9. Daily changes in plasma testoterone concentration (ng/ml SE) among male armadillos sampled over a 24-h
period in July (n = 4).
samples were averaged to obtain individual monthly

values that were then averaged to obtain group values
for respective months. The resulting values varied between 9 and 14 ng/ml each month; differences between
months were not significant (Peppler and Stone 1981).
Circulating levels of testosterone in wild animals were
similar to those observed in captive animals throughout the year. In contrast, Czekala et al. (1980) reported
seasonal variation in circulating testosterone, with the
lowest values occurring in January and the highest in
July. Their values are some of the highest reported for
any species, in the range of 100200 ng/ml. For comparison, the human range is 37 ng/ml. Different assay
techniques might explain the differences in testosterone
levels between these two studies, but do not account for
the discrepancy in findings regarding seasonal variation. The cause of this latter difference remains unknown.
Just as in females, during a 14 month-long examination of plasma testosterone levels in four captive males,
daily fluctuations in testosterone were examined in
samples taken over a 24-hour period in mid-July (figure
15.9). Resulting values ranged between 13.8 and 17.5 ng/
ml. Just as in females, hormonal fluctuations were not
apparent within this time frame, nor when compared
with levels from 365 days previously or 60 days following the sampling day. The finding of constant plasma
testosterone levels throughout the year, in combination
with the histology of the testes, indicates that the male

reproductive system does not undergo seasonal regression (Weaker 1977a,b; Peppler and Stone 1981).
Plasma gonadotropins
The levels of plasma FSH and LH determined in 4 males
in July were similar to those detected in females at the
same time. The average for FSH was 1.9 ng/ml and for
LH 6.5 ng/ml. As stated above, because this is the first
report on the levels of these hormones for either sex,
interpretation of the results is not feasible until levels
throughout the year have been measured.
Summary
The nine-banded armadillo is a unique animal. This
chapter has presented some parameters of the reproductive biology of immature and mature individuals of
both sexes. The potential for use of the armadillo as an
experimental model for investigative studies is unlimited.
Acknowledgments
The author thanks Vicki Murrell for technical assistance
with the graphs and charts and Joann Canale for help
with the collection and tabulation of the data.
16
Reproductive parameters and placentation
in anteaters and sloths
Kurt Benirschke
armadillo. But because of the ancient history of this

mammalian branch (Delsuc et al. 2002), it has been of
special interest. The comparison of placentas from different mammalian orders has been used by Mossman
(1987) and others to deduce how this extremely variable
organ has evolved. Thus the Xenarthra, with their long
and isolated existence, may shed light on the evolution
of the placenta. Moreover, the uterus simplex of anteaters, sloths, and some armadillo species makes their
reproductive physiology of special comparative value
because almost all nonprimates have a bicornuate, or
duplex, uterus. While all xenarthrans share some similarities in reproduction, differences also occur. Thus,
the nondasypodid armadillos have bicornuate uteri,
lack monozygotic litters, and differ in other aspects
such as the expansive development of the fetal adrenal
glands and the absence of delayed implantation. More
information as well as many color photographs are
available at my Web site on comparative placentation
(Benirschke 2005) and can be downloaded. Additional
case material for this Web site would be welcomed by
the author.
This chapter presents the morphology of the placentas from anteaters and sloths. It summarizes the literature on their reproductive biology and data on the
nature of implantation. These species are unusual not
only because of their uterus simplex, but also because
their placentas are both villous and trabecular with
only minimal invasion of the uterus, different from that
of Dasypodidae (see Enders this volume). It is difficult
to obtain material of this kind; thus, the deficiencies in
our knowledge are also discussed.
Resumen
En este captulo se evala la fisiologa reproductiva y el
desarrollo del feto y la placenta del oso hormiguero, el
perezoso de dos dedos y el perezoso de tres dedos. Estas
especies tienen un fundamento comn en la implantacin y un desarrollo similar de la placenta que ocurre
en un tero simple. La placenta es un rgano circular
con una estructura trabecular en la arquitectura de sus
vellosidades, una porcin basal sinusoidal y tiene una
relacin feto-maternal hemocorial. El tero tiene una
capa decidua definida con solo una infiltracin trofoblstica superficial.
Resumo
Este captulo apresenta uma reviso da fisiologia reprodutiva e do desenvolvimento placental e fetal do
tamandu-bandeira, da preguia-de-trs-dedos, e da
preguia-de-dois-dedos. Estas espcies apresentam implantao placentria fundal que ocorre em um tero
simples, e possuem desenvolvimento placentrio similar. A placenta um rgo discide de arquitetura vilosa
com estrutura trabecular, com a poro basal de forma
sinuside, e possui relao hemocorial entre a me e o
feto. O tero possui uma definida camada decdua com
infiltrao trofoblstica superficial.
Introduction
The xenarthran placenta has not often been available
for detailed study, except for that of the nine-banded
160
Reproductive parameters and placentation in anteaters and sloths
The giant anteater, Myrmecophaga tridactyla

General gestational data
The length of gestation is 142190 days with a single
young resulting. One of our term newborn infants at
San Diego Zoo weighed 1,100 g, while the prior pregnancy yielded a premature infant weighing 650 g. In
giant anteaters, estrous cycles occur at approximately
7-week intervals (Patzl et al. 1998). Ovarian cycles and
a pregnancy were examined with fecal steroid analysis
by Patzl et al. (1998). Onset of ovarian cyclic activity
was within 411 weeks after pregnancy, and pregnancy
lasted 184 days; average length of ovarian cycles was 51
days. Progestogens rose above cyclic luteal activity in
the second half of pregnancy and were 20 times higher
during the week before parturition. Estrogens rose significantly in the last third of gestation and, especially,
shortly before birth.
In Tamandua, estrus is believed to be 23 days. Additional observations on the endometrium of tamandua by Becher (1931) are useful. In his large study of
pregnant and nonpregnant uteri, he described the cyclical change of the endometrium; it had many similarities to human endometrium, and periods of estrus
were described with marked secretion of mucus from
the vagina. In later stages of the endometrial cycle, photographs suggest a decidua-like change of the endometrial stroma. Much extravasation was reported in later
stages of the cycle and in early pregnancy.
Implantation
Walls (1939) described the placenta and 10 mm fetus
of the first trimester but no other early stages of implantation of giant anteaters have been available. There
was rather more trophoblast covering the villi in Walls
specimen than is true of term placentas. It was a twolayered structure, composed of cytotrophoblast and
plasmoditrophoblast (syncytium). In his early specimen, the fetal villous capillaries had just become visible. He found no yolk sac or yolk sac membrane, but
the distal intestine protruded slightly into the umbilical cord. The allantoic canal (duct) had nearly closed.
What is difficult to understand is that he and perhaps
Wislocki (1928b)with a different anteater species
to be sureinterpreted the sinusoidal channels to be
made of solid peripheral trophoblast. I interpret them
to be composed of connective tissue with trophoblastic
cover, that is, peripheral extensions of the villi. Becher
161
(1931) made the same observation as Walls and Wislocki. Walls described the uterine artery and vein as
passing through the center of the placenta, and also
showed the fusion of villi with these peripheral sinusoidal trabeculae. Electron microscopy is needed to resolve the questions resulting from these varied descriptions. Similar unresolved issues remain as to the nature
and origin of the trabecular periphery of the placenta
and its relation to the uterine wall.
Homologous features of these placentas with those
of nine-banded armadillos (Dasypus novemcinctus)
include: (1) uterus simplex, (2) fundal implantation,
and (3) similarities in the histologic appearance of the
final placenta, especially the sinusoidal base. Enders
(1960a) summarized publications on armadillo placentation and added his own studies of D. novemcinctus
(see also Enders this volume). He divided placentation
into three stages, with the earliest being the formation
of an inverted yolk sac placenta. Nothing of this structure is left at term and it is doubtful that this exists in
anteaters. He also emphasized the implantation of armadillos over the modified fundal endometrium with
its unique uterine blood sinuses, which are apparently
also present in the giant anteater uterus. The peripheral
trabeculae, once interpreted to be solid trophoblast by
Walls (1939), possess an internal connective tissue core
with trophoblast on the surface. Wislocki (1928b) in his
description of a Cyclopes didactylus specimen (silky
anteater) makes reference to this layer as having maternal tissue . . . undergone a symplasmic transformation. Enders then described the broad trophoblastic
cell columns and red cell precursors in further placental development, all of which are seen in the anteater.
Moreover, as in the term armadillo placenta, the anteater lacks visible Langhans cells (cytotrophoblast)
on the villous periphery; its placenta has the same very
thin syncytial villous covers at term.
General characterization of the placenta
I have had five giant anteater placentas available for
study. One weighed 250 g, two weighed 270 g, another
200 g, and the last weighed 296 gall without cord and
membranes. Two are shown in figures 16.1 and 16.2. In
general, the placenta is disk-shaped, although two had
the appearance of a duplex organ. Two placentas with
single disks measured 16 14 1.5 cm. The maternal
surface had a multicotyledonary appearance with little
decidua covering the villous tissue. The latter was pale
and friable without calcification. The membranes at-
162
K. Benirschke
Figure 16.1. Macroscopic appearance of

a giant anteater placenta with unusually
prominent fetal surface blood vessels. The
two-lobed appearance is more striking in
this specimen than in that of figure 16.2.
Figure 16.2. Gross appearance of another

term placenta from a giant anteater.
tached at the placental margin and the umbilical cords

inserted near the center of the disk. The chorionic
surface vessels were unusually prominent. The uterine attachment site was probably fundal in the uterus
simplex. The placenta had a disk with a chorionic membrane that is similar to that of primates.
The maternal aspect of these placentas is similar to
that of nine-banded armadillos, in that thin bundles
of basal sinusoidal network are anchored to the villi,
without intervening endometrial glands or decidua.
Two large maternal vessels penetrate deeply into the
villous tissue, as was shown by Walls (1939). No data
have been reported concerning the precise manner
of entry of two large maternal blood vessels into the
center of the placenta. This unusual vascularization is
supported by the description of an early anteater pregnancy by Walls (1939).
Because of the unusual complexity of the basal portions of the villous tissue, it is best to illustrate this
region next. This portion of the placenta is basal and
peripheral but is also interspersed with villous tissue.
Once interpreted as solid (cyto-)trophoblast, I view it
more as peripheral placental tissue into which the villi
may expand. This region is often described, and different interpretations have been made of its origin. What
is needed now is a specific genetic analysis of these
trabeculae to determine whether they are maternal
or fetal in origin. This would be easiest by employing
cytogenetic identification of the sex chromosomes in
pregnancies with male conceptuses, or by trophoblast/
endothelium staining with specific antibodies. It is
important to recognize that this layer is shed with the
placenta at delivery. Thus, an origin from endomyometrium alone seems unlikely. Sonographic study of the
pregnant uterus with Doppler techniques in any of

these species would also be of interest as it might better
reveal the pattern of blood flow and display the entry of
the central large vessels.
In order to better understand this trabecular tissue,
sections were stained with Massons trichrome stain
(figure 16.3). They show that these trabeculae have a
central connective tissue core (not myometrium), upon
which rests the presumably single-layered trophoblast.
These trophoblastic cells are much plumper than those
163
covering the villi. In the regions shown with arrows in

figure 16.3D, the peripheral villi, including fetal capillaries, infiltrate the trabeculae and then expand into
them.
Details of the fetal/maternal barrier
This is a hemochorial, villous placenta. The villous syncytial trophoblast is in direct contact with the maternal blood, and Mossman (1987) thus considered it to
be intermediate between the labyrinthine and villous
C.
A.
D.
B.
Figure 16.3. Giant anteater placenta. A. Complete

thickness of mature placenta. Note the basal sinusoidal
region with persistent epithelial covers. Villi attach to
and intersperse among these lamellae. They were once
interpreted to be solid trophoblast. B. Masson trichrome
stain of trabecular region at base of mature placenta.
At the arrow points, one can see the beginning of
villous expansion into the solid trabeculae. C. Higher
magnification of areas similar to B with fetal capillaries
(F.V.) and solid trabeculae (T). D. Transition of villus (V)
into trabeculae at arrow.
164
K. Benirschke
that cluster around the larger fetal villous blood vessels

and likened them to decidual cells but was uncertain of
their function. They are certainly similarly prominent
in anteaters in general. The villous tissue is thicker and
much more pronounced in the central part of the placenta (figure 16.4). At its edges, it is thinner and eventually completely replaced by the sinusoidal tissue.
Umbilical cord and membranes
Figure 16.4. Section taken from near the margin of the giant
anteater placenta, as the villous tissue (left) becomes less
prominent and the sinusoids (right) take over. FV = fetal
vessel, IVS = intervillous space.
types. I believe it to be more typical of a hemochorial

villous placenta, but with an unusual basal portion of
sinusoidal tissue, much as is found in armadillos. The
labyrinthine notion probably stems from the description of putative villous anastomoses (see Wislocki
1928b). But Wislocki was also adamant that this labyrinth is dissimilar to the placental labyrinth of Callitrichids. The villi are anchored to the thin lamellae of
the spongy sinusoids, which do not lose their probably
trophoblastic surfaces, as Enders (1960a) pointed out,
although he considered that they may be endothelial
cells. Villi intermingle diffusely with this spongy region
of the placenta.
I found no direct anastomoses between villi. The
trabeculae connect some villi, as shown in flat sections
through the placenta. No cytotrophoblast is apparent in
the term placentas, while it is seen in the first trimester.
(A similar misinterpretation of the absence of cytotrophoblast was initially made for the term human placenta until electron microscopy showed its presence).
The villous surface of the term gestation placenta has
a very thin layer of syncytium (figure 16.3D). The fetal
capillaries come close to the trophoblastic epithelium,
a feature on which Wislocki commented. He also drew
attention to the unusual, perhaps unique, large cells
The umbilical cords of our placentas inserted near the

centers of the disks and had very few twists. They measured 10, 16, 20, 21.5, and 27 cm in length at delivery but
it is unknown how much cord remained on our neonates, as all survived for several days and remnants may
have fallen off after birth. There are two umbilical arteries and one vein, and no ducts; only once was a minute
remnant of a former allantoic duct seen. The surface is
a single-layered, flat, squamous amnionic epithelium.
Walls (1939) commented on the anteater cord as being
similar to that of human placentas.
The membranes are composed of amnion, with occasionally tall epithelium, and the chorion. The two
layers are fused; there are neither allantoic nor yolk
sac remnants. The amount of decidua capsularis is extremely variable. Walls (1939) found it to be absent in
most areas.
Trophoblast external to barrier and endometrium
There is very little apparent trophoblastic invasion, as
the villi implant on modified basal sinusoids that are
shed with the placenta. But no intact uterus with an
implanted placenta has been available and judgment
must thus be withheld about possible uterine vascular
trophoblastic invasion.
Little decidua is shed with the placenta. Walls (1939)
also found virtually no decidua vera, and only very little
decidua capsularis in his first trimester specimen. Very
little decidual tissue, intermixed with some trophoblast,
is thus found in the delivered placenta (figure 16.5).
Additional remarks and need for future studies
The contribution by Becher (1931) on Tamandua tetradactyla is important in this context as he had a large
number of pregnant and nonpregnant uteri, which he
described as having been well fixed for histologic studies. He pointed out that this is perhaps the reason why
his findings differed in many ways from the interpretation of Wislockis single specimen of Cyclopes didactylus (1928b).
Bechers conclusions about Tamandua are perhaps
165
the individual trabecular elements. Electron microscopy will be mandatory to better understand the contribution of the trabeculae to placental development
and its function.
Two-toed sloths, Choloepus didactylus and

C. hoffmanni
Figure 16.5. At the site of attachment in the giant anteater,

there is little decidua basalis next to the sinusoidal region.
The basal decidua is intermingled with some trophoblast.
Rare glands are present and some hyalinized villi.
equally applicable to the placenta of the giant anteater.

The placenta of tamanduas was always implanted at the
fundus and, with advancing gestation, the uterine wall
became very thin in the region of nidation. In the fundus, a vascular network developed in the musculature
that was visible externally. Becher spoke of a corpus
cavernosum. The maternal surface of the placenta had
a finely lobulated appearance. Neither yolk sac nor
allantois was observed, even in the earliest stages of
pregnancy, quite different from what Enders (1960a)
described for the nine-banded armadillo. The umbilical cord inserted centrally and the length of some cords
was given for immature stages. The study presented an
interesting flat section through the organ that displayed
the mesodermal villi (his choice of words) connected
to one another by the solid trophoblastic trabeculae.
These, Becher found (in contrast to Wislocki), were
composed of large cytotrophoblast and contained no
connective tissue. These sheets of trabeculae hyalinized
toward term and he interpreted them as not becoming
zones of possible future expansion of the villous structures. Their plump cells had no microvillous surface.
Most interesting is his observation that the placenta
was shed in the muscular (!) cavernous zone (not in
the decidua).
There is much work to be done on the nature of the
peripheral zone of xenarthran placentas. Most important will be the identification (maternal versus fetal) of
Pregnancy in C. hoffmanni lasts 11.5 months (309 days)

according to Eisenberg and Maliniak (1978, 1985),
who used special techniques to verify the gestations. It
normally produces one offspring (Meritt 1977, 1985a).
Gender assignment has been difficult and is now aided
by DNA study of hair samples (Murata and Masuda
1996). The placenta was depicted by Soma (1976) and
Benirschke and Powell (1985). No endocrine studies
are known, although the histology of fetal adrenals and
testes is identical to that of the three-toed sloth. Organs
from a 110 g C. tridactylus fetus studied by us included
adrenal glands weighing 0.311 g each, kidneys weighing
0.55 and 0.85 g, and testes weighing 0.065 and 0.058 g.
The fact that the adrenal glands were so large and apparently hyperactive in fetal life suggests that some fetal endocrine stimulus exists in utero, which is also the
case in nine-banded armadillos (Moser and Benirschke
1962). Similarly, the interstitial cell component of the
testis is prominent in fetal sloths.
Eisenberg and Maliniak (1978) were forceful in their
interpretation of gestational length and cited all previous records. While some pregnancies cited were very
long, they dismissed suggestions of sperm storage and
delayed implantation to account for the long gestation.
Neonates weigh 350454 g. Other observers have suggested that the length of pregnancy is shorter for C. didactylus. Veselovsky (1966) recorded birth weights of
376, 408, and 309 g and observed a birth 5 months and
20 days after mating in C. didactylus. Thus there may be
marked differences in reproduction of these two species that merit further attention (see Gilmore et al. this
volume).
Implantation
The exact time of implantation is unknown. The only
useful data come from Heuser and Wislocki (1935) on
three-toed sloths, of which they had several early specimens for histologic study (see below). Implantation occurs in a manner similar to that of apes, deep in the
endometrium. Subsequently, there is much modifica-
166
K. Benirschke
Figure 16.6. Choloepus didactylus gestation from the San

Diego Zoo.
Figure 16.7. Placenta of Choloepus hoffmanni with typically

bulging cotyledons. Some atrophy of lobules is seen at
right. (Courtesy Dennis Meritt, Chicago)
Figure 16.8. Site of implantation in C. hoffmanni,

with septum between the two cotyledons.
tion of the blastocyst shell and of its relation to the endometrium. Two-toed sloths have a pear-shaped uterus
simplex and the endometrium changes in pregnancy to
become classical decidua, within which the placenta
separates at delivery.
The placenta of these two species is essentially the
same and is also similar in Bradypus. It implants at the
fundus in the uterus simplex and has similar bulging
cotyledons on its surface (figures 16.6 and 16.7). I have
had two specimens of C. didactylus (from San Diego
Zoo) and one of C. hoffmanni available (Benirschke
and Powell 1985). Several others are known from the
literature. The first placenta, from which the electron
micrographs stem, came from an animal that died

from salmonellosis. The female weighed 4,040 g and
had a 6 4 cm gastric ulcer. The fetus weighed 110 g;
the placenta weighed 109 g, measured 9 cm in width,
and had a 12 cm long umbilical cord. The fetus was 20
cm in length and had remarkably large adrenal glands
(0.311 g each). The second maternal specimen weighed
5 kg and also died from salmonellosis. The lobes were
separated by thin septa and there was a thick decidua
basalis (figure 16.8). The placenta of Hoffmanns sloth
was provided by Dr. D. A. Meritt of Chicago. It came
from a term delivery, measured 20 16 1 cm and had
a 9.5 cm long umbilical cord.
Becher (1921) considered earlier publications by
Turner on the placenta of Hoffmanns sloth. The find-
Figure 16.9. Cross section through placental labyrinth of C. hoffmanni.

E = endothelium, M = maternal vessel.
ings were essentially identical to his Bradypus specimen. In order not to be too repetitive, I refer the reader
to the subsequent section on the three-toed sloth. I
found no significant differences when comparing gross
and microscopic features. This is a multicotyledonary,
but fused placenta, in which the cotyledons (lobes) are
separated by thin maternal (decidual) septa that contain large maternal blood vessels (figure 16.9). It is more
lamellar than villous, and the placenta is essentially endotheliochorial in nature. There is no vitelline tissue,
and the amnionic sac directly apposes the chorion. The
chorionic membrane has many small fetal blood vessels. The placental location is mostly fundal but extends
onto the posterior uterine surface distally. In my specimen, one lobule was completely infarcted.
Electron micrographs were published by Benirschke
and Powell (1985). While it is true that there are small
remnants of a vascular (maternal) basement membrane, and occasional other remnants of connective tissue persist, the major relationship between mother and
trophoblast is an apposition of syncytium to the hypertrophied maternal endothelium (figure 16.10). The trophoblast is largely syncytiotrophoblast with a microvillous surface. As was often remarked by Wislocki (see
three-toed sloths, below), the maternal endothelium
is prominent and in the villous connective tissue are
167
Figure 16.10. Electron micrograph of barrier in C. didactylus. BM = basement membrane, C.T. = connective tissue.
The latter is the reason this placenta has been considered
syndesmochorial.
numerous pigmented epithelioid cells, interpreted to

be macrophages filled with an electron-dense granular material (figure 16.11). They may be equated to the
Hofbauer cells of primates. The nature of the granular
content is unknown.
There is no extravillous trophoblast, and the invasion of maternal space is limited to the deciduas, where
the trophoblast destroys decidual cells; this accounts for
much of the debris found at the base. It then remodels
the decidua and maternal blood vessel walls to form the
trabeculae, or lamellae, which are apposed by the fetal
components. Maternal blood vessel invasion does not
occur.
Umbilical cord and extraplacental membranes
The umbilical cord has two arteries and one vein. It is
about 10 cm long at term and is not twisted and has
no duct remnants. The surface is smooth and clad with
thin amnionic epithelium.
There is a thick decidua basalis within which the placenta separates at delivery. The membranes have a definite decidua capsularis outside the chorion laeve. The
amnion is also very similar to that of primates, being
very thin and avascular. No allantoic sac exists at term
and, in the young specimens of Bradypus examined by
Wislocki, the allantoic diverticulum was rudimentary
and disappeared quickly.
168
K. Benirschke
Figure 16.11. At arrows are the so-called pigmented epithelioid cells (= Hofbauer cells) of
the villous stroma in C. didactylus.
Three-toed sloths, Bradypus tridactylus, B.

torquatus, and B. variegatus
A single offspring delivers after a gestation of 106 days.
Wislocki (1927), who had the most material, never saw
more than one implanted embryo. Newborns weigh
200250 g. The placental weight of term organs is
around 6080 g.
Implantation
Implantation occurs mainly at the fundus of the pearshaped uterus. At first, the entire conceptus is covered
with villous/lobular tissue. Later in development, however, the villous tissue atrophies in much of the lower
portion (over the cervix and some anterior uterine
parts) of the uterus. Implantation occurs with invasion
of the decidua.
Heuser and Wislocki (1935) studied four early gestations in B. griseus (now considered B. variegatus) and
remarked that the anterior portion of the uterus was
larger, the posterior flatter and thinner. The decidua
also had a different appearance in this material. The
allantois was rudimentary and observed only in serial
sections of the body stalk. The yolk sac was prominent
and was described in detail. No omphaloplacental circulation ever developed because the yolk sac detached
from the chorion very early. The entire blastocyst surface invaded the endometrium, removing its epithelium in the process. As a result, the maternal vessels
hypertrophied significantly. While the trophoblast was
initially cytotrophoblast, it soon became primarily and

then exclusively syncytium. This appears to be the invasive cell type, but it never breached the maternal vascular endothelium. The invasion of syncytial trophoblast
into the decidua is best described by Wislocki (1927).
He opined that the trophoblast destroys decidual cells
and portions of the maternal vascular walls. He thus
considered the maternal tissue to be gradually transformed into the lamellae of the labyrinthine tissue.
Becher (1921) provided the first comprehensive description. He reviewed earlier contributions on two-toed
sloths and highlighted the differences in their structure. He also reviewed the inaccessible contribution
by Klinkowstrom (1895) who described three developmental stages. In the youngest stage, chorionic lobules
were distributed all over the chorion and only later
did they concentrate into 1 or 2 connected lobes at the
fundus. The fetus of Bechers term specimen was 10 cm
long. The placenta had the usual lobular projections on
the fetal surface which he equated with cotyledons. He
and others were much impressed by the ease of placental separation in the decidua. The lobules (cotyledons) were separated one from another by decidual
septa. Moreover, he discussed numerous giant cells that
he considered to originate from blood cells, but this is
probably incorrect. The chorion and villi were covered
by syncytium lying directly adjacent to the maternal
vascular endothelium, making this an endotheliochorial placenta.
Wislocki (1925) made the next important observa-
tions on a specimen of B. griseus (B. variegatus). The

fetus of his specimen was 24.5 cm long. The lobulated
cotyledonary tissue covered three-quarters of the uterine surface (bell-shaped) and was constructed of two
adjoining halves. The cord inserted centrally between
the two lobes and had a smooth surface. He spoke of
placental lamellae and considered the designation
villi to be inappropriate. Wislocki found a trophoblastic syncytium surrounding the maternal blood
vessels, which were somewhat larger than capillaries,
in the labyrinth and interpreted this placenta to have
a syndesmochorial relationship because a thin layer of
connective tissue remained around many maternal vessels. In all the descriptions thus far, much emphasis was
placed on the coexistence of many regressive changes
occurring with atrophy of whole cotyledons.
Subsequently, Wislocki (1926) described 14 pregnant
uteri of various stages. Much like the fourth membrane
of the camelids, the sloths have a thin epitrichium, a
delicate membrane attached to the nostrils, anus, toes,
and other organs. He later (1927) elaborated on these
specimens and became convinced that the placenta is
an endotheliochorial organ, likening it to the barrier of
carnivores. In reply to de Langes paper (1926), Wislocki
(1928c) emphasized that the trophoblast was syncytial,
without evidence of cytotrophoblast, and strongly disputed de Langes assumption of sloths having a hemochorial placenta.
De Lange (1926) had two uteri and opined that the
trophoblast relation to the uterus was mixed, between
endotheliochorial and hemochorial. King et al. (1982)
laid all these arguments to rest with a study by electron
microscopy. In young specimens they found a syndesmochorial relation but the more advanced specimens
Figure 16.12. Bradypus tridactylus uterus with fetus attached

to placenta by short umbilical cord.
169
were definitely endotheliochorial and comparable to

the shrew placenta. This conforms to our observations
(Benirschke and Powell 1985). Two other features are
notable: (1) the numerous large macrophages within
the villous connective tissue, described as epithelioid
cells by Wislocki (1927), and (2) the unusually large
endothelium of the labyrinthine maternal vessels.
Through the generosity of Z. Silva and W. Jorge in
Brazil I was able to examine one complete uterus of
a three-toed sloth conceptus (Benirschke and Powell
1985; see figures 16.1216.14). It contained a fully grown
male fetus (27 cm long, 185 g), attached by a 15 cm umbilical cord to a multilobulated placenta weighing 60
g and measuring 11 7 cm. It had a marginal infarct
(an area of probably normal atrophy). About 20 placental lobules had fused to a single disk of placental
tissue (figures 16.1216.14). The membranes made up
approximately 50% (by weight) of the whole specimen.
The fundally implanted placenta was readily detached
from the single uterus simplex and measured 12 5 cm.
The fallopian tubes were 7.5 cm long; the distance between their origins was 4 cm. The ovaries were large
and located in an 8 3 cm bursa.
The term placenta is an endotheliochorial organ. Maternal blood vessels, lined by hypertrophic endothelial
cells, are surrounded by trophoblastic syncytium (figures 16.15 and 16.16). In younger stages of development,
the syncytium is seen to destroy the decidua and a remnant of connective tissue surrounds the vessels. Techni-
Figure 16.13. Opened uterus of three-toed sloth (B.

tridactylus) with fetus deflected to the right. It shows that
the lower portion of the formerly cotyledonary tissue is
undergoing atrophy (white areas at bottom of placenta,
at arrows).
170
K. Benirschke
Figure 16.14. Portion of the

attached placenta from a
three-toed sloth (B. tridactylus)
with the uterus below. The
two protuberant lobules are
evident. They are separated by
the septum (S) composed of
decidual tissue.
cally speaking, then, the placenta is a syndesmochorial

organ at this time.
Umbilical cord and extraplacental membranes
The cord of our specimen was 15.5 cm long, had few
spirals, and contained two arteries and one vein and
no ducts. Wislocki (1926) found the cord to be attached
to the midsagittal posterior plane of the uterus. It has a
Y-shaped bifurcation shortly before its placental inser-
tion, with the vessels then running to the two nearly

joined lobes.
Becher (1921) found numerous small amnionic nodules in his young specimen, not present in our term placenta or in Wislockis specimen (1925). Later, Wislocki
(1926, 1927) observed amnionic villi, small projections
of connective tissue that made bulges in the amnionic
epithelium. The amnion consisted of a thin, single-layered epithelium. Wislocki (1927) stated that the portions
Figure 16.15. The relation of a villous/

lamellar fetal structure in B. tridactylus
with its villous connective tissue (VCT)
that harbors numerous epithelioid
cells, corresponding to macrophages
or Hofbauer cells. The endothelium
(E) of the maternal vessel is hypertrophic.
171
Figure 16.16. Higher magnification of

maternal-fetal relation in B. tridactylus. MV
= maternal vessel. Note how closely the
trophoblastic syncytium is attached to the
endothelium.
Powell 1985); the amnionic connective tissue had numerous pigmented macrophages. Meconium, however,
has never been observed. Large quantities of decidua
capsularis and basalis separate with the placenta (see
Becher 1921). Wislocki (1925) found large glands early
in gestation but they appeared to shrink subsequently.
Although the yolk sac is a prominent structure in early
gestation, vitelline remnants have not been identified at
term.
Uteroplacental circulation
This has not been studied in any detail. The maternal
blood vessels of the decidua are surrounded by trophoblastic syncytium and have no ingrowth of trophoblast.
Physiologic studies or perfusion experiments have not
been performed.
Figure 16.17. Edge of the placenta of the three-toed sloth,
B. tridactylus, with the membranes showing decidua capsularis on the outside, very similar to human placentas.
of the membranes that lacked cotyledons (where they

had regressed, would be a better description), always
contained blood vessels (figure 16.17). We found in addition to the large normal fetal blood vessels a host
of tiny vessels in these membranes (Benirschke and
Summary
This contribution describes the placentas of the giant
anteater and the two- and three-toed sloths. These species possess a uterus simplex, with generation of single
fetuses the rule. The placentas differ slightly and have
specialized features of endometrium (decidua) and trophoblast. More electron microscopy of sloth placentas
is needed and it is desirable that Doppler sonographic
studies provide insight into placental circulation.
17
Placentation in armadillos, with emphasis on
development of the placenta in polyembryonic species
Allen C. Enders
Resumen
Resumo
Un gran nmero de especies de armadillos son monotocas; algunas tienen mellizos no idnticos, pero al menos
dos especies del gnero Dasypus presentan poliembrio
na. Todas las especies estudiadas hasta el momento
poseen una placenta definitiva en la que las vellosidades
placentarias estn baadas directamente por la sangre
materna; las vellosidades estn cubiertas por una sola
capa de sinciciotrofoblasto, con un citotrofoblasto confinado a los extremos en crecimiento. La implantacin
y formacin de la placenta han sido estudiadas ms
ampliamente en D. novemcinctus. En esta especie el
blastocisto se implanta en el fundus del tero simple,
donde los senos sanguneos maternos estn prximos
a la superficie del endometrio. Las vellosidades en formacin penetran en estos senos; al mismo tiempo la
prdida del trofoblasto anembrionario invierte el saco
vitelino, exponiendo el endodermo del saco vitelino al
lumen uterino. Subsecuentemente, las vellosidades se
extienden en la sangre materna dentro de los senos en
el endometrio del cuerpo uterino, mientras que el saco
vitelino y el embrin en desarrollo se expanden dentro
del lumen uterino. Se sugiere que las ventajas de este
mecanismo de implantacin y placentacin son suficientes, por lo tanto ellas fueron mantenidas cuando
las condiciones fueron adecuadas para el desarrollo
de cras mltiples en Dasypus. La retencin de estas
relaciones debe haber favorecido la poliembriona, en
oposicin a la poliovulacin, como un mecanismo para
producir cras mltiples.
Um grande nmero de espcies de tatus so monotocas; algumas no produzem gmeos idnticos, porm
pelo menos duas espcies de Dasypus possuem poliembrionia. Todas as espcies estudadas at o momento apresentam uma placenta definitiva, na qual as
vilosidades placentrias so banhadas diretamente pelo
sangue materno; as vilosidades so cobertas por uma
camada simples de trofoblastos sinciciais, com o citotrofoblasto confinado aos extremos em crescimento.
A implantao e formao da placenta tem sido estudada mais extensamente em D. novemcinctus. Nessa
espcie, o blastcito implanta-se no fundo do tero
simples, onde os seios de sangue maternal so prximos superfcie do endomtrio. As vilosidades formadas penetram nesses seios; ao mesmo tempo a perda
do trofoblasto abembrinico inverte o saco vitelnico,
expondo a endoderme do saco vitelnico para o lmen
uterino. Subsequentemente, as vilosidades se extendem
no sangue materno entre os seios no endomtrio no
corpo do tero, enquanto o saco vitelnico e o embrio
em desenvolvimento expandem-se dentro do lmen
uterino. Sugere-se que as vantagens desse mecanismo
de implantao e placentao so suficientes, e foram
retidos quando as condies favoreceram o desenvolvimento de descendncia mltipla Dasypus. A reteno
dessas relaes devem ter favorecido a poliembrionia,
em oposio a poliovulao, como um mecanismo para
se produzir mltiplas crias.
172
Placentation in armadillos
Introduction
Reproduction in the armadillo has been of interest to
scientists for more than 100 years. The seemingly primitive nature of the animal, but especially the formation
of identical siblings in species of one genus, attracted
attention to gestation. Nowak (1991) identified 20 species of living armadillos; currently, 21 are recognized
by the IUCN (Aguiar and Fonseca this volume). The 7
species in the genus Dasypus are believed to produce
identical young (Galbreath 1985), but only D. novemcinctus and D. hybridus have been extensively studied.
Species in at least four other genera can produce nonidentical twins; all other species produce just a single
young (Galbreath 1985). The definitive placenta of the
armadillo is of the villous hemochorial type (Adamoli
et al. 2001). Although this conclusion is based on observations of species in at least four different genera, most
additional information on placentation comes from the
genus Dasypus and the two species first discovered to
produce identical young.
Description of both early development and placentation in D. hybridus was undertaken by Fernandez
(1909), who reported polyembryony in this species, illustrated conceptuses with six and nine embryos, and
stated that as many as 12 identical embryos could occur.
Newman and Patterson (1909, 1910) described development of D. novemcinctus, again with emphasis on the
question of polyembryony. A more complete description was provided by Patterson (1913). Although polyembryony has received the greatest amount of interest,
the nature of placentation in Dasypus species is in itself
quite extraordinary and deserving of study. Such studies are important in their own right, but also because
we must understand development of the placenta in
order to achieve insight into the possible evolution of
polyembryony in armadillos.
173
continue development to any great extent during the

period of delayed implantation (Hamlett 1935; Enders
1962). During this delay, the single spherical blastocyst
remains in the fundic region of the uterus, has a distinct inner cell mass, but does not show differentiation
of an endodermal layer or indication of its later formation into identical quadruplets. In Texas and Louisiana
(U.S.A.), fertilization commonly occurs early in the
summer and implantation around November (Hamlett
1932; Enders 1966; Peppler this volume).
Implantation of the blastocyst

The following description of implantation, formation
of embryos, and placentation is based on a series of
studies of D. novemcinctus (Enders 1960a,b, 1962, 1963,
2002; Enders and Welsh 1993). The fundic region of the
uterus has a relatively thin layer of endometrium overlying a series of anastomotic venous sinuses (Enders et
al. 1958; Enders 2002). These sinuses are interconnected
with a series of basal sinuses in the endometrium of
the body of the uterus as well (figure 17.1). Reasons that
armadillos engage in delayed implantation and what
Pre-implantation stages of development

The nine-banded or long-nosed armadillo, D. novemcinctus, ovulates a single egg from the ovulation fossa of
the ovary (Enders and Buchanan 1959; Peppler this volume). This egg traverses the oviduct during cleavage;
however, when the morula reaches the confluence of
the oviduct with the uterus simplex, folds of the endometrium retain the developing blastocyst in the fundic
region of the uterus (Patterson 1913; Enders and Welsh
1993). Although the zona pellucida is lost shortly after
the blastocyst reaches the uterus, the blastocyst does not
Figure 17.1. Sagittal section of an armadillo uterus prior to

implantation. The position of the blastocyst is shown by the
black dot adjacent to the asterisk in the fundic region. Note
the basal blood sinuses (black) within the endometrium.
174
A. C. Enders
Figure 17.2. An early implantation stage in the fundus of the uterus. A. The uterus has been
everted to show the implanting blastocyst. The epiblast shows as a brighter area within the
trophoblast. B. A section of the blastocyst seen in A. Note the large single amniotic cavity, and
several vesicles below that will form the exocelom. The blastocyst cavity has collapsed but the
trophoblast (arrow) is still intact, and a small space between it and the endoderm of the yolk
sac appears on either side. The trophoblast has attached to the uterus; note the maternal blood
sinuses (*) below the site.
terminates this period remain unclear; however, when

implantation finally begins, trophoblast from the polar
region of the blastocyst penetrates the uterine luminal
epithelium (Enders 1960a, 2002). Laterally, syncytial
trophoblast masses interdigitate with the uterine luminal epithelium and underlying connective tissue, anchoring the implanting blastocyst to the endometrium.
Centrally, individual trophoblast giant cells surround
the endometrial glands (which subsequently disintegrate), and trophoblast also penetrates to the underlying venous sinuses.
Shortly after the initiation of implantation, an amniotic cavity develops by cavitation (Patterson 1913).
Clusters of cells appear between the amnion and the
polar trophoblast, becoming small vesicles that coalesce
to form an exocelom lining the amnion and trophoblast
(Enders 1960a; see figures 17.2, 17.3). The blastocyst cavity collapses; subsequently the abembryonic trophoblast
disintegrates, exposing the underlying endoderm. The
endoderm has developed from the inner cell mass but
has not proceeded more than a short distance beyond
its margin (Enders 2002). The loss of the abembryonic
trophoblast inverts the yolk sac, exposing the endoderm to the uterine lumen (figure 17.4).
At this stage of development, there is a single epiblastic plate and a large exocelom; the trophoblast
forms a ring around the margin of the site, the trophoblastic annulus, formed of multiple trophoblast cells. A
space develops between the trophoblast adjacent to the

exocelom and the larger trophoblast cells that invade
the endometrium, surrounding the uterine glands (figures 17.4, 17.5). When the venous sinuses are tapped,
blood enters this space, which Fernandez (1914) called
the trager cavity; the layer of trophoblast adjacent to the
exocelom becomes largely unilaminar, with multiple
microvilli extending into the blood spaces. This layer
of trophoblast is designated the absorptive trophoblast
(Enders 2002), as opposed to the nonmicrovillous invasive trophoblast that penetrates farther into the endometrium.
Formation of the embryos

Only after considerable expansion of the conceptus
does there begin to be a condensation of cells within
the epiblastic plate, forming first the beginning of two
embryonic shields and then two more at approximately
right angles to the first two (Enders 2002; figure 17.6).
These shields form at a stage when there is still a single
large amnion. Subsequently, the expanding exocelom
tends to collapse the amnion onto the surface of the
embryonic shields and reduces the region between the
developing amnions to four slender amniotic ducts
and a central amniotic vesicle. Later the amniotic
ducts close, leaving four separate amnions. It is only
after the wide separation of the embryonic disks and
Figure 17.3. Diagram of an early

implantation site. A ring of trophoblast
attaches to the uterine epithelium at
the margins. Individual trophoblast
cells migrate around the uterine
glands and penetrate the underlying endometrial venous sinuses.
The trophoblast across the center
of the site becomes modified with
some syncytial trophoblast and some
cytotrophoblast, and is polarized
with a microvillous surface facing the
maternal blood.
Figure 17.4. A more advanced implantation site after loss of the abembryonic
trophoblast. At this stage the endoderm is directly exposed to the uterine lumen. The uterine glands (dark) have been surrounded by trophoblast cells but
have not yet disintegrated. It is only at this stage that areas within the epiblast
(above the amniotic cavity) begin to be organized into individual embryonic
shields.
Figure 17.5. Higher magnification showing numerous

trophoblast cells around the dark gland cells. The single
layer of trophoblast overlying this area is absorptive trophoblast and is the region where placental villi will form.
176
A. C. Enders
Figure 17.6. Two embryonic shields.

The amniotic cavity has collapsed
but it is still continuous between
the forming embryos.
Figure 17.7. A. Portions of four

embryonic shields (arrows) are
seen in this section through the
developing conceptus. Note
that the increase in size of the
exocelom has widely separated
the embryos. B. Higher magnification of the embryo shown at
the bottom of A. Note the area
of the developing primitive
streak (*) between the epiblast
of the embryonic shield and
the endoderm below.
the restriction of the amnion to amniotic ducts that the

first primitive streaks develop between the epiblast and
the underlying visceral endoderm of each embryonic
shield (figure 17.7).
Formation of the placenta

As the embryonic shields are developing, the continued growth of the conceptus and increase in size of
the exocelom push the vesicle into the uterine lumen,
where the endodermal cells of the inverted yolk sac are
directly exposed to the uterine luminal epithelium and
luminal content (figure 17.8).
Condensations of cytotrophoblast appear within the
absorptive trophoblast and intrude into the underlying
blood space, initiating the formation of placental villi,
that is, primary villi. The surface of these clusters has
a thin covering of syncytial trophoblast, and mesenchymal cells lining this portion of the exocelom soon
indent the primary villi, converting them to secondary
villi (figure 17.9A). The villi thus formed expand laterally into the maternal sinuses peripheral to the fundic
implantation site. As they continue to develop, the cytotrophoblast remains as clusters of proliferating cells
at the tips of the advancing villi, but the rest of each

villus, although covered by syncytial trophoblast, has
no underlying cytotrophoblast layer (figure 17.9B). Mitotic figures are confined to the cytotrophoblast, which
therefore provides a supply of cells to the forming syncytial trophoblast as the villi advance. Between the
individual cytotrophoblast cells at the tips of the villi,
a granular extracellular material is found. The mesenchymal core of the villi becomes vascularized and is at
first a rather loose connective tissue. The advancing villi
are often somewhat flattened or leaflike as they develop.
Later, the arborization is more similar to that seen in
other villous hemochorial placentas.
As the conceptus with its four developing embryos
enlarges, the fundic region of the uterus is expanded.
The villi extend farther into the sinuses of the body of
the uterus (figure 17.10), and as development continues
the villi in the fundic region degenerate, leaving a thin
blood space in the fundus and a circumferential band
of villi within the maternal venous sinuses in the body
of the uterus. Anderson and Benirschke (1963) demonstrated that by the time the circulation of the fetuses has
developed sufficiently to enable umbilical injection, the
vascular domains of the villi for the individual fetuses
177
Figure 17.8. A. Depiction of a developing placenta within the uterus.

The inverted yolk sac intrudes into
the uterine lumen exposing the layer
of endodermal cells to uterine gland
secretions. The expanding conceptus has enlarged the fundus of the
uterus and the villi of the chorioallantoic placenta (dotted) are spreading into the sinuses (black) of the
body of the uterus. B. Hemisected
uterus through the exocelom of the
conceptus showing two of the four
developing embryos closely covered
by their amnions. A bit of the luminal
surface exposed to the inverted yolk
sac is seen (*).
Figure 17.9. A. Section of placental villi starting to develop

along the surface of the absorptive trophoblast. A cluster of
cytotrophoblast is seen at the tips; a thin layer of syncytial
trophoblast (arrow) covers each villus. Mesenchymal cells
are invading the villi, converting them to secondary villi.
B. Villi at the stage when they are invading the endometrial sinuses in the body of the uterus. Note the cluster of
cytotrophoblast cells at the tip of the villus, and the lack of
cytotrophoblast under the syncytial trophoblast (arrow)
covering the villus. Uterine gland (gl).
are separate. Consequently, the placentas are discoidal

in structure. It should be noted that, just as embryonic
disk and primitive streak formation are relatively late,
so also is the development of circulation within the placental villi.
The villi entering the blood sinuses in the body of
the uterus expand these sinuses, converting them to
intervillous spaces. This results in villi being situated
within maternal sinuses that retain their endothelial
lining, and are within arcades of endometrium (figure
17.11A). Maternal arterial blood passes in endometrial
septae between clusters of villi, entering the venous
sinuses from the endometrium superficial to the villi
(figure 17.11B). The maternal blood then drains to many
small channels at the base of the venous sinuses, entering a meshwork of veins within the myometrium.
At the same time, the inverted yolk sac is still situated
in the uterine lumen in a position to absorb materials
from the endometrial epithelium and glands.
As pregnancy progresses, the endometrial sinuses
that have been invaded become greatly enlarged. The
endometrium is pressed against the yolk sac on one
side, and the growing embryos with their amnions intrude into the exocelom and press against the yolk sac
from the luminal side, apparently causing atrophy of the
endoderm of the inverted yolk sac. The villi within the
sinuses also undergo progressive changes. The number
of cytotrophoblast tips is reduced and they eventually
disappear. The fetal stromal cells within the villi hypertrophy, become epithelioid, and develop extensive regions of granular endoplasmic reticulum (figure 17.12).
The syncytial trophoblast of the definitive villi has
abundant microvilli on its free surface and regions of
invagination of the basal cell membrane in patches
178
A. C. Enders
Figure 17.10. A. Sagittal section showing villi (*) penetrating from the area
of attachment (upper left) into the
venous sinuses of the endometrium.
B. Higher magnification of the area of
attachment of the inverted yolk sac
to the endometrium (endom). Note
the villi within the endometrial sinus.
The cytotrophoblast (*) is restricted
to the growing tips of the villi. The
endoderm (arrow) of the inverted yolk
sac is directly exposed to the uterine
lumen.
Figure 17.11. A. Placenta in late pregnancy. The numerous villi have greatly
enlarged the endometrial sinuses to
form the intervillous space. Note the
endometrial septae (arrows) containing vessels that extend from the base
of the endometrium to the thin layer
of endometrium overlying the intervillous space. B. Higher magnification
of a late placenta. A maternal vessel
(mv) in a septum is opening into the
intervillous space (ivs). The fetal connective tissue (fct) is a remnant of the
inverted yolk sac overlying the darker,
thin layer of endometrium. Note the
lack of vascularity of this fetal layer
and that the original endodermal covering, as well as the space between
the yolk sac and the endometrium, is
gone. Villi (v).
overlying the basal lamina (figure 17.12B). Numerous

invaginations and small vesicles are found near the
luminal surface of the trophoblast. Thus, the syncytial
trophoblast resembles that of many other hemomonochorial placentas.
Presence of the developing villi of the chorioallantoic placenta in maternal blood sinuses assures a continuous pathway for circulation of maternal blood. At
the same time, inversion of the yolk sac provides excellent access to products within the uterine lumen. The
retention of the blastocyst in the fundus places it not
only in the ideal area for implantation but also in the
most suitable site for the dual nature (chorioallantois
and yolk sac) of placental development.
Although placental villi were observed within spaces

in the endometrium in some of the first histological
studies, the way in which they penetrated into these
spaces and the significance of yolk sac inversion were
not completely understood. Thus Strahl (1913) and de
Lange (1933) depicted the villi situated in spaces within
the endometrium, but had trophoblast rather than endoderm exposed to the luminal surface of the uterus.
Newman and Patterson (1910) mistook the sinuses for
lymph spaces situated between the wall of the vesicle
and the uterine mucosa. Strahl (1917) sectioned a number of placentas in situ, and properly determined that
much of villus development occurred within the basal
endometrial sinuses of the body of the uterus. Unfor-
179
Development of polyembryony
Figure 17.12. A. A single mature villus. Note the close

association of the fetal vessels (arrows) to the syncytial
trophoblast covering the villus. Enlarged mesenchymal
cells (m) are abundant in the stroma of the villus. B. Electron
micrograph of a portion of a mature villus. Note the numerous microvilli and vesicles at the surface of the syncytial
trophoblast. Note also the dilated endoplasmic reticulum in
the fetal mesenchymal cell below.
tunately, he did not have the early stages and published

his findings in a somewhat obscure journal.
The nature and role of the intercellular material
found between cytotrophoblast cells at the tips of placental villi has not been clarified. Epithelioid mesenchymal cells with abundant rough endoplasmic reticulum, such as those found in armadillo placental villi,
are also seen in the placenta of such diverse species as
the tree shrew (Luckett 1968; Kaufmann et al. 1985),
short tailed shrew (Wimsatt et al. 1973), and pygmy otter shrew (Carter et al. 2006). As yet, the function of
these cells is unknown.
As Loughry, Prodhl, McDonough, and Avise (1998)

have pointed out, polyembryony has a number of disadvantages, including particularly the generation of
offspring with only one genotype to confront environmental changes. How then did polyembryony develop
in the genus Dasypus?
Many answers to this question have been proposed
over the years. Some of the historic suggestions were
quite fanciful and not based on knowledge of the way
in which the identical embryos developed from a single
blastocyst. For example, Rosner (1901, cited in Newman
and Patterson 1910), writing before information on development was available, suggested that there was a
fusion of four follicles within the ovary prior to ovulation. Newman (1917) and Stockard (1921) suggested that
delayed implantation may have resulted in deprivation
of oxygen to the blastocyst, resulting in polyembryony.
However, Hamlett (1933) pointed out that other species
with delayed implantation do not have polyembryony
and that there are examples of polyembryony in species
that do not delay. He therefore concluded that polyembryony was genetically controlled in each species.
Galbreath (1985) suggested that an ancestral armadillo with a uterus simplex and a single embryo preceded the development of polyembryony. He further
suggested that later changes produced a situation in
which an increase in litter size was both adaptive and
energetically possible. He used the observation of Buchanan (1967) that a second, unimplanted, blastocyst
was found in the uterine lumen of a female with an
implanted blastocyst in the fundic area, together with
the fact that, although females with more than one
corpus luteum had been found, none had been found
with more than one implantation site, to suggest that
the uterus simplex of D. novemcinctus was incapable of
supporting more than one blastocyst. He did not, however, give the biological reasons that the uterus could
support only one implantation site. Others have suggested that there might be room in the fundus for only a
single blastocyst (Loughry Prodhl, McDonough, and
Avise 1998). However, not only is there room for a blastocyst on both sides of the fundus, but in addition the
endometrial folds in this position are readily pushed
aside during implantation when the implanting blastocyst expands into the body of the uterus. Buchanan
(1967) suggested that after the first blastocyst expanded
a decidual response prevented further implantation
and that only the fundic area is capable of supporting
180
A. C. Enders
implantation. In agreement with Benirschke (2005),

no true decidual response, that is, conversion of stromal cells to decidual cells, is found in the armadillo.
Changes in the surface of the uterus in the fundus, or
other uterine alterations, could nevertheless result in an
inability of a second blastocyst to implant. Thus, asynchronous response of more than one blastocyst to the
uterine stimulus to implant following delayed implantation could result in failure of implantation of a second
blastocyst. Alternatively, the expansion of the implanting blastocyst might expel an additional unimplanted
blastocyst into the lumen of the uterus. No studies of
the nature of the uterine luminal epithelial cell surface
have been made, although the nature of the glycocalyx of this cell layer has been implicated in regulation
of implantation in a number of species (Carson et al.
2000).
It seems clear there must be a reason for retention
of the monovular condition and of a means of limiting
implantation of more than one blastocyst when more
than one ovulation occurs. With regard to the latter, a
possible explanation may relate to the way in which the
inverted yolk sac fills the uterine lumen. Multiple yolk
sacs within a single lumen would result in the exclusion
of some of the surface of each yolk sac from the uterine
surface. If so, then dizygotic twinning might be deleterious. In an analogous situation, dizygotic twinning
in the horse frequently results in intrauterine death of
both conceptuses. Ginther (1989) suggested that overlap of the embryos, which in the horse come to occupy
both horns and the body of the uterus, might deprive
the conceptuses of adequate embryonal-maternal exchanges. While overlap of the yolk sacs in the armadillo would not compromise gaseous exchange, it could
deprive the individual conceptuses of appropriate histotrophic nutrition. It should also be pointed out that
other aspects of implantation in the armadillo facilitate
polyembryony. These include both precocious development of the exocelom and its enlargement, and delay
of formation of the primitive streak until the epiblastic

plate has greatly enlarged, providing room for separation of the developing embryonic shields.
Implantation of the blastocyst in the fundus of the
simplex uterus in D. novemcinctus provides rapid association of trophoblast with maternal blood and the
earliest association of endoderm of an inverted yolk
sac with the uterine lumen found in any Eutherian
mammal. The advantages of retaining such an efficient
system are therefore obvious. However, there is considerable variation, both in the structure of the uterus,
which can be simplex or bicornuate, and in the number of young, varying from monotocous to polytocous,
in other armadillo species (Galbreath 1985). The hairy
armadillo, Chaetophractus villosus, has a uterus with
two uterine horns and polyzygotic twinning (Fernandez 1915a); however, the two embryos seem to develop
in the body of the uterus rather than in the horns and
were reported to have inverted yolk sacs. Adamoli et al.
(2001) reported that the definitive form of Cha. villosus
placental villi, and of several other armadillo species,
is hemomonochorial, like that of D. novemcinctus. The
early development of Cha. villosus and other armadillo
species has not been studied, and the way in which the
intervillous space and yolk sac are formed is unknown
in the majority of species.
Conclusions
Until more information is available on implantation and
placentation in species of armadillos with both simplex
and bicornuate uteri, it will be difficult to assess any of
the preceding explanations for causal mechanisms in
the evolution of polyembryony in the genus Dasypus.
It does seem probable, however, that a simplex uterus
with a single blastocyst and the peculiar mechanism of
formation of both yolk sac and chorioallantoic placenta
preceded the development of polyembryony.
18
Sequencing the armadillo genome
Jean L. Chang and James E. Adams
tiva dos genomas. No presente trabalho descreve mos

a produo do sequenciamento de baixa cobertura do
genoma do tatu atravs da abordagem de shotgun e
as novas direes em pesquisas que so possveis com a
disponibilidade deste recurso valioso para o estudo no
s dos Xenarthrans e seres humanos, mas tambm para
a gentica de mamferos e para o estudo de genomas.
Resumen
El genoma del armadillo de nueve bandas, Dasypus
novemcinctus, es el primero en ser secuenciado para
el clado Xenarthra (armadillos, osos hormigueros y
perezosos). En la actualidad se encuentra disponible
pblicamente un low-coverage draft assembly de una
nica hembra de armadillo que cubre ~80% de 2,75 Gb
del genoma eucromtico. Este armadillo es uno de una
serie de 16 mamferos catalogado para secuenciacin
genmica de light-coverage. Su secuencia ayudar a
identificar elementos evolutivamente conservados en el
genoma humano a travs de anlisis genmicos comparativos. En este captulo se describe la generacin de
un low-coverage, whole-genome shotgun armadillo
assembly y nuevos caminos de investigacin para los
estudios genticos y genmicos de xenartros, humanos
y mamferos, que son posibles con la disponibilidad de
este valioso recurso.
Mammalian genome sequencing and human

genome annotation
Study of animal genomes is a technologically intensive enterprise employing much unfamiliar terminology. Because of this, we wish to begin by referring the
reader to appendix 18.1, a glossary of important terms.
We hope inclusion of this glossary will facilitate comprehension of what follows for the nonspecialist.
To date, complete or near-complete genome sequences have been generated from five mammals. The
first mammalian genome project, sequencing the euchromatic portion of the human genome, was launched
in 1990 and completed in 2003 (International Human
Genome Sequencing Consortium [IGHSC] 2001,
2004). Subsequent projects have produced deep-coverage (4- to 8-fold sequence redundancy) whole-genome
assemblies for the mouse (Mouse Genome Sequencing
Consortium [MGSC] 2002), rat (Gibbs et al. 2004),
chimpanzee (Chimpanzee Sequencing and Analysis Consortium [CSAC] 2005), dog (Lindblad-Toh et
al. 2005), and opossum (Broad Institute unpublished
data). The first pairwise mammalian genome analysis
was performed between Homo sapiens and Mus musculus following the initial assembly of the mouse genome. That analysis revealed ~5.2% of the human genome was strongly constrained by selective processes,
Resumo
O genoma do tatu galinha, Dasypus novemcinctus,
o primeiro mamfero da ordem Xenarthra (tatus,
preguias e tamandus) a ser completamente seqenciado. A assemblia deste genoma est disponvel agora
e um esboo baseado em seqenciamento cobrindo
duas vezes o tamanho total do genoma (cobertura baixa
ou light-coverage) de um indivduo do sexo feminino, que inclui cerca de 80% das 2,75 Gb do genoma
eucromtico. O tatu um de uma srie de 16 mamferos que tero seus genomas seqenciados com baixa
cobertura. O seqenciamento deste genoma ajudar
na identificao dos elementos conservados evolutivamente no genoma humano atravs de anlise compara181
182
J. L. Chang and J. E. Adams
a much higher percentage than expected based on the

considerably smaller proportion of the genome (~1.5%)
consisting of protein-coding sequences (MGSC 2002).
Sequence conservation suggests the presence of essential genomic elements. The amount of genome under functional selection was determined by measuring
interspecies conservation rates across regions of sequence and comparing that rate with the rate of neutral
substitution (MGSC 2002). This method allowed quantification of the portion of the genome under selection,
but, because of limited overall divergence time, it could
not pinpoint the specific sequences of functional importance embedded within regions displaying positive
selection. The recent addition of the dog genome in
the comparison allowed further correlation of regions
with high conservation between the three species. For
example, from the intersection of these data we have
identified very strongly conserved regions that contain
genes regulating development and body plan patterning (Lindblad-Toh et al. 2005). In general, functional
elements residing in the nonprotein-coding portion of
the human genome demonstrate higher than expected
conservation between the studied species. However,
comparison of three genomes is still insufficiently sensitive to exactly delineate such elements.
Our ability to annotate the human genome is augmented by comparative genome analysis. Efforts to
identify genes within the human genome using purely
computational, rule-based protocols (ab initio or from
the beginning methods) have been attempted (IHGSC
2001), but these analyses suffer from a lack of sensitivity
(inability to detect known genes) and specificity (inappropriate labeling of nongenes as genes). Currently,
the most effective efforts also utilize homology-based
searches against complementary DNA (cDNA) and
expressed sequence tag (EST) databases and incorporate other external sources of sequence information
(IHGSC 2004). Nonetheless, the gene sets resulting
from these projects are not comprehensive or free of
spurious gene predictions, despite the incorporation
of prior knowledge from the literature and other databases. The use of algorithms that incorporate an
analysis of cross-genome conservation for the identification of protein-coding sequences (e.g., Twinscan,
SGP2, SLAM) provide the most robust procedures to
date for the accurate identification of genes (Parra et al.
2003, Wang et al. 2003, Dewey et al. 2004), and attest
to the potential of comparative genomics for genome
annotation of additional functional elements. In parallel with gene identification, methodologies continue
to be developed for de novo identification of nonprotein-coding RNA sequences (Jones-Rhoades and Bartel
2004; Xie et al. 2005). These protocols develop rules for
annotating subsets of nonprotein-coding RNAs, however, as with gene annotation, our understanding of the
complete set of rules is still very rudimentary. Because
our current understanding of the structure underlying
basic genomic units of information is simply not mature enough for comprehensive ab initio annotation of
genomes, we rely on the observation of evolutionary
conservation as an indicator of the presence of functional elements.
Whereas the identification of transcribed sequences
is difficult, the identification of regulatory elements is
even more challenging, because many transcriptionfactor binding sites are short (612 bp) and variable
in sequence. Current methods to analyze a genome
in isolation are limited in their ability to annotate real
functional elements. Classical methods for annotating
novel sequence, including the identification of putative transcription-factor binding sites, use databases of
known elements from existing research (e.g., TRANSFAC, MAPPER, TRED, Rfam, NONCODE, RNAdb;
see Wingender et al. 2001, Griffiths-Jones et al. 2005,
Liu et al. 2005, Marinescu et al. 2005, Pang et al. 2005,
Zhao et al. 2005). To date, the most effective means for
rapidly recognizing novel functional elements include
cross-species comparisons (Kellis et al. 2003) and analyses of clustered binding sites (Bulyk 2003).
Rationale for low-coverage genome sequencing

To accurately identify conserved sequence elements
likely to be functional and important in most or all
mammalian genomes, a proposal to broadly sample
genomes across the widest diversity of eutherian mammals was put forth to the National Human Genome
Research Institute (NHGRI, see figure 18.1). Specifically, the utility of low-redundancy sequencing for
comparative genome analysis was assessed (Margulies
et al. 2005). The species chosen (table 18.1) are similar
enough at the nucleotide level to permit accurate sequence alignment to the assembled human genome, but
have diverged sufficiently so that functional sequences
under selection should stand out compared with neutrally evolving flanking regions. Based on the model of
Eddy (2005), a cumulative branch length of four substitutions per site is necessary for detection of imperfectly
conserved six-base-pair elements (e.g., a transcription
factor binding site), with a false positive rate of 1 error
183
Figure 18.1. Phylogeny of placental

mammals selected for low-coverage
sequencing. Species slated for lowcoverage (2x) sequencing, including
armadillo (bold), are shown relative
to completed (underlined) and in
progress (italic) deep-coverage genome species. Supraordinal clades
are designated by labeled black bars
on far right.
per 10 kb of sequence. To obtain the necessary aggregate divergence time, species with the longest branch
length within eutheria were preferentially selected for
sequencing.
The primary intent of generating low-coverage genome sequences, such as for the armadillo, is to compare a broad and representative range of mammalian
genomes with the human genome for discovery of elements that are evolutionarily conserved across eutherian mammals. Sequencing technology and capacity
have advanced considerably in recent years, such that
reduced costs and accelerated speed make it possible to
consider sampling genomes beyond the major model
organisms (NHGRI 2002). Nevertheless, the cost of
Table 18.1. Mammalian Species Selected for Low-Redundancy Sequencing

Common name
Scientific name
African savanna elephant

Chinese pangolin
Domestic cat
Eurasian common shrew
European hedgehog
European rabbit
Ground squirrel
Guinea pig
Large flying fox
Lesser hedgehog tenrec
Little brown bat
Nine-banded armadillo
Northern tree shrew
Rock hyrax
Small-eared bushbaby
Two-toed sloth
Loxodonta africana
Manis pentadactyla
Felis catus
Sorex araneus
Erinaceus europaeus
Oryctolagus cuniculus
Spermophilis tridecemlineatus
Cavia porcellus
Pteropus vampyrus
Echinops telfairi
Myotis lucifugus
Tupaia belangeri
Procavia capensis
Otolemur garnettii
Choloepus hoffmani
Distance to humana
0.323
0.305
0.292
0.414
0.438
0.310
0.300
0.423
0.294
0.484
0.290
0.307
0.301
0.396
0.278
0.324
Notes: At this writing an additional 8 mammal species were being considered for low coverage sequencing
by the National Human Genome Research Institute (NHGRI).
a. Substitutions per site.
184
genome sequencing is still considerable. Thus, a major

design consideration of the mammalian low-redundancy sequencing project was to achieve comprehensive coverage of each additional species at the lowest
possible cost, in order to maximize the number of different mammalian genomes that could be sequenced.
Theoretical calculations using the Lander-Waterman
equation (Lander and Waterman 1988) predicted that
63%, 87%, and 95% of the genome would be included at
1, 2, and 3 coverage, respectively. We note that even
though genome coverage beyond the threefold mark
yields progressively less new sequence information, increased levels of redundancy are crucial for very high
sequencing accuracy and the ability to perform independent genome assemblies. To strike the best balance
between coverage of a particular species and coverage
across many species, a sequencing level of twofold redundancy was chosen for the project.
The nine-banded armadillo as a xenarthran

genome project
The biomedical importance of Dasypus novemcinctus
as a model for leprosy and the accessibility of a working research colony (Truman this volume) greatly influenced its selection as the representative xenarthran. D.
novemcinctus is the best studied of the Dasypodidae, as
many chapters in this book attest, and as is evidenced
by the listing of more than 200 publications concerning this species in the PubMed database. It is also the
most common armadillo and the most widely distributed of all xenarthran species, ranging from Peru and
Argentina to the southern United States (Taulman and
Robbins 1996).
Armadillo sequence data in GenBank prior to the
low-coverage sequencing effort consisted mainly of
three types: high-coverage BAC sequence produced by
the ENCODE consortium (ENCODE Project Consortium 2004), BAC end sequences (BES), and PCR-derived sequencing of specific nuclear and mitochondrial
gene fragments for molecular phylogenetic studies
(Hss et al. 1996; Delsuc et al. 2001; Delsuc et al. 2003).
Assessment of genes derived by PCR is biased in favor
of acquiring data from the more conserved regions of
the genome. PCR-based cloning of target genes is possible only if there is sufficient conservation in the regions surrounding the gene for primers to successfully
anneal and generate the desired amplification product.
Because genes having high rates of divergence are less
likely to be accessible by PCR-based methods of clon-
ing, estimates of genetic distances between species using PCR-based data have a greater chance of underestimating differences between the species studied. The
comprehensive and nontargeted nature of whole-genome shotgun sequencing is a better means for estimating the rates of divergence at the nucleotide level by
providing data without need for prior knowledge of the
regions sequence content.
ENCODE armadillo BACs were selected for orthology to the 44 regions targeted by ENCODE that comprise ~1% (30 Mb) of the human genome. Targets for
~15 Mb of the ENCODE project were manually identified based on gene content or preexisting comparative
genomic data; the remaining 15 Mb were divided into
thirty 500 kb regions using a stratified random-sampling strategy, varying both gene density and nonexonic mouse-human conservation (NHGRI 2003). The
quality of ENCODE armadillo BAC sequence is very
high: BACS were sequenced to 8 coverage with subsequent manual analysis to determine order and orientation of all contigs (ENCODE Project Consortium
2004). The high depth of coverage ensures that the vast
majority of bases were sequenced multiple times, such
that a very accurate call of the actual base composition
could be performed. The consensus derived from the
redundancy of deep-coverage reads allowed the depth
of data to compensate for local poor-quality sequence
in any single read, and ensured uniform, high-quality
sequence.
The available D. novemcinctus BAC end sequence
reads were part of a multimammal project to sample
eutherian BAC libraries. Sequence from 694 clones
from the VMRC-5 D. novemcinctus BAC library (see
appendix 18.2) were included. Generation of pairs of
BAC end sequence from these 140 kb average insert
clones yielded 1,288 D. novemcinctus nucleotide entries in GenBank (accessions: CL003069CL003763,
CZ268190CZ268782). These entries were submitted
to the genomic survey sequence (GSS) division of GenBank by McGill University and the Genome Quebec
Innovation Centre (MUGQ).
Whole-genome sequencing and assembly

Estimating heterozygosity
The low-coverage armadillo project also provides the
groundwork for future deep-coverage sequencing,
through the careful selection of a DNA donor suitable
for a full-depth genome project. One of the critical fac-
tors in whole-genome assembly is the level of heterozygosity of the individual sequenced. Unlike clone-based
sequencing, where only a single haplotype is represented in any assembly of a given clone, the whole-genome shotgun approach generates sequence from both
haplotypes of a diploid donor. Thus, assembly algorithms dealing with sequence derived from partially
inbred or outbred individuals must be able to tolerate
and combine data representing two haplotypes for any
given segment of the genome. Aligning reads from
two different haplotypes of the same genetic locus results in conflicting sequence data: both haplotypes are
valid representations of the region and the nucleotides
differing between the two haplotypes will tend to be
mismatched bases of high sequence quality. Because
genome duplications are also present in mammalian
genomes, it becomes increasingly difficult to distinguish between alleles (two different copies of the same
locus) and duplications (two different loci with similar
sequences) as the level of heterozygosity rises. Hence,
the more polymorphisms in the DNA donor, the more
fragmented the resulting assembly is likely to be.
To identify the best DNA donor with regard to heterozygosity, we collected and screened a set of candidate armadillos to determine their heterozygosity levels, with the goal that the individual would continue to
be accessible for a higher coverage genome project at
a later date. In mammals, the Y chromosome contains
large duplications that are more suitable for BAC-based
sequencing and would largely confound any whole-genome assembly approach. Hence, eight female armadillos were collected, as we preferred to sequence from a
female individual and generate full 2 coverage of the
X chromosome. These animals were provided by Richard Truman of the National Hansens Disease Programs
(NHDP) at Louisiana State University School of Veterinary Medicine and included six females from the Port
Hudson region of Baton Rouge, Louisiana, and two
from northern Arkansas. One advantage of sampling
North American animals was that reports in the literature indicate that D. novemcinctus in North America
are significantly less heterogeneous than their South
American counterparts (Huchon et al. 1999).
An initial small-insert library was created to generate
~3,000 random, whole-genome shotgun reads. These
reads were screened for known mammalian repetitive
elements using RepeatMasker (A. F. A. Smit, R. Hubley,
and P. Green unpublished data). PCR primer pairs were
designed to amplify ~300 random genomic loci with
the goal of selecting ~200 pairs for the final assay set.
185
All PCR assays were characterized on three armadillo

DNAs from the test panel to select the final collection
of assays. Primer pairs that were chosen yielded robust
and reproducible PCR reactions across all three test
DNAs, and had sharp, single bands of the expected size
following agarose gel electrophoresis of the amplified
products.
PCR products were obtained for the eight candidate donor DNAs and sequenced. Because the pool
of sequencing templates contained both alleles for
each individual, the resulting reads were analyzed using Polyphred, an algorithm for automated tagging of
polymorphic bases (Nickerson et al. 1997). Heterozygous bases undetected by the automated protocol were
identified by manual scoring. Two rounds of manual
scoring were undertaken to minimize human error.
The manual review allowed bases that were misidentified by the automated algorithm to be corrected and
relevant bases in regions of mixed base quality were assigned base calls. Polymorphic insertions and deletions
were rare and were not included in this analysis. More
than 80,000 bases were assessed for each candidate armadillo, with an average heterozygosity rate across the
candidate panel of one heterozygous base per 1,750 bp.
Although the sample size was insufficient to fully represent the genome, it allowed us to identify one armadillo
that stood out in the dataset; the lowest heterozygosity rate determined from the panel was 1 in 2,600 bp
(range: 1/1,3001/2,600). This Port Hudson, Louisiana,
armadillo (NHDP accession: 3136) was chosen as the
DNA donor for the genome sequencing project.
Creating genomic libraries for whole-genome
sequencing
DNA from the chosen female donor (Broad Institute
Genome Sequencing Sample Repository ID: 240.0) was
purified from blood cells and physically sheared to the
desired sizes of 4 kb for plasmids and 40 kb for fosmids
via hydroshear to avoid any potential biases resulting
from restriction enzyme digestion. The fragments were
subjected to two rounds of size fractionation to ensure a
tight insert size range, which is essential for subsequent
assessment of assembly quality. The size-selected DNA
molecules were end-repaired, ligated into the appropriate vectors, then transformed into electrocompetent E.
coli host cells. Transformants were picked and arrayed
into liquid medium in 384-well microtiter plates. For
the fosmid library, glycerol stocks of each clone were
archived at -80C. Because no attempt was made to preselect the DNA content of the subclones, the genomic
186
Figure 18.2. Sequencing and scaffold assembly strategy. Genomic DNA is sheared and cloned to produce 4-kb (small circles) and 40-kb (large circles) insert sequencing libraries that are end-sequenced
to generate paired-end reads, represented by bold dashes connected with dotted or dashed lines.
A contiguous consensus sequence, or contig (bold line), is derived from overlapping reads (dashes).
Scaffolds are formed using information from paired-end reads that span contig gaps (arcs) to order and
orient contigs.
libraries collectively represented all clonable regions

of the genome, with variance in depth of coverage due
largely to stochastic effects.
Estimating euchromatic genome size
In order to estimate the size of the euchromatic portion of the armadillo genome, we sequenced the ends
of more than 25,000 fosmid clones and determined
their orthologous alignment positions on the human
genome assembly. Only 45.2% of the fosmid end sequences could be orthologously mapped because of sequence divergence and independent retroposing DNA
element activity. From this set of mapped reads, we

identified a subset of 3,887 fosmid clones where both
ends could be assigned to unique positions in the human genome. We calculated the human syntenic length
encompassed by armadillo fosmids to generate an estimate of armadillo genome expansion or contraction.
We observed that the human genome has expanded
(syntenic length: 41,340 6,288 bp) relative to the syntenic armadillo segments. If the syntenic clone lengths
observed reflect the overall genomic rate of expansion
or contraction, then we expect the euchromatic portion
of the armadillo genome to be 96.6% that of the human
genome, or 2.75 Gb. From this we calculated that a 2

whole-genome shotgun project would require the generation of 5.5 Gb of sequence data.
Genome sequencing and assembly
Two types of libraries were generated for whole-genome shotgun sequencing: 4 kb DNA inserts in plasmid
vectors and 40 kb DNA inserts in fosmid vectors. We
implemented a two-tiered sequencing strategy, where
the pairs of terminal sequences would be obtained
from subclones of each vector type. Based on our 2.75
Gb estimate of the armadillo euchromatic genome, we
sought to generate paired end sequence from 7.5 million plasmids and 94,000 fosmids. Thus, the average locus in the armadillo genome should have been covered
by sequence arising from approximately two independent sequencing reactions (i.e., 2 sequence coverage).
In addition, each locus should have been contained in
approximately five fosmid subclones (i.e., 5 fosmid
physical coverage).
The strategy of employing paired-end reads allowed
the sequence to be linked in the assembly process.
First, we identified all reads displaying overlapping sequences and created assembled sequence contigs. Second, we built scaffolds by placing the individual contigs into their correct order and orientation, using the
linkage information from the paired ends (figure 18.2).
The scaffolds thus consist of the ordered and oriented
sequence contigs interleaved with gaps, where the gap
sizes are inferred from the size of the spanning subclones.
The scaffolds of low-coverage mammalian genomes
can be further improved by utilizing phylogenetically
close reference genomeseither a finished sequence
or a high-quality draft assembly. For the armadillo, we
referenced human genome build 35 (hg17) and canine
genome version 2 (canFam2).
Although the gross genome organization of placental mammals varies widely in terms of chromosome
number and chromosome size, genome organization
at the local level is often preserved across species. The
mapping of a region from one mammalian genome to
its orthologous region in another forms the basis for
maps of conserved synteny (MGSC 2002, Murphy et
al. 2004, Lindblad-Toh et al. 2005). Human-mouse
syntenic segments have a mean length of 6.9 Mb, with
more than half the segments being longer than 16.1 Mb
(MGSC 2002). This suggests that mammalian syntenic
segments are large enough to support an assisted assembly procedure.
187
Assisted assembly is a multistep process that limits

the use of information from the reference genome to
corroboration of data in the de novo assembly. First,
whole-genome shotgun reads were aligned to the reference genome: for armadillo we used blastz to align
the armadillo reads to hg17 and canFam2. At this stage
we tagged the set of reads that aligned uniquely to the
reference genome. We also formulated a list of genomic
staples, pairs of reads that aligned logically to the reference genome (i.e., each end read aligned uniquely
and both reads from the same insert mapped to the reference genome in a manner consistent with the original
subclone).
Second, short read-read overlaps were confirmed if
the reads involved independently aligned to the reference genome and produced the expected overlap. In
areas where read overlap for contig formation in the
initial assembly did not meet the minimum overlap
threshold, the confirmed overlaps permitted incorporation of sequence into the assisted assembly to extend
contigs (figure 18.3A), form new short contigs, or fill
gaps in the initial assembly, resulting in a substantial
read usage increase and additional sequence contiguity.
Third, information implying a linkage between two
scaffolds in the initial assembly was confirmed if the
two scaffolds were anchored by staples that had unique
and logical alignment to the reference genome. Potential connections could be inferred from single links but
were considered insufficient for scaffold formation or
extension. However, single links that had been confirmed on the reference genome could be safely used to
join scaffolds (figure 18.3B), improving the contiguity of
the resulting assembly.
Finally, staples within scaffolds were used with the
reference genome to highlight potential misassemblies.
Anchored scaffolds were scrutinized for inconsistent
mapping on the reference genome. Further analysis allowed identification of weak joins in the de novo scaffold (figure 18.3C) or sites of syntenic breakage. This
resulted in better accuracy and understanding of the
final assembled genome.
dasNov1: The first armadillo genome assembly

Assembly statistics
The first D. novemcinctus genome assembly resulted
in 2.15 Gb of sequence. The level of sequence coverage
specified for a genome project determines the amount
188
Figure 18.3. Enhancement of de novo assembly by alignment of whole-genome shotgun reads to a

reference genome. A. A short read-read overlap (solid circle) is confirmed by the reference genome
(thick line), resulting in extension (dotted line) of the de novo assembly contig (black line). B. Validation of single linkage (gray arc) by the order of staples (vertical bars) in scaffolds on the reference
genome allows two scaffolds to be merged. C. Illogical pattern of staple (vertical bar) placement on
the reference genome may indicate regions of misassembly (dotted circle) that may be corrected on
further inspection.
of quality sequence produced based on the expected genome size. Twofold coverage of the 2.75 Gb armadillo
euchromatic genome implies the production of 2 2.75
Gb, or 5.5 Gb, of sequence. To attain 2-fold coverage of
the armadillo genome in good quality bases, 8,442,780
whole-genome shotgun reads were generated by the
Broad Institute and submitted to the National Center
for Biotechnology Information (NCBI) trace archive:
7,500,461 plasmid reads and 942,319 fosmid reads. Af-
ter filtering for sequence quality, vector contamination,

and other minor quality issues, 8,098,969 reads were
used as input for the assembly. Version 1.0 of the D.
novemcinctus genome assembly (dasNov1) consists of
848,432 contigs (Genbank accession: AAGV01000001AAGV01848433). Contigs were ordered and oriented
using paired-end, double-linkage information to form
the 304,391 scaffolds in the dasNov1.0 assembly (Genbank accession: CH482385CH666571).
The sequence contiguity of low-redundancy genome

assemblies is inherently low because of the limited sequence and linkage information available at reduced
coverage levels. Sequence contiguity is assessed by the
contig N50 length metric: the contig length in which
the typical nucleotide resides (appendix 18.1). The average contig N50 of all 2 redundancy genomes assembled to date is 2.9 kb (Sante Gnerre personal communication). For dasNov1.0, the contig N50 is 2.7 kb. This
is much shorter than the values for the mouse (mm3)
assembly (25 kb with 6.5 coverage, MGSC 2002) or
the dog (canFam2: 180 kb at 7.5 depth, Lindblad-Toh
et al. 2005). In contrast to the intrinsically short contig
length of reduced coverage sequencing, the difference
between the contig N50 of the mouse and dog assemblies is thought to be due to: (1) improvements in assembly algorithms, and (2) inherent differences in the
repetitive content of the two genomes, including both
retrotransposable elements and segmental duplications
(Lindblad-Toh et al. 2005).
Similarly, scaffold length, that is, the cumulative
length of linked contigs plus the length of the gaps as
inferred from paired-end data, is limited with lower
coverage. For the armadillo assembly process, two
scaffold N50s are pertinent: 15.5 kb for the de novo
armadillo assembly and 45.7 kb for the human and
dogassisted assembly of dasNov1. These statistics can
be compared with the scaffold N50s of deep-coverage
genome assemblies, which are 17 Mb and 45 Mb for
mouse and dog respectively (MGSC 2002; LindbladToh et al. 2005).
Assembly coverage
The portion of the armadillo genome represented by
dasNov1 was estimated by comparing the assembled
genome size, 2.14 Gb, with the estimated genome size,
2.75 Gb, or 77.8% genome coverage. A sequence-based
approach to assessing genome coverage measured the
quantity of dasNov1 contigs that uniquely aligned to
independently generated, finished sequence: 40 kb
fosmids from the genome project and ENCODE BACs
from the VMRC-5 library. Coverage of 9 fosmids and
15 BACs by dasNov1 sequence ranged from 45% to 73%,
reflecting the variability of local coverage in a lightly
sequenced genome assembly. Structural accuracy of the
order and orientation of contigs within scaffolds was
generally congruent with the sequence from finished
clones, although patterns of interlaced scaffolds were
often seen. At higher sequence coverage levels, these
physically interlaced scaffolds would likely merge into
189
longer scaffolds with more contiguous sequence as a result of additional linkage information. On the nucleotide level, differences between the assembly sequence
and fosmid sequences were expected because of heterozygosity within the sequenced DNA donor. Additional differences, including insertions and deletions,
would be expected in any comparison of the dasNov1.0
assembly with sequence generated from the VMRC-5
BAC library because the latter sequence was derived
from a different animal; specifically, a male armadillo
captured, sampled, and released in the Lake Iamonia
area of northern Florida.
An alternative assessment of assembly coverage
quantifies the expected gene content in the assembly.
To determine how well the low-coverage armadillo
genome assembly was able to capture known proteincoding genes, we compared the Consensus Coding Sequence (CCDS) collection, a set of core human proteincoding regions with high-quality annotation, with the
dasNov1 scaffolds. Using NCBI blastn (match penalty
= -1, word size = 7, expect cutoff = 1 10-5), 4.1% of
the 14,795 CCDS genes were not found in the dasNov1
assembly. If we define a gene as being found by the
assembly when at least 10% of its bases are captured in
dasNov1, then 95.3% of CCDS genes have been found.
In terms of near-complete coverage, 37.9% of CCDS
genes have >90% coverage in the armadillo assembly,
with 13.9% completely captured (figure 18.4).
Because not all reads generated to fulfill the 2 coverage goal were incorporated into the assembly, sequence
coverage at 2 depth results in single-coverage sequence
for a significant portion of the assembly. In fact, 20.5%
of the reads generated were not incorporated into the
assembly, resulting in only 1.16 coverage in Q20 bases
(defined in appendix 18.1). Thirty percent, or 644 Mb,
of the assembly was single-coverage sequence. Unlike
the uniform sequence quality in deep-coverage assemblies such as the human and mouse genome projects,
the regions of the armadillo assembly with single-pass
sequence had wider variation in base-to-base sequence
quality than the 70% assembled with redundant coverage.
The quality of individual bases is highly dependent
on sequence coverage, because redundant sequencing
ensures the correct assignment of each base. The minimum base quality standard for a finished genome sequence is a quality score of 40, often denoted as Q40;
80.5% of bases in dasNov1 achieved a base quality score
of 40. For comparison, 98.3% of the 7.5 dog assembly
are Q40 bases (Lindblad-Toh et al. 2005). At the other
190
Figure 18.4. Human genes found by nucleotide BLAST to armadillo scaffolds. Coverage of human genes
from the CCDS gene set is assessed by a nucleotide search in armadillo scaffolds using discontiguous
MegaBLAST. The number of genes (left y-axis) detected at a given percent gene coverage (x-axis) is represented by vertical bars. The cumulative percent of genes (right y-axis) found at the selected gene coverage
level or better (x-axis) is indicated by a dashed line.
end of the quality spectrum, 4.5% of bases in the armadillo assembly had quality of less than Q20, compared
with only 0.4% for the dog assembly. An important
high-quality subset within the <Q20 class of bases were
heterozygous loci in the donor individual, where allelic
differences resulted in one of two different high-quality
bases appearing in the data with equal frequency. The
assembler was forced to choose one of these two alleles
at random and assign an artificial base quality of zero.
5 physical coverage of the genome. As such, the fosmid library, in conjunction with the genome assembly, represents a powerful tool for isolating regions of
interest in the armadillo genome. If a fosmid-read pair
flanks a region of interest, a physical clone of the region
can be acquired by simply identifying the fosmid clone
corresponding to the read pair and ordering the clone
from the fosmid repository (BACPAC Resource Center,
Oakland, Calif.).
Benefits of a fosmid sequencing library

Fosmid-derived data provide powerful tools for assessing the quality and reliability of genome assemblies.
The DNA packaging system of the phage imposes a
strict size restriction on the packaged fosmid inserts.
This focused insert size distribution makes anomalous
distances between paired fosmid-insert reads a strong
indicator of potential assembly problems, for example,
if the predicted size based on the assembly deviates
significantly from the expected insert size. Each pair of
fosmid reads represents the insert boundaries of a 40
kb clone spanning the region between the sequenced
reads. Although sequence from the armadillo fosmid inserts made up <10% of the total dataset, endsequenced fosmids were utilized so that, on average,
each location in the genome could be found in approximately five fosmid clones. This is referred to as
Using the Dasypus novemcinctus genome

sequence
Whole-genome sequence accelerates the pace of molecular research by providing the nucleotide sequence for
genes of interest from the sequenced species through
in silico genome gazing. The fragmented nature of a
low-coverage genome presents particular challenges for
finding genes of interest. The following section outlines
a method for finding genes of interest in the current
dasNov1 assembly. Although details of genome assembly files and Web-based tools may change (see current URLs in appendix 18.2), the underlying reasoning
should remain applicable for finding genes of interest in
any low-coverage genome assembly.
There are two ways to identify D. novemcinctus
genes using orthologs available from other mammals

(human, mouse, dog, etc.). The first approach is a
faster but potentially less sensitive method, using the
nucleotide sequence of the known orthologous gene to
search for nucleotide similarity in the: (1) trace archive,
(2) whole-genome shotgun (wgs) database of genomic
contigs, or (3) database of genomic scaffolds, accessible
through the NCBI Genomic Biology portal. Because of
the evolutionary distance between available gene sequences, use of cross-species megaBLAST (Basic Local
Alignment Search Tool) is advised, whether searching
dasNov1 contigs and scaffolds or the collection of raw
sequencing reads for armadillo in tracedb. The discontiguous megaBLAST alignment algorithm is optimized
for sequence alignment of diverged sequences (Altschul
et al. 1990; Ma et al. 2002).
Screening against the scaffolds or contigs of the genomic assembly is a more streamlined approach, because alignment of the trace files has already been performed and the fidelity of the sequences is higher than
that of raw sequence found in the trace archive. However, a search of the trace archive should be performed
if matches to contigs are not found, because ~20% of
sequenced reads are not currently incorporated into
the assembly. For higher sensitivity (but longer compute time), word size should be set as low as possible
and percent identity requirement set to none, or the
smallest mismatch penalty possible should be selected.
If trace files are identified, the matching sequence
can be entered into a text file individually in FASTA
format. FASTA files can be copied and pasted into
CAP3 (contig alignment program) to obtain more
comprehensive alignments of the trace files (Huang and
Madan 1999). To verify the contig generated by CAP3,
the contig should be submitted to BLAST again using
translated query vs. protein database (blastx). The
blastx algorithm is used because of the degeneracy in
the genetic code. The query is submitted and translated
in six frames (three forward and three reverse) against
the nonredundant (nr) database, resulting in amino
acid alignments.
A second approach to identifying orthologous genic
sequence searches the contigs or scaffolds using the
protein sequence derived from the gene of interest.
Nucleotide databases are queried with an amino acid
sequence by using the protein query vs. translated
database (tblastn) method to identify traces, genomic
contigs, or scaffolds containing the gene of interest.
Once again, the sequence obtained can be verified as
with aligned trace files. One caveat of utilizing the trace
191
archives or contigs is that they may have sequencing

errors that result in mis-sense mutations, early stop
codons, additions, or deletions that result in frame
shifts. Protein-based searches are potentially more sensitive because synonymous (wobble) base changes
are penalized in a nucleotide search but do not affect
a tblastn search. Furthermore, tblastn searches utilize a
word size of three amino acids, which is approximately
equivalent to a nucleotide discontiguous word size of 6
or 7, a more sensitive setting than the minimum word
size of 11 for discontiguous megaBLAST.
Once a contig or scaffold has been identified, the
structure of the assembly around the gene of interest
can be understood by studying the AGP file, which lists
all scaffolds and information about their component
contigs and gaps, including contig and gap length, order, and orientation. Two types of assembly gaps are
possible: (1) fragment gaps, where the gap is spanned
by a sequenced clone and the size of the gap has been
estimated on the basis of paired-end linkage information, or (2) clone gaps, which indicate breaks between
scaffolds due to a lack of linkage information.
If portions of a gene or region of interest appear as
a fragment gap in the assembly, then generating the
missing sequence that lies within the gap may be fairly
straightforward. The assembly.reads file from dasNov1,
available from the Broad Institute FTP site, can be
searched by contig name to find pairs of fosmid reads
(nomenclature: G734P8XXXX series of reads) that
flank the gap in a logical manner (i.e., unique placement and correct orientation). Direct sequencing of the
fosmid insert should recover the missing region. Alternatively, if the gap size is small, it may be possible to
recover the missing region by using sequence flanking
the gap to design PCR primers to amplify the sequence
in the gap using genomic DNA as input template.
Gap-flanking sequence can be entered into web-based
primer generation programs such as Primer3 (Rozen
and Skaletsky 2000) to design primers for PCR amplification. Sequencing of the amplified product may recover the missing fragment and close the gap between
contigs. For a desired sequence that falls into a clone
gap, the investigator can search the assembly.reads file
for fosmids that point in the direction of the gap and
sequence the fosmid in order to walk into the gap.
With additional sequence that extends into the gap, the
assembly and trace archive can be searched to identify
additional clones that might walk further into the gap,
obviating the need for traditional lab-based genomic library screening procedures.
192
New avenues of research

With the availability of multiple mammals for crossspecies comparisons, the immediate goal is alignment
of the armadillo assembly to the finished human sequence for genome annotation of conserved sequence
elements. Once aligned, the most expedient way to annotate the armadillo genome will be to transfer human
gene models to the armadillo sequence where the data
are consistent with the existence of a gene in the armadillo assembly. Although an exhaustive gene catalog of
the armadillo is not possible at 2 coverage, the majority of genes are at least partially represented.
Research in the Dasypus system will clearly benefit
from the wealth of genic sequence available in dasNov1.
Orthologous genes from complete or near-complete
genomes (human, mouse, rat, dog) are useful but are
often inadequate substitutes for the actual Dasypus
protein. For example, only D. novemcinctus regularly
develops lepromatous-type leprosy when inoculated
with Mycobacterium leprae (Truman this volume);
however, suitable reagents and tools for immunological studies are not commercially available. Although
there are commercially available human, murine, and
porcine reagents, few of them cross-react well in the
Dasypus system. This gap between available mammalian reagents and the Dasypus system seems to stem
from local sequence differences found in Dasypus. For
example, tumor necrosis factor alpha (TNF-a) is well
conserved across mammalian species. Degenerate
primers designed to amplify the conserved regions of
TNF-a have been used to find the sequence in many
mammals for cladistic studies. Although these primer
sets worked on other xenarthrans (silky anteater [Cyclopes didactylus], three-toed sloth [Bradypus tridactylus], and naked-tail armadillo [Cabassous unicinctus]),
they failed to produce amplified product in Dasypus (de
Jong et al. 2003). Thus, it appears that any efforts to advance Dasypus as a model system, whether for leprosy
infection or other reasons, will require development of
species-specific immunological reagents.
Both complete and partial genic sequences can be
used to generate amino acid sequence that, validated by
cDNA information, can be subcloned into expression
vectors to generate recombinant proteins. The dasNov1
contigs were queried for 33 target genes chosen on the
basis of: (1) cytokines (Th1 and Th2) important in immune responses to M. leprae, (2) representation of immune cell markers (CD markers), (3) potential markers
for monocyte recruitment (CCL2 and CCR2), and (4)
association with susceptibility in human leprosy (TLR2, PARK2, and PACRG). Contigs from dasNov1 either
partially or completely capture the coding sequence
(cds) of each target gene (table 18.2). Several of these
genes have been annotated and submitted to NCBI, and
a biologically active, recombinant D. novemcinctus IL-2
has been produced from E. coli (J. Adams unpublished
data).
Recombinant proteins can be used as bioactive molecules or to produce antibodies. Translated nucleotide
sequences yielding putative amino acid sequences that
align well with other mammalian orthologs can be used
to identify epitopes. These in turn can be used to generate polyclonal antibodies for use in enzyme-linked
immunosorbant assays (ELISA) and fluorescence-activated cell sorting experiments (FACS). Stretches of
aligned amino acid sequences can be selected to generate epitopes by using in silico predictions of hydrophobicity (Kyte-Doolittle Hydropathy PlotMembrane Plot), antigenicity (Hopp-Woods Hydrophilicity
PlotAntigenic Plot), and exposure (Surface Exposure
Plot). Classically, a gene would be isolated by screening a genomic library for clones containing the gene
of interest, often using a radioactively labeled orthologous gene sequence as the probe. The identified clones
would then be sequenced and subcloned to expression
vectors, followed by purification of a recombinant protein product and injection for production of antibodies.
Now, however, protein coding sequence can be isolated
in silico, allowing development of D. novemcinctus reagents to proceed more quickly.
The catalog of genic sequences derived from dasNov1 could also populate a first-generation armadillo
gene chip if expression profiling experiments are desirable. Alternatively, primers can be generated to produce
reagents (primers and probes) for reverse transcribed
polymerase chain reaction (rtPCR) experiments to
quantify gene expression in the Dasypus system.
Having a genome assembly in hand also allows for
rapid development of microsatellite markers for studying the genetic structure of populations and subpopulations or assessing gene flow between populations.
Microsatellites can be studied using minimal amounts
of DNA when assayed by PCR. Studies like those elegantly described by Prodhl et al. in this book have
been limited by a small marker set, because of the laborious nature of marker development. Traditionally,
microsatellite markers are developed by cloning DNA
fragments that have been selected for simple sequence
tandem repeat content (SSTRs). These <1 kb fragments
Table 18.2. Coverage of Selected Armadillo Genes Found in

DasNov1 Contigs
Genes
NCBI accession number
Cds (%)a
1. Cytokines and chemokines

CCR2
IFN-g
DQ094083
IL-2
DQ092925
IL-4
DQ096648
IL-6
IL-8
DQ099902
IL-10
IL-12 p35
IL-12 p40
IL-27
MCP-1 (CCL2)
TNF-a
DQ096649
> 95
100
100
> 42
> 75
100
> 75
> 90
> 85
> 85
> 95
100
2. Cell markers
CD3 (T-cell)
CD8 (T-cell)
CD14 (Macrophage)
CD19 (B-cell)
CD20 (B-cell)
CD44 (Monocyte)
CD68 (Macrophage/monocyte)
FOXP3 (T-cell)
> 90
> 50
> 90
< 50
> 75
> 85
> 90
> 50
3. Associated with susceptibility

FOXP3
Ninjurin1
PACRG
PARK2
TLR-1
TLR-2
TLR-3
TLR-4
TLR-5
TLR-6
DQ113655
TLR-7
TLR-8
TLR-9
TLR-10
TNF (a promoter)
> 50
> 50
> 92
> 23
> 85
> 74
> 75
> 79
> 80
100
> 26
> 74
> 64
> 96
> 50
a. Percentage of coding sequence covered by dasNov1 contigs.
are then sequenced and analyzed for SSTRs of useful

repeat length with suitable unique flanking sequence
for primer design. In silico microsatellite marker development obviates the need for lab work prior to primer
design. The genome can be searched rapidly for appropriate SSTRs and, if desirable, even be assessed for
equal distribution across the genome as inferred from
armadillo/human genomic alignments. The N50 contig
length of 3 kb suggests that most contigs are suitable
for microsatellite design. The newly designed markers
could subsequently be tested in a population sample to
193
determine their prevalence and, thus, their utility for

intra- and interpopulation applications.
Genome sequence of greater than 1 depth also
suggests the possibility of developing SNP markers.
Redundant sequencing can be mined for high quality mismatched bases in overlapping reads to create a
database of candidate SNPs for the genome. SNPs are
a powerful tool for mapping genetic loci because they
are abundant across the genome; they are perhaps most
familiar because of their use in mapping disease genes.
SNPs can be unique to subpopulations, reflecting common ancestry within that population. Thus, the genetic
cause of heritable phenotypes can be identified if sufficient SNPs are queried and cross-referenced in properly
selected affected and control groups. An armadillo genome opens the possibility for developing these genetic
tools to localize the genetic components of quantitative
traits in the armadillo biomedical model.
Conclusions
With the generation of a Dasypus novemcinctus genome
assembly, immediate access to the majority of armadillo
genes and other sequences of interest is now available.
This resource will enable accelerated development of
tools and new avenues of research for the understanding of xenarthran biology. Efforts that would improve
the genome assembly include (1) deep coverage sequencing of D. novemcinctus for a highly contiguous
assembly amenable to detailed annotation efforts, (2)
anchoring of deep coverage scaffolds to chromosomes
by fluorescent in situ hybridization (FISH) or genetic
mapping techniques, (3) development and sequencing
of full-length cDNA libraries to complement annotation efforts, and (4) comparative genomic analysis with
the forthcoming low coverage sloth (Choloepus hoffmanni) genome to lend insight into the evolution of
the Xenarthra clade. The rapid pace of genomic studies
should lead to the achievement of these goals soon. For
example, sequencing of the sloth genome is already underway at the Genome Sequencing Center of Washington University School of Medicine, and deep coverage
sequencing of the armadillo genome will be performed
at the Baylor Human Genome Sequencing Center with
expected completion in 2007.
Acknowledgments
Special thanks are due to L. Gaffney for help with the
manuscript, figures, and tables. We are grateful to S.
194
Gnerre, J. Chou, and M. Lara for scientific input and

helpful advice. In addition, we thank K. Lindblad-Toh,
P. Samollow, and K. Dewar for critical feedback about
the manuscript. We acknowledge all the members of
the Broad sequencing platform and assembly team for
their dedication and hard work.
Appendix 18.1. Glossary of genome sequence

analysis terms
Assembly: The process of aligning a large number of
sequencing reads, detecting valid sequence overlap,
and merging the data to create a consensus sequence
that is a representation of the original source DNA.
Assisted assembly: An assembly process that utilizes
synteny with a reference genome to improve the
contiguity and quality of a de novo assembly.
BAC clone: Bacterial artificial chromosome vector
bearing a large genomic DNA fragment, typically
100200 kb.
Base quality (Q20, Q40, etc.): A logarithmic scale
metric denoting the probable accuracy of a base call
in the raw read. A score of 10 denoted as Q10 = 10%
probability of containing a sequencing error, Q20 =
1% error, and Q40 = 0.01% error.
Contig: The consensus sequence derived from overlapping whole-genome shotgun reads.
De novo assembly: An assembly process that uses only
whole-genome shotgun data from the species under
assembly.
Depth (or coverage): The average number of highquality, independently generated representations at
each genomic base from a collection of whole-genome shotgun reads.
Euchromatic genome: The portion of the genome that
can be cloned and sequenced with current technology; believed to contain all genes.
Genome annotation: Delineation of the location and
function of genomic features (e.g., genes, promoters, regulatory elements, retrotransposons, segmental duplications).
Haplotype: Haploid genotype; a set of SNP markers
statistically linked to a single chromatid.
Heterozygosity level: The quantity of bases within a
single genome bearing different alleles for each copy
within a diploid chromosome; does not include insertions and deletions.
Microsatellite: Highly polymorphic, locus-specific, codominant genetic loci characterized by short repeat
motifs of 26 bp, also known as variable number
tandem repeat markers (VNTR); commonly used as

markers in population genetics studies.
N50 length: The standard metric for describing assembly contiguity. N50 equals the length, L, where half
of all nucleotides in the assembly occur in contigs or
scaffolds of at least length L.
Paired-end sequence: Single read sequence from each
end of a genomic fragment of known size and subcloned into any sequencing vector (e.g., plasmid,
fosmid, BAC).
Physical coverage: The number of genome equivalents
represented by the cumulative length of all clone inserts from a genomic library.
Scaffold: The ordered and oriented arrangement of
contigs using paired-end linkage information.
Single nucleotide polymorphism (SNP): Single-nucleotide, fixed differences found between individuals:
frequently common genetic variants found within
the population.
Synteny: The preserved order of features along the genomic axis between related organisms.
Whole-genome shotgun (WGS): A sequencing strategy wherein genomic DNA is broken into random
small fragments that are subcloned and sequenced.
Appendix 18.2. Internet resources for armadillo

genome analysis
blastx, blastz, tblastx
www.ncbi.nlm.nih.gov/BLAST/
Broad Institute FTP site
www.broad.mit.edu/ftp/pub/assemblies/mammals/armadillo/
Two files: Assembly.agp and assembly.reads.
CAP3
For program documentation: genome.cs.mtu.edu/cap/
cap3.html.
To download program: seq.cs.iastate.edu/
Consensus coding sequence collection
www.ncbi.nih.gov/CCDS/
Discontiguous megaBLAST
http://www.ncbi.nlm.nih.gov/blast/Blast.cgi?PAGE=
Nucleotides&PROGRAM=blastn&MEGABLAST
=on&BLAST_PROGRAMS=megaBlast&PAGE_
TYPE=BlastSearch&SHOW_DEFAULTS=on
Database: Other [Whole-genome shotgun reads (WGS)]
Organism: nine-banded armadillo (taxid:9361)

Entrez Query: Dasypus novemcinctus [ORGN]
Program Selection: More dissimilar sequences (discontinuous megablast)
Discontiguous megaBLAST trace archive
http://www.ncbi.nlm.nih.gov/blast/Blast.cgi?PAGE
= Nu c l e o t i d e s & P R O G R A M = b l a s t n & B L A S T
_SPEC=TraceArchive&BL AST_PRO GRAMS=
megaBlast&PAGE_TYPE=BlastSearch
Use database Dasypus novemcinctusWGS.
NCBI genomic biology
www.ncbi.nlm.nih.gov/Genomes/
195
NCBI trace archive

www.ncbi.nlm.nih.gov/Traces/trace.cgi?
Primer3
frodo.wi.mit.edu/cgi-bin/primer3/primer3_www.cgi
RepeatMasker
www.repeatmasker.org/cgi-bin/WEBRepeatMasker
VMRC-5 BAC library
bacpac.chori.org/armadillo5.htm
19
Chromosomal studies in the Xenarthra
Wilham Jorge and Helio Rubens Jacintho Pereira Jr.
process of endogamy in small populations, resulting in

reduced fertility. In addition, cytogenetic studies allow
determination of possible evolutionary relationships
among different groups and the separation of species
within a genus. In 1969, Benirschke called attention
to the importance of cytogenetics for animals kept in
zoos. For example, individuals of the same or related
species might have different chromosomal numbers. If
these animals were kept in the same enclosure and allowed to reproduce, this could jeopardize the offspring.
Managing reproduction of wild animals requires correct systematic identification of species such that genetic patrimony is preserved and reproduction is not
jeopardized by hybridization or other factors that can
reduce population viability. Traditional and molecular
cytogenetics aid these efforts by allowing the global genetic characterization of a species.
Resumen
Se realiza una sntesis de los trabajos de citogentica
del Orden Xenarthra desde 1925 a 2004, la que muestra
la importancia del anlisis cromosmico en la identificacin sistemtica de las especies teniendo en cuenta
el manejo reproductivo y la preservacin del patrimonio gentico. El numero diploide, la determinacin
cromosmica del sexo y la evolucin del cariotipo son
analizadas en las especies de perezosos (familias Brady
podidae y Megalonychidae) de armadillos (familia
Dasypodidae) y osos hormigueros (familias Cyclopedidae y Myrmecophagidae).
Resumo
Uma reviso de 1925 a 2004 sobre a citogentica na Ordem Xenarthra realizada, mostrando a importncia
da anlise cromossmica na identificao sistemtica
das espcies, tendo em vista o manejo reprodutivo e
a preservao do patrimnio gentico. O nmero
diplide, a determinao cromossmica do sexo e a
evoluo do caritipo so analisados nas espcies de
preguias (famlias Bradypodidae e Megalonychidae),
de tatus (famlia Dasypodidae) e tamandus (famlias
Cyclopedidae e Myrmecophagidae).
A history of xenarthran cytogenetics

The first studies on chromosomes of the Xenarthra were
done with nine-banded armadillos (Tatusia [now Dasypus] novemcinctus) by Newman and Patterson (1910)
and Newman (1912). These authors found 31 chromosomes in the spermatogonia and 32 in the oogonia.
They attributed the odd number in the male to the existence of an XO/XX mechanism. Painter (1925) found 60
chromosomes in somatic cells of the same species. This
discrepancy was not resolved until better techniques
became available some 40 years later. In 1962, Beath
et al. analyzed peripheral blood of D. novemcinctus to
show that this species had 2n = 64 chromosomes. Saez
et al. (1964) described Dasypus hybridus as also having 2n = 64 chromosomes and with 32 bivalents in the
meiotic cells. The latter study further reported differ-
Introduction
Cytogenetic studies in wild animals, especially those in
danger of extinction, are valuable tools for ecologists
and conservationists, who use them for sex determination in animals lacking sexual dimorphism and for
surveying genetic problems such as deletions, translocations, and inversions that might be attributed to the
196
ences in the karyotype involving some big metacentric

chromosomal pairs and small acrocentric ones between
two subspecies of D. hybridus.
The year 1969 was a hallmark due to the publication
of the book Comparative Mammalian Cytogenetics by
Kurt Benirschke, in which five species of armadillos
were analyzed (Benirschke et al. 1969): D. hybridus and
D. novemcinctus (both 2n = 64), Euphractus sexcinctus
(2n = 58), Chaetophractus villosus (2n = 60), and Cabassous centralis (2n = 62). Also in the same year, the remarkable contribution of Corin-Frederic documented
the atypical mechanism of gender determination in the
sloth Choloepus hoffmanni. Ten animals (6 females and
4 males) from the Lincoln Park Zoo had been captured
at different times in Panama, Costa Rica and Ecuador.
Both males and females had 2n = 49 chromosomes. The
females had an XO sex chromosome constitution in the
somatic cells and males had the material from the Y
translocated to an autosome. Meiotic study of males
showed normal behavior of the sex chromosomes, suggesting the mechanism of sex determination was normal despite the translocation. The author proposed that
females represented the extreme expression of compensation for an inactive X chromosome.
Several years later, Meritt and Benirschke (1973) described the karyotypes of Tolypeutes matacus and Zaedyus pichiy. These authors identified the three-banded
armadillo, T. matacus, as the xenarthran with the lowest number of chromosomes (2n = 38), and drew attention to the remarkable morphological and karyotypical
differences within the order. In 1977, with the advent
of banding techniques, Jorge et al. analyzed the C- and
G-banding patterns of E. sexcinctus, Cha. villosus, D.
novemcinctus, D. hybridus, and Z. pichiy. Even though
the diploid numbers of these species are similar, comparison of banding patterns indicated karyotype evolution in some groups of autosomes and the X chromosome could be explained by Robertsonian fusion and
pericentric inversion. The same work also described a
species of C. hoffmanni with 2n = 50 chromosomes in
both a male and female. The female presented two X
chromosomes with different C-band patterns and the
male possessed a supposed Y of atypical size, around
50% bigger than the first pair of autosomes. This Y chromosome also presented an atypical C- and G-banding
pattern.
In 1980, Sonta et al. described the karyotype of Choloepus didactylus as 2n = 53 chromosomes for males
and females. Females presented a single X chromosome
197
and in males the Y chromosome was translocated to

autosome 21.
In 1982, Jorge described the karyotype of Chlamyphorus truncatus, a species restricted to the region of
Mendoza, Argentina, as 2n = 58 chromosomes.
Jorge et al. (1985a) analyzed 26 individuals of the
families Bradypodidae (sloths) and Dasypodidae (armadillos). From these, 7 were analyzed with C- and
G-banding. The authors demonstrated principally that:
(1) Sloths of the genus Bradypus have a Robertsonian
mechanism of karyotype evolution. The existence of an
additional chromosome in some males suggested two
mechanisms of sex determination: one with XY (XX/
XY) and the other with XYY (XX/XYY) males; (2)
Sloths of the genus Choloepus presented a karyotype
polymorphism with diploid numbers varying from 49
to 53 and even 64 chromosomes; females were either
XO or XX, and males had the Y chromosome translocated to an autosome; and (3) In all armadillos the sex
chromosome mechanism was XX/XY. D. novemcinctus
and D. hybridus demonstrated a constant 2n number
of 64, but with differences in some chromosome pairs.
The diploid number in Chl. truncatus (2n = 58) was
quite similar to that of Euphractini species.
Barroso and Seunez (1991) analyzed the karyotypes
of species in the armadillo genera Dasypus, Euphractus, and Cabassous, confirming the data of Benirschke
et al. (1969) and Jorge et al. (1985a). In Dasypus they
confirmed the diploid number of 64 (D. novemcinctus
and D. septemcinctus), although one specimen of D.
novemcinctus with 2n = 65 (due to a supernumerary
chromosome) was observed. The results showed some
structural differences in karyotypes between D. septemcinctus, D. novemcinctus, and D. hybridus. E. sexcinctus
and Ca. unicinctus (tatouay?) had diploid numbers of
58 and 50 respectively, confirming the findings of Benirschke et al. (1969) and Jorge et al. (1977).
Jorge et al. (1985b) and Jorge (2000) analyzed the
karyotype of the silky anteater, Cyclopes didactylus, a
threatened species of the Amazonian region. This species has 2n = 64 chromosomes and is cytogenetically
distant from anteaters in the family Myrmecophagidae,
underlining the remarkable differences between them.
Lastly, and most recently, Dobigny et al. (2005) used
the newly developed technique of chromosome painting to analyze patterns of karyotype evolution within
the three major lineages of xenarthrans. Their results
indicated a relatively low rate of genomic repatterning
and may provide clues to the architecture of the ances-
198
W. Jorge and H. R. J. Pereira Jr.
Table 19.1. Characterization of Species in the Genus Bradypus According to Diploid Number (2n), Fundamental Number
(FN), and Sex Chromosomes (Female/Male)
Species
B. pygmaeus
B. torquatus
B. tridactylus
B. variegatus
Origin
2n
FN
Sex
References
Boca del Toro, Panama

Reserva Poo das Antas, Rio de Janeiro
Island of Curari, Manaus, Amazonia
Belem, Para
No data
50
52
54
No data
64
56
56
No data
XX/XY
XX/XY
XX/XY, XX/XYY
5,6
3
1,2
1,2,4
Sources: (1) Jorge 1982, (2) Jorge et al. 1985a, (3) Jorge and Pinder 1990, (4) Goldschmidt and Almeida 1993, (5) Anderson and
Handley 2001, (6) Anderson and Handley 2002.
tral xenarthran karyotype (see also Delsuc and Douzery this volume).
The evolution of the karyotype in the genus

Bradypus
Figure 19.1 shows the karyotypes for three of the four
extant species in this genus. Further details are provided in table 19.1. The karyotypes of two species, B.
tridactylus and B. variegatus, may have resulted from
a mechanism of fusion or fission. In general, fusion
consists of the union of two telo/acrocentric chromosomes, making a meta/submetacentric pair, while, alternatively, fission creates a dissociation of two meta/
submetacentric chromosomes into two telo/acrocentric chromosomes. It is often difficult to decide which
process has occurred. Fusion could have generated the
union of pairs 23 and 24 of B. variegatus, resulting in
pair 25 of B. tridactylus (figure 19.1), or fission of pair
25 of B. tridactylus could have led to the origin of pairs
23 and 24 in B. variegatus. According to Holmquist and
Dancis (1980), which process actually occurs depends
on the presence of latent telomeres and on the sequence
of DNA in the centromeres.
Jorge and Pinder (1990) described the karyotype of
the maned sloth B. torquatus. It has the lowest chromosomal number (2n = 50), and is substantially different from the other species. B. torquatus is restricted to
the coastline forest of Sergipe, Espirito Santo, and Rio
de Janeiro states (Brazil) and is in danger of extinction
(Fonseca et al. 1994). Its phenotype is so different that
Wetzel (1985a) put the maned sloth into another genus
(Scaeopus torquatus), although this has not been widely
accepted. Beyond the reduction of the diploid number,
structural modifications in chromosome pairs 1, 3, 6,
13, and 20 have occurred. The existence of a low number of chromosomes with a high number of structural
rearrangements suggests a larger phylogenetic distance
between this species and the others in the genus, possibly justifying a systematic revision of the group.
Goldschmidt et al. (1995) dyed the nuclear organizing regions (NOR) in B. variegatus and marked pair
25 (metacentric) for comparison with B. torquatus. In
B. variegatus there was an intercalary localization of
NORs on pair 25, while in B. torquatus they were telomeric, suggesting a fusion mechanism of chromosomal
evolution.
Bradypus pygmaeus was considered a new species
by Anderson and Handley (2001, 2002). This species is
found on the Isla del Escudo in Panama. The insularity
of living on an island seems to have promoted a substantial reduction in body size, as has been documented
for other island species. Characteristics of hair indicate
similarity between animals of the island and the mainland; however, there are no cytogenetic or DNA studies
yet to determine the degree of similarity.
Chromosomal determination of sex in Bradypus
Table 19.1 shows that sex determination in Bradypus
is of the XX/XY kind, except for B. variegatus. In this
species, Jorge et al. (1985a), Goldschmidt (1992), and
Goldschmidt and Almeida (1993) found an additional
chromosome (2n = 55) in some males, interpreted as a
second Y or a supernumerary chromosome (Jorge et al.
1985a; see also figure 19.1C).
The data for B. variegatus just described came from
wild populations in Manaus and Rio de Janeiro. In a
recent study, Gonalves (2003) analyzed a population
of B. variegatus located at a public square in the city
of Teofilo Otoni, Minas Gerais state, Brazil. According to senior residents, sloths had been living in the
area since the founding of the municipality in 1878;
however, with increased urbanization they became restricted to the square. Consequently, the population became highly inbred as evidenced by the birth of albino
animals, young with other sorts of anomalies, and the
premature deaths of some juveniles. In a chromosomal
analysis, two kinds of sex determination were found in
this population: females were XX, but males were either XY or XYY. It remains unknown how these two
A.
B.
C.
D.
Figure 19.1. Karyotypes of the three-toed sloths (Bradypus). A. B. variegatus. Pairs 23 and 24 may
result from fission of pair 25 of B. tridactylus. B. B. tridactylus. The karyotype is very similar to B.
variegatus except for the last four pairs. Pair 25 may have resulted from fusion of pairs 23 and 24
of B. variegatus. C. B. variegatus. In this specimen, 2n = 55; the additional chromosome in some
males is interpreted as a second Y (XYY). D. B. torquatus. The karyotype is substantially different
from the others in the genus. Reduction of diploid number and structural modifications in pairs
1, 3, 6, 13, and 20 are observed.
200
Table 19.2. Characterization of Species in the Genus Choloepus According to Diploid Number (2n) and Sex Chromosomes
Species
Origin
2n
Females
Choloepus
BrazilAmazonia
51, 52
X monosomy
(no species specified)
and Goiania
53, 64
C. didactylus
Japan Zoo
53
X monosomy

BrazilAmazonia
53
-

Brazil-Belem
53
-

BrazilManaus
52, 64
-
49
X monosomy
C. hoffmanni
Panama, Costa Rica, Ecuador

Colombia, Panama
50
Two X chromosomes
Males
No homologous partners
(B-chromosomes)
Translocation of Y
chromosome to autosome 23
-
-
-
Translocation of Y
chromosome to autosome 21
Atypical Y chromosome
Reference
4
3
5
4
4
1
2
Sources: (1) Corin-Frederic 1969, (2) Jorge et al. 1977, (3) Sonta et al. 1980, (4) Jorge et al. 1985a, (5) Jorge unpublished data.
systems have been maintained in three very distant and

distinct regions (Manaus, Rio de Janeiro, and Teofilo
Otoni). Sloths occur in other public squares in Brazil
(e.g., Consentino 2004), but no chromosomal analyses
are available yet.
Chromosomal polymorphisms in the genus

Choloepus
The phenotypes of two- and three-toed sloths might
suggest the separation of the two genera in a not too
distant past. However, the work of Patterson and Pascual (1968) indicates a long separation between Bradypus and Choloepus ancestors, a finding reinforced by
subsequent studies (Sarich 1985; Webb 1985a; Greenwood et al. 2001; Delsuc and Douzery this volume).
Although there are no studies of genomic homology
as yet, the two Choloepus species (C. hoffmanni and C.
didactylus) also differ dramatically from one another
in, among other things, geographic distribution (Wetzel 1985a), hair color (Meritt 1985a), size, weight, age
at sexual maturity, pregnancy duration, and inter-birth
intervals (Taube et al. 2001).
Two-toed sloths are characterized by a polymorphism in chromosome number, with 2n values ranging
from 49 to 64 (table 19.2). Additional chromosomes
with no homologous partners are interpreted as B
chromosomes; there is also female X monosomy and
Y-autosome translocation (Jorge et al. 1985a). The latter two mechanisms lead to an odd diploid number in
both sexes (table 19.2, figure 19.2). Recently, Kovacs et
al. (2001) and Dobigny et al. (2005) found a 2n = 65
karyotype in C. didactylus. The specimen of Dobigny et
al. exhibited a big Y chromosome, probably a Y-autosome translocation, and a single unpaired acrocentric,
leading to the odd diploid number.
Taking other findings about each species in turn,

Sonta et al. (1980) reported C. didactylus females with
X monosomy, while in males the Y chromosome was
translocated to the small metacentric autosome number 21. The size and form of the translocated Y was very
similar to that described by Corin-Frederic (1969) in
C. hoffmanni, except that here autosome 23 was translocated to the Y chromosome. Jorge et al. (1977) described, with G- and C-banding techniques, individuals of C. hoffmanni with 2n = 50 chromosomes in both
sexes. In contrast to other studies, the female in this
sample had two X chromosomes with different patterns
of C-banding. The male had one atypical Y chromosome, with an exaggerated large size (50% bigger than
the first pair of autosomes), and with a region of heterochromatin occupying more than half of the long arm.
There was no evidence that the Y chromosome was the
result of a translocation with an autosome. Finally, using human chromosomes as probes, Svartman et al.
(2005) found that C. hoffmanni has a karyotype similar
to the proposed ancestral Eutherian karyotype of 44,
46, or 48 chromosomes.
Female monosomy (i.e., XO) may represent an extreme form of dosage compensation, with the total
elimination of the inactivated X chromosome. Hayman and Martin (1974) reported this phenomenon in
females from four genera of the family Peramelidae
(Marsupialia). These authors and Hayman (1990) proposed that this could be related to the mechanism of
X inactivation in mammals. According to Cooper et
al. (1993), X elimination has a paternal origin. For example, Johnston et al. (2002) analyzed the inactivation
and reactivation of the X chromosome in Isoodon obesulus (Marsupialia: Peramelidae) and concluded that
it is the paternal X that is inactivated and eliminated
in this species, probably at the beginning of meiosis in
A.
B.
201
C.
Figure 19.2. Karyotypes of two-toed sloths (Choloepus). A. Female with X monosomy. B. Male with two supernumerary chromosomes. C. C-banding pattern. Notice the big supernumerary (or B chromosome) with typical banding pattern.
females. Whether something similar has occurred in

Choloepus remains unknown.
Cytogenetics of the Dasypodidae

All living armadillos are members of the family Dasypodidae (Wetzel 1985b). Four subfamilies have been
proposed based on ecological and molecular data (Delsuc and Douzery this volume; but see Gardner 2005);
we treat each in turn below.
Dasypodinae
This subfamily contains just the genus Dasypus, which
in turn contains the largest diversity of species (7)
among living xenarthrans. Of these, three have been
karyotyped; all exhibit 2n = 64 chromosomes (table
19.3). However, the apparent constancy of the diploid
number hides considerable diversity in chromosome
morphology (Beath et al. 1962; Saez et al. 1964; Benirschke et al. 1969; Barroso and Seunez 1991). There
are 18 pairs of big and medium acrocentric chromosomes, where no differences are observed among species. However, differences occur in one of the two sets
of big submetacentric pairs: D. hybridus has the second
most metacentric pair in relation to the other species.
Differences also occur in numbers of pairs. D. novemcinctus presents 6 pairs of submetacentric chromo-
somes (Beath et al. 1962; Benirschke et al. 1969), while

D. septemcinctus has 5 submetacentric pairs (Barrosso
and Seuanez 1991) and D. hybridus exhibits 5 metacentric pairs (Saez et al. 1964). Additionally, D. novemcinctus has 5 pairs of small acrocentric chromosomes; D.
septemcinctus and D. hybridus have 6. Finally, the X
chromosome is the most metacentric one in D. septemcinctus and D. novemcinctus and the most submetacentric one in D. hybridus.
Euphractinae
The subfamily Euphractinae contains the following
genera: Euphractus, Chaetophractus, and Zaedyus (see
Gardner 2005 for a different view), all of which present low variation in diploid sets (5862 chromosomes,
table 19.3). Despite the similarity in diploid numbers,
chromosomal evolution in this group cannot be explained exclusively by the Robertsonian mechanism of
translocation. This is at least partly because of evidence
such as the noncorrespondence of a huge number of
acrocentric chromosomes in Cha. villosus and Z. pichiy
with the number of metacentric and submetacentric
chromosomes in E. sexcinctus, as well as differences in
the fundamental numbers of each species (table 19.3).
Pericentric inversions might be another mechanism for
chromosomal reorganization in the evolution of karyotypes in this group.
202
Table 19.3. Chromosomal Features of Armadillo Species

Subfamily
Species
Chlamyphorinae

Dasypodinae

Euphractinae

Tolypeutinae

Chl. truncatus
Chl. retusus
D. hybridus
D. kappleri
D. novemcinctus
D. pilosus
D. sabanicola
D. septemcinctus
D. yepesi
Cha. nationi
Cha. vellorosus
Cha. villosus
E. sexcinctus
Z. pichiy
Ca. centralis
Ca. chacoensis
Ca. tatouay
Ca. unicinctus
Ca. sp.
P. maximus
T. matacus
T. tricinctus
2n
58
-
64
-
64/65
-
-
63/64
-
-
-
60
58
62
62
-
50
-
46
50
38
-
-
-
SM
-
SM
-
-
SM
-
-
-
A
SM/A
A
SM
-
M
-
M
SM
M
-
-
-
A
-
A
-
-
A
-
-
-
A
SM/A
A
M
-
A
-
A
M
A
-
Banding
C and G
-
G
-
G, NOR and B chromosome
-
-
G, NOR and chromosome fusion
-
-
-
C and G
C, G and NOR
G
-
-
-
-
-
-
-
-
References
7
No data
1
No data
2,4,6
No data
No data
8
No data
No data
No data
6
6,8
5,6
4
No data
8
No data
9
3
6
No data
Sources: (1) Saez et al. 1964, (2) Beath et al. 1965, (3) Benirschke and Wurster 1969, (4) Benirschke et al. 1969, (5) Meritt and Benirschke
1973, (6) Jorge et al. 1977, (7) Jorge 1982, (8) Barroso and Seunez 1991, (9) Pereira et al. in preparation.
Note: A = acrocentric chromosome, M = metacentric chromosome, NOR = nucleolar organizing region, SM = submetacentric chromosome, SM/A = submetacentric/acrocentric chromosome.
Tolypeutinae
This subfamily has seven species, with major differences among the genera. For example, chromosome
number exhibits the largest variation of any subfamily,
from Tolypeutes matacus (2n = 38) to Cabassous centralis (2n = 62, table 19.3). Within Cabassous, two karyotypes have been published (table 19.3). However, two
other specimens (both male) have not been classified
yet, and are currently designated as Cabassous sp. (table
19.3). The karyotype (2n = 46) of one of these animals is
depicted in figure 19.3.
Chlamyphorinae
This subfamily contains two species in the genus Chlamyphorus (note: Gardner 2005 eliminates this subfamily, placing both speciesin separate generawithin
the Euphractinae). The karyotype of Chl. truncatus has
been described (table 19.3), but there are few reports
about the biology or cytogenetics of the other species,
Chl. retusus.
Cytogenetics of the Vermilingua

Living anteaters comprise two families. The Myrmecophagidae contains two genera: Myrmecophaga,
with only one species (the giant anteater), and Tamandua, with two species of vested anteater. A second family, Cyclopedidae, contains one species of silky anteater
(Cyclopes didactylus). Karyotypes have been published
for all species (table 19.4) and indicate minimal variation across species. The molecular phylogenies proposed by Delsuc et al. (2001, 2002) and Barros et al.
(2003) indicate that Cyclopes originated earlier than Tamandua and Myrmecophaga. Therefore, its karyotype
could be considered ancestral. One notable trend in the
Vermilingua is reduction of the diploid set through fusion events of acrocentric chromosomes.
Cytogenetic data have been critical in describing the
systematics of anteaters. Prior to 1975, two species of
vested anteater were recognized: T. tetradactyla and T.
longicaudata. Cytogenetic analyses of these two spe-
203
Figure 19.3. Karyotype of Cabassous sp. The X

chromosome is a medium metacentric, and
the Y chromosome is a small acrocentric.
individual described by Hsu and Benirschke (1969) was

actually T. mexicana.
In 2004 Pereira et al. identified a new karyotype in
the genus Tamandua with 2n = 56 chromosomes (figure 19.4). We have examined a large number of these
animals genetically and morphologically. As described
below, the data suggest a new species of anteater may
have been found (referred to here as Tamandua sp.).
The chromosomal set of Tamandua does not present acrocentric chromosomes (except for the Y), thus
standardization of the karyotype follows that of Pereira
cies showed identical diploid numbers (2n = 54, Hsu

1965, Jorge et al. 1977). These results supported data on
geographical distribution and morphology obtained by
Wetzel (1975) that ultimately led to these species being combined as T. tetradactyla. In 1969, Hsu and Benirschke described the karyotype of a male T. tetradactyla from Chiapas, Mexico with 2n = 54 chromosomes.
However, in 1975 Wetzel divided T. tetradactyla into
two species: T. tetradactyla from northern to southern
South America, and T. mexicana from central Mexico
to the northern portion of South America. Thus, the
Table 19.4. Karyotypic Data for Anteater Species

Species
Collection sites
Sex
2n
References
T. tetradactyla

T. mexicana
Tamandua sp.

C. didactylus

M. tridactyla
Philadelphia Zoo, U.S.A.

Colombia
Chiapas, Mexico
Fundao Zoobotnica, Minas
Gerais, Brazil (found in Atlantic
rainforest, Sao Paulo State, Brazil)
Japura River, Brazil
Manaus, Brazil
Herman Park Zoo, U.S.A.
M
F
F
M
54
54
54
56
Third pair
SM
SM
Third pair M
Small SM
No data
Small SM
A
2
3
2
6
F
M & F
F
64
64
60
No data
Third pair SM
No data
No data
Smallest chromosome
No data
4
5
1
Sources: (1) Hsu 1965, (2) Hsu and Benirschke 1969, (3) Jorge et al. 1977, (4) Jorge et al. 1985b, (5) Jorge 2000, (6) Pereira et al. 2004.
Note: See Table 19.3 for definitions of abbreviations.
204
et al. (2004). Comparison of karyotypes in this genus

shows that in group I, T. tetradactyla presents 4 pairs
of big metacentric chromosomes, T. mexicana 5 pairs,
while Tamandua sp. presents only 3 (Pereira et al.
2004). Group II consists of 3 pairs of big submetacentric chromosomes in all species. Group III comprises
7 pairs of medium submetacentric chromosomes in T.
tetradactyla, 11 pairs in T. mexicana, and 9 submetacentric ones in Tamandua sp. In Group IV, 7 pairs of
metacentric chromosomes of medium size are found
in T. mexicana, while T. tetradactyla and Tamandua
sp. each possess 12 pairs. The X chromosome is a large
metacentric, equivalent to the third pair of autosomes
in all species. The Y chromosome is a small acrocentric. These karyotypic differences could be explained
by chromosomal fusion or other processes of structural
rearrangement. In Tamandua sp., the majority of the
autosomes are small metacentrics, possibly originating
through fusion of acrocentric chromosomes found in
the other members of the Myrmecophagidae.
In our opinion, the data described above merit revision of the genus Tamandua into three species rather
than the two proposed by Wetzel (1975). Whether this
taxonomic addition is ultimately approved or not, our
work demonstrates that the members of this family require more studies of their taxonomy and cytogenetics.
Figure 19.4. Karyotype of Tamandua sp. The X chromosome

is a large metacentric, and the Y chromosome is a small
acrocentric.
20
Genes and demes
Population genetic analyses of the Xenarthra
Paulo A. Prodhl, W. J. Loughry, and Colleen M. McDonough
a primeira viso detalhada da estrutura gentica de

populaes selvagens dentro desse grupo. Muitas dessas anlises genticas continuam a focalizar nas relaes filogenticas dentro do grupo Xenarthra, ou entre representantes dos Xenarthra e outros grupos de
mamferos. At ento, anlises genticas de populaes
estavam limitadas a investigaes (a) do impacto da poliembrionia e da determinao do sucesso reprodutivo
em populaes do tatu galinha, Dasypus novemcinctus,
e (b) padres de variao gentica ambos entre populaes de D. novemcinctus, ou entre D. novemcinctus e
outras espcies de tatu em estudos filogeogrficos. No
entanto, com relao a perspectiva de conservao no
futuro prximo, existe a necessidade de investigaes
envolvendo um maior nmero de amostras de outras
populaes, e tambm de outras espcies.
Resumen
Muchos aspectos de la biologa de los xenartros son
difciles de investigar porque los mtodos tradicionales
de estudio son de uso limitado. Para muchos animales
los anlisis genticos pueden proveer una excelente alternativa para tratar numerosas cuestiones. Este tipo de
estudios en xenartros recin est comenzando: hace slo
10 aos que se han desarrollado marcadores moleculares adecuados para proveer una detallada observacin
de la estructura gentica de poblaciones silvestres. Numerosos anlisis genticos continan enfocados hacia
las relaciones filogenticas dentro de los xenartros o
entre los xenartros y otros grupos de mamferos. Los
anlisis de gentica de poblaciones han sido limitados
al estudio de: (a) impacto de la poliembrona y determinacin del xito reproductivo dentro de poblaciones
y (b) patrones de variacin gentica entre poblaciones,
ambos estudios aplicados al armadillo de nueve bandas
(Dasypus novemcinctus); en otras especies, han sido utilizados como componente de estudios filogeogrficos.
Sin embargo, los aspectos que nos interesan de la conservacin hacen necesarias, en un futuro muy cercano,
ms muestras tanto de poblaciones como de especies.
Beginning some 20 years ago, the development of modern molecular genetic techniques (e.g., DNA fingerprinting, Jeffreys et al. 1985; Burke and Bruford 1987)
for the study of wild animal populations triggered an
explosion of research that has resulted in the emergence
of the new field of molecular ecology (see Avise 2004
for a recent review). These studies have been remarkable in providing new insights into animal behavior
and ecology. Perhaps the most obvious example of this
comes from studies of birds where genetic analyses of
offspring showed that many species traditionally considered monogamous were, in fact, not (e.g., Quinn et
al. 1987). Discoveries such as these suggest previously
unsuspected complexities in animal populations and
have led to entirely new and novel research questions.
What is perhaps most striking about some of the insights derived from molecular genetic analyses of wild
populations is how completely undetectable they were
Resumo
Muitos aspectos da biologia dos xenartras so difceis
de estudar porque os mtodos de observaes tradicionais so de uso limitado. Para muitos animais, as
anlises genticas provaram ser uma forte alternativa
para responder muitas questes biolgicas. Esses tipos de estudos, no grupo Xenarthra, esto apenas no
comeo: faz somente 10 anos desde que o desenvolvimento de marcadores genticos adequados forneceu
205
206
P. A. Prodhl, W. J. Loughry, and C. M. McDonough
by conventional observational techniques. For example,

in the case of monogamous birds, in many instances the
species had been studied extensively over long periods
of time, and yet these observational studies failed to
reveal fundamentally critical aspects of the birds biology.
Reliance on observational methods can be even
more problematic. If molecular genetic analyses indicate we are missing things even in species that are relatively easy to observe, what then about species where
observation is more difficult? As many of the other
chapters in this book attest, this is clearly the case for
most xenarthrans, where individuals are often highly
cryptic. In cases where detailed and numerous observations of known individuals are difficult or impossible to
obtain, it seems that molecular genetic analyses may be
our primary source of insight.
In studying wild animal populations, it is now standard practice to collect some form of DNA sample (e.g.,
blood, hair, or tissue) as a component of ones sampling
effort. These genetic samples are then analyzed in a variety of ways. Within a single population, genetic data
are often used to estimate genetic relationships between
individuals and determine parentage of juveniles. Genetic analyses between populations of the same species are typically designed to map patterns of genetic
variation onto known spatial and historical features
of that taxon (i.e., a phylogeographic analysis, Avise
2000). Lastly, genetic data can be used to examine evolutionary relationships among taxa. This latter subject
has been a major focus of many recent genetic analyses of the Xenarthra and is covered thoroughly in this
book by Delsuc and Douzery (see also Garcia 2003). In
this chapter we will focus on the application of genetic
markers to analyses either within or between populations of a single species.
Within-population analyses
Polyembryony
As in so many instances in the Xenarthra, only the ninebanded armadillo, Dasypus novemcinctus, has been extensively studied to date. However, even this statement
is somewhat misleading because, at this time, virtually
all our knowledge of within-population genetic structure comes from a single population in north Florida
(see McDonough and Loughry this volume for more
details about this population). While this population
has been intensively sampled, it remains to be seen how
representative the data from this site are relative to populations living in other parts of the species range. This
is a potentially serious concern because nine-banded
armadillos were first reported at this Florida site in 1972
(Stevenson and Crawford 1974). Thus, descriptions of
population structure there might represent recent adaptation to abnormal conditions. Comparisons with
populations in more ancient parts of the species range
(e.g., South America) indicate that genetic, behavioral,
and morphological differences can exist (Wetzel and
Mondolfi 1979; Wetzel 1985b; Loughry, Prodhl, McDonough, and Avise 1998; Loughry and McDonough
1998a; Huchon et al. 1999; McDonough et al. 2000),
although it is not clear to what extent (if any) these
differences impact the basic genetic structure of armadillo populations. In any case, one important task for
the future will be to broaden our sampling to include
not only descriptions of population genetic structure in
those species that have not yet been sampled at all, but
also to examine additional populations in those species that have already been studied, in order to confirm
previously described patterns.
Genetic analyses of the nine-banded armadillo
population in north Florida were designed specifically
to examine the impact of polyembryony. As described
elsewhere in this volume (Enders, Peppler), armadillos
in the genus Dasypus are the only vertebrates known to
exhibit obligate polyembryony (Newman and Patterson
1910; Galbreath 1985; Craig et al. 1997), whereby multiple offspring are produced from a single fertilized egg.
In the case of D. novemcinctus, polyembryony results
in females giving birth to litters of quadruplets each
spring.
Embryological studies in the early 1900s were the
first to suggest that nine-banded armadillos were polyembryonic and that littermates were thus, presumably, genetically identical (Newman and Patterson
1910; Hamlett 1932). However, phenotypic differences
often occur among littermates (Storrs and Williams
1968; Loughry and McDonough 2002), suggesting that
embryological data alone were insufficient to conclusively demonstrate that siblings were indeed genetically
identical. Genetic confirmation of polyembryony was
clearly required.
Surprisingly, this was not done for many years. This
was probably because early genetic analyses of armadillos using protein electrophoresis showed that populations in the United States were virtually monomorphic
(Ramsey and Grigsby 1985; Moncrief 1988), possibly
as a result of some form of population bottleneck due
207
Figure 20.1. Representative micro

satellite gel showing complete
genetic uniformity among ninebanded armadillo littermates.
Litters 14 contain lanes for all
four siblings of each litter. Note
that there is no genetic variation
within litters, but that juveniles
from one litter differ genetically
from those in other litters. Similarly, no two randomly selected
adult members of the population
were genetically identical to one
another. Reprinted from Prodhl
et al. (1996).
to their recent colonization of the region (Humphrey

1974; Taulman and Robbins 1996). In any case, lack of
variation made confirmation of polyembryony impossible.
The development of microsatellite DNA markers
provided access to new levels of genetic variation. Although less variable than populations in more ancient
parts of the species range (see below), U.S. populations
of nine-banded armadillos were sufficiently variable to
allow confirmation of polyembryony (Prodhl et al.
1996). Using seven microsatellite loci, each with 37 alleles, Prodhl et al. (1996) showed that presumed littermates were in fact genetically identical to one another
and were distinct from individuals from other litters
and the population at large (figure 20.1). While the genetic evidence clearly demonstrated the occurrence of
polyembryony in D. novemcinctus, similar analyses are
still required for the other six species in the genus to
confirm that they are also polyembryonic. Such data
will be essential if we are ever to begin unraveling the
evolution of this odd reproductive mode.
While important in its own right, the genetic confirmation of polyembryony was critical in allowing subsequent examination of a long-standing issue in armadillo behavioral ecology, namely the potential for kin
selection. Because of the high coefficient of relatedness
among littermates, armadillos, like other vertebrates
that also produce genetically identical offspring, were
long considered potential model systems for the study
of kin selection (Bell 1997). The first explicit mention of
this was by Dawkins (1976): Nine-banded armadillos
are born in a litter of identical quadruplets. . . . it has

been pointed out that some strong altruism is definitely
to be expected, and it would be well worth somebodys
while going out to South America to have a look.
In order for kin selection to be important, several criteria must be met. For example, kin must have opportunities to interact with one another. Thus, in D. novemcinctus, the question became: are littermates spatially
clustered so as to promote interactions that might be
favored by kin selection? As McDonough and Loughry
describe elsewhere in this volume, long-term mark-recapture data provide one means of answering this question by tracking patterns of juvenile recruitment and
movement (see also Loughry and McDonough 2001).
Genetic confirmation of polyembryony provided an alternative, and much faster, approach: genetic screening
of individual tissue samples allowed identification of
all possible sibships in a population because the occurrence of a complete match in microsatellite markers between any two individuals meant they had to be littermates. This approach had the added benefit of allowing
identification of sibships among animals first caught as
adults or as solitary juveniles, where litter membership
was unknown.
When Prodhl et al. (1996) performed such an analysis they found: (1) Few adults had any sibs present.
Only 16 of 196 adults sampled had at least one sibling
present (these 16 animals were divided into two groups
of triplets and five groups of twins), and (2) While juvenile siblings were often found in tight spatial clusters where interactions among kin might be possible,
208
Figure 20.2. Spatial dispersion of adult () and juvenile () cloneships in a population of nine-banded armadillos. Adult littermates occurred infrequently in the population and were usually widely separated from one another. Reprinted from Prodhl
et al. (1996).
adult siblings were widely separated from one another

(mean distance between adult sibs was 455.9 m; figure
20.2), suggesting low rates of interaction. These conclusions were subsequently confirmed by long-term
observational data: recruitment of juvenile siblings was
low (possibly because of high juvenile mortality from
predation, McDonough and Loughry 1997a) and, in the
few instances where multiple littermates were recruited,
they were widely separated from one another (Loughry
and McDonough 2001). Thus, for adult armadillos in
this population, it appears that opportunities for kin
selection were minimal.
Although the findings above do not preclude the
possibility of kin selection occurring among more distant relatives or at different ages (e.g., perhaps among
juveniles), all the evidence to date supports the hypothesis that kin selection has been relatively unimportant
in shaping armadillo behavior and population structure
(see Loughry et al. 2005 and McDonough and Loughry

this volume for further discussion of this issue). Again,
we wish to reemphasize that this conclusion is based
on data from a single site at the extreme northern edge
of the species range, in an area that had only been inhabited by armadillos for about 2530 years. It remains
to be seen whether kin selection is also relatively unimportant in D. novemcinctus populations in more ancient parts of the species range or in the other reputedly
polyembryonic species in the Dasypus genus. Interestingly though, the data from D. novemcinctus do seem
to support a general pattern seen in vertebrates that
produce genetically identical offspring (i.e., the parthenogenetic reptiles, amphibians, and fishes; Cole 1975;
Cuellar 1977; Vrijenhoek et al. 1989) in which no obvious kin selection has ever been reported (Loughry et
al. 2005). The fact that nine-banded armadillos fit this
broader pattern might lend support to the argument
that data from this one population are not abnormal or

unrepresentative.
Reproductive success
Determination of reproductive success is critical for
understanding evolutionary processes within a population and also for describing the mating system. Population biologists are developing a number of analytical
techniques for reconstructing pedigrees for wild populations strictly from genetic data (Blouin 2003). Such
techniques are essential in species such as the ninebanded armadillo where fitness cannot be estimated
by observing male and female reproductive behavior
and/or interactions between adults and juveniles. There
are several reasons for this. First, although males and
females form reproductive pairs for some portion of the
breeding season (Jacobs 1979; McDonough 1997), observations of copulations are rare, so, as in many species
(Gowaty 1996; Griffith et al. 2002), it is difficult to know
whether social partners are actually genetically related
to the young produced. Second, there is no evidence
that males provide parental care or associate with their
offspring after birth, so there are virtually no behavioral
cues one can use in assigning paternity of young. Third,
assignment of maternity is also problematic. Females
share burrows with their young prior to, and for some
time after, weaning. However, after juveniles emerge
from their natal burrows, they are rarely seen above
ground in association with the mother. Siblings often
forage with one another, but it is unusual to see them in
close proximity to an adult. Thus, if one wishes to determine the maternity of a litter, the only means of doing
so is to excavate a burrow and identify the adult female
sharing the burrow with a set of juveniles. This is, to say
the least, extremely labor intensive and often impractical, and may even be unreliable because adult females
may share burrows with one another (Lehmann, cited
in Kalmbach 1943; Bushnell 1952; Pacheco and Naranjo
1978; Herbst and Redford 1991; Schell 1994).
The above reasons all point toward the need to use
genetic markers for determining armadillo reproductive success. Nonetheless, behavioral data provide some
bases for predictions about expected outcomes. For example, contrary to the typical mammalian pattern of
extreme polygyny and conspicuous male sexual dimorphism (Kleiman 1977), observations of male-female reproductive pairs suggest minimal levels of polygyny:
females are usually seen paired with just one male during the breeding season, whereas males are often seen
paired with 23 different females (McDonough 1997).
209
In addition, although in some populations males are

slightly larger than females (McDonough 1997), more
commonly there is no obvious sexual dimorphism
(Loughry and McDonough 1996, 1998a).
Genetic analyses have generally confirmed this picture. Using the same microsatellite markers described
above, Prodhl et al. (1998) used a maximum-likelihood
approach (Thompson and Meagher 1987) to construct
pedigrees from genetic data. In total, they analyzed 68
armadillo litters, sampled between 1992 and 1995, and
found that (1) most adults in the population apparently
failed to reproduce at all in the four years of the study;
(2) reproductive success was low and relatively similar
for males and females; and (3) there was no relationship
between reproductive success and any obvious phenotypic feature (e.g., body size) of either males or females
(Loughry, Prodhl, McDonough, Nelson et al. 1998).
There was some weak evidence of polygyny in that six
males sired more than one litter during the same breeding season (out of a total sample of 138 males, 46 of
which were reproductively successful). However, the
data primarily supported the view that nine-banded
armadillos do not fit the pattern of typical polygynous
mammals and that, over the four years analyzed, fitness
was random with respect to those features of males and
females measured. Whether these results hold over longer periods of time or for populations in other locales is
currently unknown, but once again suggests an important area for future study.
Between population analyses

Avise (2000) defined phylogeography as the study of
the principles and processes governing geographic
distributions of genealogical lineages. At this writing,
there have been 14 such studies involving four xenarthran species, and we review each in turn.
Nine-banded armadillos represent an excellent system in which to perform phylogeographic analyses
because the historical record of their colonization of
the United States is so well documented (Humphrey
1974; Taulman and Robbins 1996). As mentioned earlier, initial genetic analyses of D. novemcinctus populations in the United States using protein electrophoresis
revealed virtually no variation (Ramsey and Grigsby
1985; Moncrief 1988). More recent work has shown that
this monomorphism is confined to the United States
and that populations in more ancient parts of the species range in Latin America are more variable (Huchon
et al. 1999). However, the limited usefulness of protein
210
data has led to analyses using more refined molecular

markers.
Consistent with Huchon et al.s findings from protein
data, Loughry, Prodhl, McDonough, and Avise (1998)
presented preliminary data from nuclear microsatellite
markers that also showed reduced genetic variation in a
Florida population of D. novemcinctus versus one from
the Atlantic coastal rainforest of Brazil. Huchon et al.
(1999) extended their work on protein variation by using part of the mtDNA D-loop sequence to compare
populations in French Guiana with those in Louisiana,
Texas, and Mississippi (U.S.A.). They found clear differences in both the type and number (2 versus 10) of haplotypes present in North and South American populations. Elrod (2000) also examined variation in D-loop
sequences among four populations from Arkansas and
Texas (U.S.A.), finding reduced genetic variation in all
populations relative to other mammals, but higher sequence diversity in a semi-isolated population than in
either a completely isolated or two free-ranging ones.
These data are all consistent with the notion of a population bottleneck or founder effect in the colonization of
the United States by nine-banded armadillos, although
at this time we do not know precisely when or where
this bottleneck occurred.
Within South America, Frutos and Van Den Bussche
(2002) found substantial variation (25 haplotypes) when
examining the mitochondrial cytochrome-b sequences
from five populations of D. novemcinctus in Paraguay.
Interestingly, most variability occurred within, rather
than between populations, suggesting substantial gene
flow due to long-distance dispersal of females. Two
more recent studies used either mitochondrial (Delsuc et al. in preparation) or nuclear and mitochondrial
markers (Prodhl et al. in preparation) from a large
number of populations in the United States and Latin
America. Preliminary results confirm the occurrence
of a bottleneck during the colonization of the United
States and reveal a high level of phylogeographic structure across the species range, with well-defined genetic
entities corresponding at least in part to morphologically proposed subspecies (Wetzel and Mondolfi 1979).
However, at the time of writing, analyses of these data
were incomplete, so they cannot be described further
here.
Unlike the case in D. novemcinctus, which is relatively widespread and abundant, phylogeographic analyses of other xenarthran populations have been largely
motivated by conservation concerns. Moraes-Barros
and coworkers provided the first description of genetic
variability in sloths (Moraes-Barros et al. 2002, 2006;

Moraes-Barros 2003), using both mitochondrial and
nuclear markers to examine populations of the endangered maned sloth, Bradypus torquatus, and the more
widespread B. variegatus in Brazil. In both species,
there was very low genetic variability within populations but dramatic differences between them. In related
work, Lara-Ruiz (2004) examined three populations of
B. torquatus in Brazil and found extreme genetic differences between them. Indeed, while genetic diversity
within each population was very low, the genetic differences between populations were of the same magnitude often observed between species. This is perhaps
not surprising as gene flow between populations would
seem extremely unlikely given the fragmented habitats
occupied by B. torquatus and the limited movements
of individuals (see Chiarello this volume). In total,
these results suggest serious reconsideration of certain
management plans for this species in Brazil, where the
current practice is to release all rehabilitated or confiscated animals at one site, thus mixing distinct genetic
lineages.
Lastly, Garcia et al. (2005) identified six microsatellite loci in the giant anteater, Myrmecophaga tridactyla,
which could be used in population genetic analyses of
this species. A preliminary study conducted with these
markers at Emas National Park in Brazil examined
samples from 30 animals. Two notable results were (1)
a high endogamy coefficient with homozygous excess,
indicating inbreeding, and (2) parentage analyses showing that males using contiguous areas were not closely
related (Leite et al. 2004). Extension of population genetic and ecological studies, not only at Emas, but at
other sites as well (see Rodrigues et al. this volume),
will be imperative in developing effective management
plans for conservation of this species (J. E. Garcia personal communication).
Summary
Genetic analyses of the Xenarthra are in their infancy
(Garcia 2003). Apart from D. novemcinctus, virtually
no other xenarthran has been subjected to intense
genetic scrutiny. This cannot be attributed to technical problems; numerous genetic markers have already
been identified and the basic screening procedures are
relatively straightforward. Indeed, with the sequencing
of the D. novemcinctus genome (Chang and Adams this
volume), we now have a multitude of genetic markers
available. What is required is the difficult task of ob-
taining genetic samples from wild xenarthran populations. As we have described above, there are numerous
reasons why more genetic sampling is needed. While
some of this may be targeted at answering questions of
basic scientific interest, much effort will undoubtedly
be directed at the more practical issues of conservation.
With so many xenarthrans listed as vulnerable or endangered (Aguiar and Fonseca this volume), it is critical that we obtain genetic data from as many species
and as many populations as quickly as possible.
211
Acknowledgments
John Avise and the staff of the Tall Timbers Research
Station provided invaluable help with our work, which
has been funded over the years by various Faculty
Research Awards from Valdosta State University, the
American Philosophical Society, the Organization of
American States, and Earthwatch. Thanks to J. Garcia,
N. Moraes-Barros, and S. Vizcano for comments on a
draft of this chapter.
4
Living Xenarthra
Conservation
21
Conservation status of the Xenarthra
John M. Aguiar and Gustavo A. B. da Fonseca
histria evolucionria em uma concentrao alm da

proporo representada pelo seu nmero despretensioso de espcies. Em conjunto com a funo crtica que
exercem nos ecossistemas onde vivem, esta linhagem,
com sua filogenia antiga, focaliza nossa ateno no status de conservao delesque bastante preocupante
para a ordem inteira. A Lista Vermelha do UICN atualmente inclui cinco espcies em categorias de ameaa:
Bradypus pygmaeus, a preguia-pigmeu recentemente
descrita, agora reconhecida como Criticamente Em
Perigo (CR); Bradypus torquatus, a preguia-de-coleira,
permanece Em Perigo (EN); e Chaetophractus nationi,
Dasypus pilosus, Priodontes maximus e Tolypeutes
tricinctus so considerados Vulnerveis (VU). Sete espcies adicionais so classificadas como Quase Ameaadas (NT), mais duas como Dados Insuficientes (DD),
e dezesseis como Menor Risco (LC). Embora menos
espcies so ameaadas entre os xenarthra do que em
outras ordens, estes nmeros s prognosticam o perigo
enfrentado por todos os xenarthra se a destruio de
habitat, a perseguio, e a explotao intensiva continuassem com esta atual severidade. Aqui revisamos
o status de conservao de cada espcie nesta ordem
excepcional, resumimos a distribuio e as ameaas
sobrevivncia deles, e fornecemos recomendaes para
ao nos nveis regional e nacional.
Resumen
Siendo uno de los cuatro principales clados de mamferos, los xenartros actuales representan una riqueza de
historia evolutiva que est concentrada desproporcionadamente en su modesto nmero de especies. Junto a
su rol crtico en el funcionamiento de los ecosistemas,
la antigedad de este linaje obliga a prestar mucha atencin a su estado de conservacinla situacin es, como
mnimo, preocupante para el orden en su conjunto. Actualmente, la Lista Roja de la UICN incluye seis especies de xenartros en categoras amenazadas: Bradypus
pygmaeus, el perezoso pigmeo que fue descrito hace
poco, est reconocido como En Peligro Crtico (CR);
Bradypus torquatus, el perezoso de tres dedos, sigue
clasificado como En Peligro (EN); y Chaetophractus
nationi, Dasypus pilosus, Priodontes maximus y Toly
peutes tricinctus son considerados Vulnerables (VU).
Adicionalmente, siete especies estn clasificadas como
Casi Amenazadas (NT), dos como Datos Insuficientes
(DD) y diecisis como Preocupacin Menor (LC). A
pesar de que menos especies de xenartros que de otros
rdenes estn consideradas amenazadas, estos nmeros
no hacen ms que expresar el peligro al que estn expuestos los xenartros si la destruccin de su hbitat y
su persecucin y explotacin intensiva siguen con la
misma severidad que en la actualidad. Aqu se revisa
el estado de conservacin de cada especie de este orden
excepcional, se resume su distribucin y los peligros
que estn enfrentando, y recomendamos acciones a
nivel regional y nacional.
Introduction
Thirty-one species constitute todays Xenarthraa
group of sometimes outlandish, often radically specialized creatures that account for barely one-half of one
percent of the worlds current mammal diversity. By
the numbers, they would seem to be no more than irrelevant oddities, strange relics from a faded chapter of
South Americas past.
Resumo
Como um dos quatro clados principais de mamferos,
os xenarthra existentes incorporam uma riqueza de
215
216
J. M. Aguiar and G. A. B. Fonseca
But the handful of xenarthran species now alive in

the Neotropics are survivors, not only of South Americas long sojourn as an island continent, but also of a
time far deeper in evolutionary history, close kindred
to the earliest of the placental mammals. Long imagined to be the oldest surviving remnants of these first
ancestors, the Xenarthra have been redefined by recent
molecular analyses as one of the four primary clades of
mammalian diversity, and potentially a basal offshoot
of the founders of the eutherian line (Madsen et al.
2001; Murphy, Eizirik, Johnson et al. 2001; Springer et
al. 2005; Delsuc and Douzery this volume). Not at all
the inconsequential fringe their numbers might suggest, the Xenarthra in fact occupy a pivotal position:
an ancient clan, monophyletic despite all appearances,
whose members stand equal to great clades encompassing hundreds or thousands of species apiece. This
unique concentration of evolutionary history marks the
xenarthrans as invaluable; should they be lost, some of
the earliest chapters of mammalian evolution will remain forever unread.
Despite their importance for understanding this history, most xenarthrans are inadequately studied or effectively unknown. Rarely conspicuous or easily found,
many species are nocturnal burrowers, while others
are lifelong residents of the high forest canopy; some
exist in environments so harsh and remote that even
the sternest field biologists are given pause. Owing to
the general lack of knowledge about their ecology and
population dynamics, their conservation status remains uncertain, long after detailed assessments have
thoroughly covered other mammal groups.
Human impacts on ecosystems, which had been a
low-level pressure for thousands of years, have abruptly
heightened in the past century, and many species of
xenarthrans have suffered critical declines. To address
the expanding issues of xenarthran conservation, the
International Union for the Conservation of Nature
(IUCN)/Species Survival Commission (SSC) Edentate Specialist Group (ESG) was reconstituted in 1994,
in large part to better organize xenarthran researchers
for a new assessment of the twenty-nine species then
recognized. (Although no longer widely used, the term
edentate was retained for continuity in publications).
The following year, the newly reorganized ESG invited
a group of active xenarthran specialists to contribute
written assessments for the 1996 IUCN Red List, which
included among its goals an evaluation of the conservation status of all mammal species then described.
Although not every member of the Specialist Group
was able to contribute, enough replies were received

to allow an analysis of all twenty-nine species. The final results, published in 1996, listed three xenarthrans
as Endangered, five as Vulnerable, two more as Lower
Risk/Near Threatened, and an additional six as lacking
the necessary data for a full assessment (IUCN 1996;
Fonseca and Fonseca 1997; table 21.1).
In 2004 the Edentate Specialist Group organized an
Edentate Species Assessment Workshop to re-evaluate
the Xenarthra using the most recently revised Red List
criteria (IUCN 2003). Hosted by Conservation International do Brasil in Belo Horizonte, the workshop was
conducted in concert with the Global Mammal Assessment, an initiative of the IUCN with the mandate of
assessing the conservation status of all known mammal
species. Drawing on the extensive field experience of a
diverse group of xenarthran researchers, the workshop
re-evaluated the twenty-nine original species and included two more that had been recently described. The
status summaries and range maps developed during the
workshop were published in Edentata, the newsletter
and journal of the Edentate Specialist Group, together
with notes on the discussions that took place during the
workshop (Aguiar 2004; Fonseca and Aguiar 2004).
One of the newly described species, the Panamanian
pygmy sloth (Bradypus pygmaeus), became the first
xenarthran to be considered Critically Endangered, owing to its small population and miniscule range. Only
one species, the maned sloth (B. torquatus), retained its
status as Endangeredthe other two were downgraded
owing to the application of the new Red List criteria
while four species were listed as Vulnerable. Another
seven were classified as Near Threatened, two as Data
Deficient, and sixteen as Least Concern (table 21.1).
In addition to the conservation assessments themselves, the Edentate Species Assessment Workshop
produced new range maps for each species, working
from base maps supplied by the Global Mammal Assessment. The final versions which appear here are derived from those working maps and represent the best
estimatebased on current fieldwork, local expertise,
and conservative extrapolationof where each species
is distributed today. (Please note that maps for the two
species with the smallest ranges, Bradypus pygmaeus
and Dasypus yepesi, are not presented here). Much of
the information we present in this chapter, especially
on the particular threats to each species, derives from
the workshop proceedings and follow-up comments
from the workshop participants. The Edentate Specialist Group extends its appreciation to each of them, for
217
Table 21.1. Conservation Status and Main Threats to the Extant Xenarthrans

Species
1996 Status
2006 Status

Sloths

Bradypus pygmaeus
NE
CR B1ab (i,ii,iii)

Bradypus torquatus
EN A1cd
EN B1ab (i,ii,iii)

Bradypus tridactylus
LR
LC

Bradypus variegatus
LR
LC

Choloepus didactylus
DD
LC

Choloepus hoffmanni
DD
LC
Anteaters

Cyclopes didactylus
LR
LC

VU A1cd
NT

Tamandua mexicana
LR
LC

LR
LC
Armadillos

Cabassous centralis
DD
DD

Cabassous chacoensis
DD
NT

Cabassous tatouay
LR/NT
LC

Cabassous unicinctus
LR
LC

Chaetophractus nationi
VU A1d
VU A2d

Chaetophractus vellerosus
LR
LC

Chaetophractus villosus
LR
LC

Chlamyphorus retusus
VU A1c
NT

Chlamyphorus truncatus
EN A1abcd
NT

Dasypus hybridus
LR
NT

Dasypus kappleri
LR
LC

LR
LC

Dasypus pilosus
VU B1+2c
VU B1ab(iii)

Dasypus sabanicola
DD
LC

Dasypus septemcinctus
LR
LC

Dasypus yepesi
NE
DD

Euphractus sexcinctus
LR
LC

Priodontes maximus
EN A1cd
VU A2cd

Tolypeutes matacus
LR/NT
NT

Tolypeutes tricinctus
VU A1bcd
VU A2bc

Zaedyus pichiy
DD
NT
Primary threats to
native populations
hunting, minimal range
population fragmentation
hunting, habitat loss
habitat loss
habitat loss
hunting, fire, highway mortality,
habitat loss, persecution
fire, habitat loss, highway mortality
fire, hunting, habitat loss, highway mortality
habitat loss and degradation
hunting, habitat destruction
intense hunting and habitat loss
hunting
hunting
intense persecution, habitat loss
habitat loss
intense hunting and habitat loss
unknown
hunting
habitat loss
hunting
unknown
unknown
hunting, habitat destruction
hunting, habitat loss and degradation
Sources: Aguiar (2004), Fonseca and Aguiar (2004), IUCN (2006).

Notes: All species were assessed using the IUCN (2001) criteria. Abbreviations are those used by the IUCN (2003): CR = critically endangered; EN = endangered; VU = vulnerable; NT = near threatened; LC = least concern; DD = data deficient; NE = not evaluated. For 1996
designations, LR = low risk, which is similar to the current LC category. For definitions of status subcategories, see IUCN (2001).
without their dedication to the Xenarthra and their

willingness to share their expertise, the following pages
would not have been possible.
Sloths
Of the six species of sloths now recognized, four are
thought to be in essentially stable condition, with the
caveat that slow-moving vertebrates are almost universally hunted throughout the Neotropics. The two
species of two-toed sloths, Choloepus didactylus and C.
hoffmanni, are not considered to be under any particular threat overall. C. didactylus, by far the more wide-
spread of the two, is plentiful throughout the northern

Amazon and the Guyanas, occupying some of the most
pristine rainforest remaining in South America (figure
21.1A). Its congener to the west is in a more complex
situation; like that of many mammals in the region, the
range of C. hoffmanni is discontinuous (figure 21.1B),
with populations in Central America and coastal Colombia separated by the Andes from their kindred in
the southwestern Amazon. If these divided populations
are recognized as subspecies, the northernmost would
almost certainly qualify for threatened status, owing to
the rampant degradation of rainforest throughout Central America. The Amazonian population is in no im-
218
A.
B.
C.
Figure 21.1. Current distributions

of sloth species. Note that, because
of its extremely small range (see
text), the distribution for Bradypus
pygmaeus is not shown.
A. Choloepus didactylus.
B. C. hoffmanni.
C. Bradypus tridactylus.
D. B. variegatus.
E. B. torquatus.
D.
E.
mediate danger, however, and represents a stronghold

for the species.
The four species of Bradypus, the three-toed sloths,
span the full gamut of conservation concern. Like
Choloepus didactylus, both Bradypus tridactylus and
B. variegatus are widespread (figures 21.1C and 21.1D,
respectively). These species are locally abundant and
present within a spectrum of protected areas, from
inner-city parks to wilderness reserves. The immense
area occupied by B. variegatusthe largest range of any
slothalmost certainly harbors local genetic variants
that, according to preliminary genetic analyses, may
be valid subspecies (Moraes-Barros et al. 2002, 2006;

Moraes-Barros 2003; Lara-Ruiz 2004). One or more of
these potential subspecies may well prove to be under
local threat; but as a whole, the variegated sloth is one
of the most abundant and secure of all xenarthrans.
The opposite is true of the pygmy sloth, B. pygmaeus,
described in 2001 and now recognized as Critically Endangered, with a single population that only numbers
in the hundreds. Confined to a tiny island off the Caribbean coast of Panama (Anderson and Handley 2001),
the pygmy sloth has evolved in isolation over the past
9,000 years, ever since its mainland progenitors were
trapped when rising sea levels created the islet of Escudo de Veraguas. No more than 5 km2 in total area, this
was the first of several islands to form in what became
the coastal archipelago of Bocas del Toro. Each of the
seven main islands in this group harbors a breakaway
population of B. variegatus, their body size dwindling
through the millennia as each island was cut off in turn;
but only on Escudo de Veraguas, the oldest and most
distant of the group, has this insular dwarfism resulted
in a separately recognized species. Unusual for sloths,
B. pygmaeus are mangrove specialists, able to survive in
an environment too harsh with salt and sun and waterstress for most other mammals to endure for long. Less
than one-third of their island is covered in mangrove
forest, however, giving the pygmy sloths a halo of habitat covering no more than 1.5 km2the entire global
range of the species.
The pygmy sloth was detected and described in a
comprehensive survey of island sloths by Anderson
and Handley (2001), and the new species is the centerpiece of their research on dwarfism and insular evolution in three-toed sloths (Anderson and Handley
2002). Its importance to island biogeography, however,
is not matched by its protected status: there are no legal
safeguards for the pygmy sloth, which is hunted freely
by the local fishermen and native people, who set up
seasonal camps on this otherwise uninhabited island.
There is no presence of authority on the island, and no
enforcement of wildlife law; the island itself is a part of
the Comarca Indigenous Reserve, a semiautonomous
region of coastal Panama granted to and governed by
the Kuna Indian nation. The fishermen in the temporary camps are more intent on the marine resources
than on the island itself; but as long as their hunting is
uncontrolled, this least-known and rarest of the Xenarthra will be suspended on the edge of evanescence.
The prevailing conditions are very different for the
maned sloth, B. torquatus, endemic to the remnants of
the Atlantic Forest on the coast of southeastern Brazil.
Once ranging from Rio de Janeiro north to Pernambuco, maned sloths now survive in three isolated regions: southeastern Bahia, near Ilheus; the mountains
of southern Esprito Santo; and the vicinity of the city
of Rio de Janeiro itself (figure 21.1E). Although population estimates are uncertain, the catastrophic loss of
forests throughout the region in the past centuryand
especially within the last 30 yearshas entirely eliminated maned sloths from their former habitat in Sergipe, Alagoas, northern Bahia, and Pernambuco. The
remaining strongholds are divided by a broad gap in
219
northern Esprito Santo, caused partly by forest loss

and partly by the presence of surviving deciduous
forests, which are unsuitable habitat for B. torquatus.
But even the presence of apparently palatable forest is
no guarantee that maned sloths will thrive, and what
protected areas exist are not always ideal. The Fazenda
Unio Ecological Reserve, for instance, contains extensive eucalyptus plantations, and the Poo das Antas Biological Reserve has many grassy and disturbed areas,
all of which are inhospitable to B. torquatus.
The resulting distribution of the maned slothtwo
smaller enclaves south of the main gap, and a larger redoubt to the north in Bahiais reflected in strong patterns of genetic isolation, which may bear the ancient
stamp of Pleistocene refugia. The population in each
of these three subregions is markedly distinct from the
other two, with the northernmost population in Bahia
showing the greatest divergence of all (Lara-Ruiz 2004).
This pattern, which parallels those of other Atlantic
Forest mammals (Costa et al. 2000), implies a degree
of isolation and differentiation that long predates any
recent human changes to the landscape, and suggests
that the northernmost population may in fact be a distinct subspecies. In addition to their marked genetic
distinctions, each population also shows a pattern of
internal homogeneity, the combination of which suggests there is no gene flow among the three disjunct
regions. This, in turn, obliges special care to be taken
with any program of translocation or captive release, in
order to avoid the genetic contamination that would result from an ill-considered transfer between otherwise
isolated regions.
Anteaters
Among the four species of anteaters, the least worrisome is the silky anteater, Cyclopes didactylus, which
adapts easily to a remarkable spectrum of habitat types
within its three-thousand-mile range (figure 21.2AB).
Like all other xenarthrans, the silky anteater is helplessly vulnerable to commercial logging, forest-burning, and other methods of wholesale destruction, but
its small size (150300 g) and nocturnal, unobtrusive
habits protect it from the intensive hunting that claims
countless other Neotropical mammals. Ranging from
southern Mexico to the mouth of the Amazon, found
throughout the Amazon basin and much of the Guyana
Shield, silky anteaters as a whole may be considered secure for as long as their varied habitats remain intact.
This is, however, a general estimate based on their
220
great range and ecological flexibility, but we know next

to nothing of their geographic variation. Occupying
such a wide variety of local habitatsfrom dense rainforest and riparian vegetation to the savannas of the
Cerradothe silky anteater may well include a number
of distinct subspecies awaiting recognition. No genetic
studies have been completed, but the one population in
greatest need of attentionand most likely to resolve as
a distinct subspeciesis an outlier in northeastern Brazil, near the city of Recife in the state of Pernambuco.
Lighter in its pelage than other Cyclopes, this population is found only in Pernambuco, where the Atlantic
Forest has been virtually destroyed. If confirmed to be
taxonomically distinct, this population might qualify
as Endangered or Critically Endangered. Resolving the
taxonomic status of this easternmost population should
be the first priority for any conservation research focusing on the silky anteater, ideally as part of a phylogeographic examination of Cyclopes populations throughout its range.
The remaining two anteater genera, Tamandua and
Myrmecophaga, cover a far greater area of the Neotropics than Cyclopes (figure 21.2CF), but their wide
range alone offers no sure protection. Both species of
Tamandua, the lesser anteaters, are comparatively secure when considered as a wholebut this is no comfort
to the countless individuals killed by cars on highways,
or caught in the wildfires and semicontrolled burns that
follow the continual expansion of agricultural lands.
Beyond the direct mortality they cause, these fires also
affect lesser anteaters in subtler ways, fragmenting and
degrading whatever habitat remains between the newly
cleared farms, and forcing the anteaters to cross dangerous open ground to travel between the shards of
their former habitat.
Much larger and easier targets than silky anteaters,
the lesser anteaters are vulnerable to hunting, although
the degree depends upon the local culture and how edible they are considered to be. Tamanduas in the Isoso
district of the Bolivian Chaco, for instance, are not eaten
by the local people, although they are sometimes killed
by domestic dogs (Noss et al. this volume). In Honduras they are considered inedible and commercially useless, but their fierceness is exaggerated in local legend,
and they may be killed out of misplaced fear (Marineros and Gallegos 1998). Sometimes they are simply shot
for no reason other than fleeting amusement; in some
parts of Central America they are casually hunted as
meat for domestic dogs (Reid 1997) and they are often
hunted by the dogs themselves. In the Bolivian Chaco
they are killed for their skins, and their dried tongues
are sometimes kept as a good-luck charm (Cullar and
Noss 2003).
Given the differences in human attitudes, as well as
the condition of local ecosystems, certain regional populations may be under particular threat. The Livro Vermelho of the Brazilian state of Rio Grande do Sulnear
the southernmost limits of the tamanduas distributionlists Tamandua tetradactyla as Vulnerable owing
to the severe degradation of its forest habitat (Fontana
et al. 2003). There are also unresolved taxonomic issues,
especially with T. mexicana in Central America, where
the status of its several subspecies requires careful revision. Despite these issuesand the pandemic threat
to individuals throughout their respective rangesthe
two species of lesser anteater are still widespread and
populous, with a strong presence in protected areas in
many parts of the Neotropics.
The situation is less reassuring for the giant anteater,
Myrmecophaga tridactyla, the largest of the Neotropical anteaters and as broadly distributed as the tamanduas (figure 21.2EF). Subject to all the same threats
which the latter endure, the giant anteater is far more
vulnerable: its greater mass of flesh ensures that it is
actively hunted throughout its range, rather than opportunistically shot. Like the giant armadillo, its fellow
heavyweight in the xenarthran clan, the giant anteater
is a large, relatively long-lived animal that survives in
a remarkable diversity of landscapes and vegetation
types, from floodplain rainforest to the near-desert of
the Bolivian Chaco. One of the most morphologically
extreme of an already specialized group, it is capable
of surviving wherever ants and termites are found. But
that versatility is also its weakness: the giant anteater
survives in an area only as long as the social insects
on which it dependsand if they are extirpated, Myrmecophaga will vanish soon after.
Giant anteaters fall prey to all the hazards of human
impact: they are killed on highways, by domestic dogs
and during deliberate burnings of the grasslands they
often inhabit. Hunting is a constant danger, and severe
enough in certain regions to decimate local populations. Sport and subsistence hunting are both prevalent in Bolivia (Ergueta S. and Morales 1996). In the
Paraguayan Chaco, their tough, leathery skinreputedly too thick for vultures to penetrateis cured and
worked into riding chaps and other equestrian gear
(Meritt this volume); the hide is used for leatherwork
in the Bolivian Chaco as well (Cullar and Noss 2003).
At the opposite end of its range, Reid (1997) considers
A.
B.
D.
C.
F.
E.
Figure 21.2. Current distributions of anteater species.

A. Cyclopes didactylus, detail of South American distribution.
B. C. didactylus, detail of Central American distribution.
C. Tamandua mexicana.
D. T. tetradactyla.
E. Myrmecophaga tridactyla, detail of South American
distribution.
F. M. tridactyla, detail of Central American distribution.
222
it the most endangered mammal in Central America

for the dangers it faces from fire and persecution.
Those which survive the bullets and burningand
the constant habitat transformationare often isolated
and reproductively disjunct from neighboring demes,
and so the momentum of local extinctions may present a serious threat to the species as a whole. Moreover,
giant anteaters often bear an undeserved reputation as
vicious, hostile animals that should be shot on sight for
the sake of human safety. Although their aggression
is nine-tenths bluff and self-defense, it has provoked
fear and retaliation in many parts of their range. In Rio
Grande do Sul, the southernmost state of Brazil, giant
anteaters are customarily killed at every opportunity,
and their bodies discarded uneaten (Fontana et al.
2003). In Venezuela they are likewise killed from fear
and thoughtlessness, but also for the opportunity to
take their large foreclaws (Linares 1998).
Against these many threats, the only reliable
strengths of the giant anteater are its vast range and its
ecological versatility; for the time being, these appear
to be enough. Although individuals are at risk wherever
they go, and populations may disappear as agricultural
lands expand, many other populations remain relatively
intact. The species is present in many protected areas in
the Amazon and the Brazilian Cerrado, and for the moment it is not in immediate danger of extinction.
Armadillos
Like the giant anteater, the giant armadillo (Priodontes maximus) survives across much of South America,
from northern Argentina to the Caribbean coast of
Venezuela and the Guyanas (figure 21.3), but nowhere
is it truly abundant. Although capable of persisting in a
variety of habitats, the giant armadillo is by no means
evenly distributed across the continent; its actual occurrence is unpredictable from place to place, but it is
often encountered in the Chaco and the llanos of Guyana (D. Meritt personal communication).
The giant armadillo may reach weights of 40 kg or
more, and such a massive animal represents a spectacular catch for a subsistence hunter. Giant armadillos were historically pursued throughout their range by
indigenous peoples, some of whom still rely heavily on
this species for their protein intake (Leeuwenberg 1997;
Parera 2002; D. Meritt personal communication, see
also Noss et al. this volume). Given their rarityand
their understandable caution around humansthere is
no precise information on remaining populations; but
Figure 21.3. Current distribution of the giant armadillo,

Priodontes maximus.
the species has most likely endured a two-tiered decline,

in which the original populations were first decimated
decades ago, and the remnants are now being whittled
down to nothing (A. Rylands personal communication). Estimates of the most recent losses vary a great
deal, but a major decline has clearly taken place over
the past ten years, and the entire population may have
been cut by as much as 50% over the past three decades
(P. Lara-Ruiz personal communication).
In Bolivia, where giant armadillos are hunted both
for food and for raw materials for handicrafts, Ergueta
S. and Morales (1996) warn that it has become rare, and
without conservation action may become endangered
in that country. In Ecuador, it is avidly pursued for its
exquisite meat (Tirira S. 2001) as well as for its shell
and claws in the making of ornaments. In Venezuela,
giant armadillos are intensively hunted wherever they
are found, even in forest reserves (Linares 1998). Recently a new sort of predation has made an impact: the
capture of giant armadillos for the wildlife trade, in
which the black-market price for a grown Priodontes
may reach into the tens of thousands of dollars. The
details of this traffic are unclear, but many giant armadillos are thought to have died in confinement before
they could be sold or delivered.
Although giant armadillos are the most likely to
draw immense prices, owing to their rarity and novelty value, most other armadillo species are also hunted
wherever they are foundsometimes only casually, but
often with a force that threatens local extinctions (Noss

et al. this volume). The intensity of hunting in any given
region depends in large part on the economic climate
of the local human communities. Aside from the value
of the meat itself, either eaten directly or sold in markets, armadillos are also harvested for the pet trade
and many thousands more to supply raw materials for
traditional handicrafts. Of these, the most devastating
to armadillo populations has been the guitar-like charango, one of the most popular symbols of Andean culture. Originating in colonial Bolivia, these folk instruments are direct descendants of the Spanish vihuela de
mano, popular in the 1500s and soon imported to the
New World metropolis of Potos. Although originally
built of wood, the Andean natives in their forest-poor
mountains quickly adapted the instrument to use an
armadillos shell as the sounding box, a design which
has survived for centuries. Many modern charangos
are hand-crafted of wood, which musicians consider
finer-sounding and more durable; but many more are
still constructed from the shells of quirquinchos (Chaetophractus nationi and Cha. vellerosus) for traditional
markets and the tourist trade.
For the quirquinchos, as with most other armadillos, ceaseless hunting is the primary threat to their
existencefollowed closely by the continuous conversion of native habitat to croplands and pasture.
For most species, this steady agricultural creep brings
with it a swarm of associated dangers: chemical runoff
A.
B.
223
from pesticides and fertilizers, cascading changes to

the surrounding soil and biota, the threat of car strikes
on newly opened roads. But changing landscapes may
also offer unexpected benefits, and for a few of the more
adaptable opportunists, new fields are a magnet for the
wealth of upturned insects they provide. Chaetophractus vellerosus (figure 21.4A) is one of these; it is heavily
hunted for food in Argentina and Bolivia, and also for
shells to supply the inexhaustible market for charangos. But Cha. vellerosus is drawn to the insects in newly
cultivated fields, and in the Paraguayan Chaco the quirquinchos have become much more common in the past
ten years to the point where they are now hunted as
agricultural pests (Meritt this volume, M. Superina personal communication).
To the north in the Bolivian altiplano, Cha. nationi
(figure 21.4B) is also intensively hunted to supply raw
materials for the charango trade, as well as for other
instruments and small handicrafts used for divination.
The combined pressure is ferocious: nearly 2,000 individuals are harvested each year (Cceres 1995), despite
a core population which may number no more than
13,000 (Peredo 1999). In the past few years hunters have
had to travel much farther than usual to find enough
quirquinchos, and they often take Cha. villosus to compensate (Peredo 1999). Apart from the intensity of human predation, Cha. nationi also faces the loss of its
habitat in the Andean puna. Because of its preference
for sandy, high-altitude soils, Cha. nationi suffers from
C.
Figure 21.4. Current distributions of armadillos in the genus Chaetophractus. A. Cha. vellerosus. B. Cha. nationi. C. Cha. villosus.
224
the industrial excavation of sand for use in road-building and general construction (Peredo 1999). This combination of hazards has caused a stark decline in recent
yearseasily exceeding 30% in the past decade. The
species occurs in one park in Bolivia, the Parque Nacional Sajama, but none in Argentina, and its situation
is considered serious (Abba and Cassini this volume).
Ergueta S. and Morales (1996) emphasize the intensity
of hunting for handicrafts and charms, and note that
it is rare even within its preferred habitat. Noss et al.
(this volume) describe it as severely endangered and
point out that Bolivia must now import other species of
Chaetophractus from Argentina to bolster the trade in
handicrafts.
By contrast, the third member of the genus, Cha.
villosus (figure 21.4C), is coping well with heavy subsistence hunting in the Chaco and elsewhere in Argentina,
as well as a commercial meat trade in Bolivia (Ergueta
S. and Morales 1996). Although pursued by subsistence
hunters and attacked by dogsand often shot for sport
by wealthier ArgentinesCha. villosus remains abundant throughout its generous range.
Tough and resilient, Euphractus sexcinctus (figure
21.5) is also plentiful and broadly distributed. Yellow
armadillos are frequently hunted, especially in the Cerrado of Brazil, but their meat has a reputation for its terrible taste, and while subsistence hunting is a constant
pressure (Ergueta S. and Morales 1996) some hunters
in the Bolivian and Paraguayan Chaco will avoid them
(Meritt this volume; Noss et al. this volume). When it
is hunted in the Bolivian Chaco, its fat is often used as
a multipurpose folk remedy (Cullar and Noss 2003),
although some indigenous groups avoid it for any purpose because they believe that Euphractus feeds on
buried human corpses (Meritt this volume). The yellow
armadillo thrives in disturbed landscapes, readily scavenging crop leavings, insects and roadside carrion, all
of which are plentiful in areas of expanding agriculture
(Dalponte and Tavares-Filho 2004). This adaptability,
together with its ample range and mixed culinary reputation, ensures it is one of the most stable of armadillo
species.
Likewise tolerant of landscape disturbance, Tolypeutes matacus (figure 21.6A) does well in the presence
of limited agriculture, although it requires a good balance of native habitat as well. Despite this flexibility, T.
matacus is still suffering from the overall effects of habitat lossand because it reproduces slowly and does not
take cover in burrows, it is far more susceptible to hunting (Noss et al. this volume), although it can be a swift
Figure 21.5. Current distribution of the yellow, or sixbanded, armadillo, Euphractus sexcinctus.
and cunning target when in flight (Meritt this volume).

In the Bolivian Chaco, the three-banded armadillo is
savored above all other prey (Cullar and Noss 2003)
and although its decline is not yet critical, the hunting
pressure is almost certainly unsustainable.
Brazils only endemic armadillo, Tolypeutes tricinctus (figure 21.6B), is also heavily huntedto the extent
that its population may have fallen by more than 30% in
the past decade. Now vanished from much of its former
range, the three-banded armadillo had gone unseen for
decades, and some researchers speculated that it might
have gone extinct (Cole et al. 1994). Several sightings in
the early to mid-1990s, however, confirmed that the species was still aliveand in fact scattered across a much
wider region than first assumed, covering the Caatinga
and the central Cerrado of northeastern Brazil (Santos
et al. 1994), although scarce and fragmented throughout
(Machado et al. 1998). But these regions are among the
most impoverished in the Western Hemisphere, with
rampant malnutrition, illiteracy, and unemployment in
human communitiesthe worst possible environment
for a slow-breeding, surface-dwelling armadillo with
the defensive habit of rolling into an easily portable ball.
Many of the sertanejos, the rural people of the Caatinga,
rely on wild game for much of their protein as well as
a source of income, and so the three-banded armadillo
remains at exceptional risk throughout its range.
Subsistence hunting also poses a threat to all four species of Cabassous, the naked-tailed armadillos, together
225
Figure 21.6. Current distributions of armadillos in

the genus Tolypeutes.
A. T. matacus.
B. T. tricinctus.
A.
B.
with the omnipresent issue of habitat degradation; but

each species is affected differently. Ca. tatouay and Ca.
unicinctus, which have the broadest distributions in the
genus (figures 21.7A and 21.7B, respectively), are common throughout Brazil, the Guyanas, and the entire
Amazon basin. Of the two, Ca. unicinctus is more likely
to be vulnerable; it is easily captured and its meat is
popular, and in the state of Minas Gerais in east-central
Brazil, hunting has become a major threat (Machado et
al. 1998). Farther south in Paran, Ca. tatouay is hunted
as a nuisance, on account of the damage it does to fields
while excavating its many burrows (Mikich and Bernils
2004). Although both species are generally intolerant of
degraded and converted landscapes, they are present in
a great many protected areas, and the outlook for both
species is considered robust overall.
Ca. centralis, by contrast, occupies a much smaller
range in Central America (figure 21.7C), where it often prefers dry forests but has proven able to tolerate
secondary forests, as well as a mix of forest and agricultural land. One of the most fossorial of the armadillos,
Ca. centralis is difficult to study, but what is known suggests a more selective and insectivorous diet (Redford
1985a)which may account for its ability to survive
near agricultural disturbance. Much like Tolypeutes,
it is reported to roll itself into an armored ball when
threatened (Genoways and Timm 2003), but Ca. centralis may not be hunted as severely as its congeners
elsewhere because of its heavy, unpleasant musk (Reid
1997). It is avoided in Honduras, for instance, on ac-
count of local folklore which holds that its disagreeable

odor is poisonous, owing to its fabled habit of feeding
on human corpses (Marineros and Gallegos 1998). So
little is known about the species that no detailed assessment may be made; but much of its former habitat is
long since splintered and worn away, and populations
may be in a more fragile state than we now recognize.
The fourth and least-known member of the genus,
Ca. chacoensis, occupies a compact range in western
Paraguay and northern Argentina (figure 21.7D), where
it specializes in the dry shrublands known as chaco seco
(D. Meritt personal communication). Although subject
to attacks by dogs and low-level subsistence hunting,
the loss of its habitat to agriculture is the most pressing concern. The conditions it favors are miserable for
most humans to endure, and long-term field research
is extremely difficult in its remote habitat. Few specialists have ever seen it, and it appears to be exceptionally
rare (Meritt this volume). Although this species may
not be threatened at present, it is steadily losing habitat,
a trend that shows no hint of abating.
In a parallel situation and environment are the fairy
armadillos, Chlamyphorus retusus and Chl. truncatustwo of the most specialized of the armadillo clan,
adapted to swim beneath sand and loose soil in the
Chaco and central Argentina. Although morphologically similar, these two species may have converged on
their extreme fossorial habit from unrelated forebears,
and some researchers now consider Chl. retusus as the
distinct genus Calyptophractus (Gardner 2005). Dimin-
226
A.
C.
B.
D.
utive and nocturnal, living almost completely subterranean lives, the fairy armadillos are nearly impossible
to study in the field, and there are no firm data on any
aspect of their ecology, behavior, or population structure. Naturally rare and confined to specific subhabitats, neither species is likely to have a population large
enough to absorb the damages they now endure.
Chl. truncatus, the pink fairy armadillo, is endemic
to sandy plains and xeric grasslands in central Argentina (figure 21.8A), where they are amiably tolerated
by the local people and overlooked by their dogsalthough they fall occasional victim to domestic cats.
Figure 21.7. Current distributions of armadillos

in the genus Cabassous.
A. Ca. tatouay.
B. Ca. unicinctus.
C. Ca. centralis.
D. Ca. chacoensis.
In local memory they are thought to have been much

more common several decades ago, and a recent intensive search yielded no trace of live individuals (Superina 2006). The prime threat to Chl. truncatus today is
the wholesale destruction of native grasslands in favor
of pasture and farmland. Together with these intrusionswhich severely degrade whatever land is not
convertedcome the more insidious effects of agricultural chemicals, crippling the soil and its vital microfauna, and affecting all those dependent on them.
Although pink fairy armadillos are known from several
parks in Argentina, and there is provincial legislation in
227
Figure 21.8. Current distributions of armadillos in

the genus Chlamyphorus.
A. Chl. truncatus.
B. Chl. retusus.
A.
B.
place to protect them (Provincia de Mendoza 1998), the

constant degradation of habitat is a threat they cannot
evade.
A far more precarious situation holds for Chl. retusus,
the greater fairy armadillo, which survives in a compact
area centered on the Chaco of Bolivia, Paraguay, and
northernmost Argentina (figure 21.8B). A specialist in
sandy soils, through which it burrows like a mole with
its outsized foreclaws, the greater fairy armadillo is unable to prosper in the dense clay substrate that covers
most of the Chaco and is thus only thinly distributed
across its modest range. Although it is known from
several protected areas, including the Defensores del
Chaco National Park of northwestern Paraguay (Meritt
this volume), Chl. retusus suffers constant losses from
merciless persecution by local people. They do not kill
it for food, or as a garden pest, despite hunting several
other armadillo species; rather they kill it from their
engrained conviction that to see one, even by chance,
is an evil omen that foretells the death of a close family memberunless the unfortunate messenger can be
slain on the spot, averting the sudden calamity.
This anathema runs far deeper than anything directed against giant anteaters, which are often viewed
as dangerous, but not as objects of deadly fear. The full
effects of this prejudice are unknown, at least in the
strictest sense, as there are no data on the demography
of the greater fairy armadillobut clearly the relentless
destruction of a rare, haphazardly distributed species is
a critical concern. Its best defenses are its natural rar-
ity and its nocturnal, subterranean lifestyle, which hide

it from potential persecutors. A campaign of environmental education would require a tremendous effort
against the cultural resistance it would face; but without
a long-term change in the beliefs of the local peoples,
the fairy armadillo will remain in danger for as long as
humans are near.
Also under constant threat from human pressure,
though to a lesser degree, is the Patagonian pichi, Zaedyus pichiy. With the most southerly range of any
xenarthran (figure 21.9), Z. pichiy endures intensive
hunting both for subsistence and for entertainment, depending on the local culture of its hunters. In addition
to this heavy poaching, pichis are also threatened by
the conversion and degradation of their habitat, even
in the arid regions they prefer (Superina this volume).
Naturally uncommon throughout their range, pichis
are ill-suited to withstand long-term hunting pressure,
and their populations are already in decline.
Human impact is much less pronounced on the most
widespread of armadillo genera: Dasypus, the most
speciose and, thanks to one species in particular, also
the most well known. That species, D. novemcinctus, is
a sturdy invader which has expanded from the Neotropics to penetrate a broad wedge of North America
as well (Taulman and Robbins 1996; Merriam 2002).
Ranging from Uruguay and Argentina to the heartlands
of central Kansas (figure 21.10A,B), D. novemcinctus is
easily the most populous of any xenarthran species, and
although it is frequently huntedand increasingly poi-
228
Figure 21.9. Current distribution of the pichi, Zaedyus pichiy.
soned as a nuisance in North Americaits survival is

by no means in doubt.
Several other species of Dasypus are also widely
distributed, to a greater or lesser extent, but so little is
known about their ecology and behavior (McDonough
and Loughry this volume) that any assessment of their
status is preliminary at best. These include D. septemcinctus, a widespread savanna-dweller (figure 21.10C);
D. sabanicola, found in savannas and gallery forest in
northern Colombia and Venezuela (figure 21.10D);
and D. kappleri, known primarily from forest patches
in savannas of the Amazon and Orinoco river basins
(figure 21.10E, see also Wetzel 1985a,b). Each of these
species is hunted to some degree; Pacheco and Naranjo (1978), for instance, commented that D. sabanicola
is highly prized as food in Venezuela, and Linares
(1998) confirmed that it is intensively hunted by rural
people. D. sabanicola and D. kappleri are considered to
be under Proteccin Especial in Venezuela, while the
extreme scarcity of the latter in Ecuador may be due
in part to subsistence hunting (Tirira S. 2001). For the
present, however, these three species are considered
to be of minimal conservation concern, with populations in many national parks throughout their respective ranges. D. septemcinctus in particular is remarkably
widespread, and appears able to adapt to secondary
habitat and a certain degree of human disturbance.
Much less flexible is D. hybridus, which appears

sensitive to even slight changes in the landscape due to
agriculture (Abba and Cassini this volume). Primarily
a grassland species (figure 21.10F), it has disappeared
from many areas of its former range, owing to heavy
land use and severe hunting pressure. Much more
widespread decades ago, the species now shows a rapid
decline, and is considered to be on the threshold of
threatened status. Also of some concern, though without numerical evidence of a decline, is the hairy longnosed armadillo, D. pilosus. Once thought to occupy
several regions in the Andes (Wetzel 1985b), it is now
known to occur only in a single narrow zone of northcentral Peru (figure 21.10G). It has been reported from
the Peruvian departments of Amazonas, Hunaco, and
San Martn (A. Abba personal communication), but the
only recent sightings have been from the Ro Abiseo
National Park in San Martn, at the northern extreme
of its estimated range (J. Loughry personal communication). Too little is known about this species for any
firm conclusions, but its tightly restricted range is cause
enough for apprehension.
The least known of all Dasypus is also the most
recently recognized: D. yepesi, described in 1995 and
known only from the provinces of Salta and Jujuy in
northwest Argentina (Vizcano 1995; Vizcano and
Giallombardo 2001). Apparently tolerant of a variety of
environmental conditions, from xeric lowlands to humid montane forest, it has the smallest known distribution of any armadillo, and as such deserves particular
attention.
Prospects
Each of the species detailed above exists in a unique
situation, shaped by its ancestry and present environment; each has its own dangers to face and challenges
to survive. But all of them are imperiled, to a greater or
lesser degree, by the same great forces that have swept
across the continent in the past century. These fundamental alterations, forced upon virtually every Neotropical biome, are driven solely by the presence and
hungers of a single populous and meddlesome species.
This rogue primate is a relatively recent arrival, but it
has made its presence felt throughout the hemisphere,
and may have been directly responsible for the extinction of the larger xenarthrans during the Pleistocene.
Its effects on the modern Xenarthra are essentially twofold: first, in its wholesale transformation of landscapes,
which it swiftly reshapes for its own benefit; and sec-
A.
B.
C.
D.
E.
F.
Figure 21.10. Current distributions of

armadillos in the genus Dasypus. Note that,
because of its extremely limited range (see
text), the distribution for D. yepesi is not
shown.
A. D. novemcinctus, detail of South American distribution.
B. D. novemcinctus, detail of Central and
North American distribution.
C. D. septemcinctus.
D. D. sabanicola.
E. D. kappleri.
F. D. hybridus.
G. D. pilosus.
G.
230
ond, its fierce appetite for wild mammals in general,

and many xenarthrans in particular.
And here the xenarthrans deep history of dogged
survival finally fails them, through no fault of their
own; for they have evolved to fend off or evade those
predators which operate singly and at close range. The
Xenarthra have no innate defense against a cooperative society of hunters who excel at killing from afar,
who collectively re-engineer entire biomes to suit their
own desires. The Xenarthra are thus caught between
these two aspects of this most effective predator: those
which survive the burning of grasslands and stripping
of forests are all the more exposed to roaming hunters,
while those who flee the hunters have ever fewer scraps
of habitat in which to find refuge. And the survivors
cannot quickly replenish themselves; from Priodontes
and Myrmecophaga to Bradypus and Chlamyphorus,
the Xenarthra are slow breeders, and only those species
with the lucky knack of polyembryonyin particular D. novemcinctusare capable of reproducing fast
enough to answer human demand.
For almost all the Xenarthrawhatever their situation as we currently understand itthere is an inevitable slippage toward the day when individuals are too
few, and too lost upon the land, to find one another
and perpetuate themselves. On the day when the final
giant anteater is shot for its claws in Venezuela, or out
of baseless fear in Rio Grande do Sul, the gene-lines
in those areas will be forever gonethe whole of the
ancient tapestry unraveling with every pinpoint death.
The calculus of species extinction admits to no final reprieve; this rogue primate with its clever fingers and
endless hunger is famous for ex post facto remorse,
when all chance of changing course is safely past and
gone.
So if the mathematics are to be altered for a nonzero solution, they must be adjusted now, while most
populations are still viable and large reaches of varied
habitat, most notably in the Amazon and the Chaco,
remain substantially intact. The economies of a dozen
tropical nations are geared toward intensive resource
exploitation; and those few of us who see innate worth
in living xenarthransnot for their genes or any practical use, but for themselves alonewatch the waves of
humankind churning across the land, and are tempted
to despair.
But despite the overwhelming pressures across the
continent, there is still much that can be done. The
first stages are already well underwayfield research
is in progress on many species, difficult as it often is.
Although data from more species are clearly required

(Delsuc and Douzery this volume; Prodhl et al. this
volume), genetic data will allow us to identify important populations for future conservation measures, as
exemplified by current efforts to describe phylogeographic patterns in sloths and silky anteaters.
And yet this by itselfthe natural progression of
our current, incremental approachwill not be sufficient to rewrite the equations of loss. Raw deforestation
is too vast and impersonal, perhaps, for us to directly
confront; but we do have an opportunity to transform
one of the great weaknesses of the Xenarthra (from a
human perspective) into a compelling strength.
In conservation terms, the Xenarthra have been effectively invisiblethey are too few species for grandiose numerical claims; and the clade as a whole is, at
least for now, much less threatened than primates or
cetaceans, parrots or panda bears. Also, in contrast to
these other groups, the reputation of the Xenarthra
ranges from poor to darkly superstitious: denounced
as corpse-eaters and demon messengers, shot on sight,
and left to rot, or else out of mind completely on account of simple ignorance.
But here lies a great opportunity, because the Xenarthra have no want of charisma waiting to be tapped.
If naked mole-ratswriggling, wrinkled, buck-toothed
creatures, compelling in their own faintly hideous
waycan capture public attention, how much more the
mellow and trusting gaze of an orphaned sloth? Giant
anteaters, for their part, are so breezily comical, and yet
so carefully dignified in their gait, that they are instantly
charismatic without the slightest effort. And the silky
anteater in particularthere is no staid, professional
way to say itis irresistibly cute, golden-glowing and
all a-fluff with sleepy charm. For all our predatory zeal,
we still have a deep affection for the engagingly outr
of the animal world, and we should let this work to the
benefit of the Xenarthra. Silky anteater, pygmy sloth,
pink fairy armadillo: their names alone hint at fascination, and their stories would bring a wider audience to
recognition of the Xenarthra as a whole.
Public awareness by itself, however, is only one aspect
of a proactive approach to xenarthran conservation.
One of the most critical components is field education,
directed toward those human populations who have the
most impact on sympatric xenarthrans. This will not be
easy: campesinos, caboclos, sertanejos and pobladores
they all have their folklore and deep beliefssome of
them rational and rooted in careful observation, and
some of them a good deal less so. Changing their more
destructive convictions, or even only deflecting them,

will often be a difficult challenge: convincing chaqueos
that fairy armadillos are not messengers of impending
death, or gauchos that giant anteaters are not outsized,
menacing verminor sertanejos that the most delicious and easily hunted of all tatus should, in fact, be
carefully spared. But in countries where park rangers
snare protected wildlife for their own families food, all
this and more will be vital to prevent many species from
being hunted into oblivion.
Field researchers must play a pivotal role in altering
public perceptions of the Xenarthra. They are at once
the best informed of the dangers xenarthrans face, the
most motivated to care about what becomes of them,
and in the best position to share this understanding
with local residents. We are well aware that in some
cases this may be next to impossible, and too vigorous
an advocacy of too alien a perspective may endanger
the research itself. But national laws and provincial decrees will be ineffective, especially in remote and thinly
monitored regions, if there is not at least a seed of local
belief to support them.
In addition to assessing the status of xenarthran species, the IUCN/SSC Edentate Specialist Group works
actively to promote their conservation. The ESG supports efforts by researchers to study and conserve xenarthran populations wherever they survive, and publishes
articles and field sightings in the journal Edentata. In
support of the research community, Edentata is freely
distributed among all those interested in the preservation of the Xenarthra; there are no page charges to
contributing authors and no subscription fees, and all
back issues are available for download at edentata.org.
Our aim is to provide a forum for everyone involved in
xenarthran conservationstudents, researchers, zookeepers, rehabilitators, ecotourists, and moreand to
promote as broad an interchange of ideas as possible,
creating a network among those who might otherwise
never interact.
The ESG also directly finances small, focused research
projects with an emphasis on field ecology and conservation, through the Edentate Conservation Fund. This
small-grants program is designed to stimulate research
on xenarthrans in both native and transformed habitat,
with a particular emphasis on species whose status is
threatened or endangeredor completely unknown.
Most grants are for approximately US$1,000, and are
directed toward students and researchers in habitat
countries. At the time of this writing, the Edentate Conservation Fund has supported more than half a dozen
231
projects, and welcomes applications on a continuing

basis. We encourage students in particular to develop
projects that will contribute to the understanding and
conservation of xenarthrans in their home countries.
These are small steps, but vital ones, and they will be
built on by the current and future community of xenarthran researchers. We cannot recommend specific
guidelines or legislation here; the situation for every
species varies from region to region, and the primary
function of the Edentate Specialist Group is to promote
research and the exchange of conservation data. At
present, the prognosis for the Xenarthra is far from uniformly grim; some species are still safe and numerous,
and others have proven to be tough and adaptable. But
too many species have lost too much of their former
extent, and the circumstances that threaten them are
no less dangerous to the rest. We who recognize these
many threatsand the inherent value of the full Xenarthran clanshould act with all our strength and vigor
today, while the prospects are favorable that the group
as it now stands may yet survive in full.
Acknowledgments
Our thanks to the participants in the 2004 Edentate
Species Assessment Workshop, without whom much
of this information would not have been readily available. The participants included Agustn Abba, Teresa Anacleto, Adriano Chiarello, Erika Cullar, Jim
Loughry (in absentia), Dennis Meritt, Flvia Miranda,
Gustavo Porini, Paula Lara-Ruiz, Rafael Samudio Jr.,
Mariella Superina, and Sergio Vizcano. The workshop
was chaired by Gustavo Fonseca and organized by John
Aguiar, Adriano Chiarello, Adriano Paglia, and Anthony Rylands, with support from Wes Sechrest and
the Global Mammal Assessment.
We would also like to thank Jim Loughry and Sergio Vizcano in particular for the invitation to contribute to this volume, and we extend our appreciation to
Reed Noss and Wes Sechrest for their careful reviews
of a preliminary draft. Thomas Lacher Jr. and Mariella
Superina both made invaluable comments on the concluding paragraphs. Kim Meek, senior graphic designer
for CABS at Conservation International, prepared the
range map illustrations, which were first published in
Edentata 6 (2004). J. M. A. extends a special thanks to
Mariella Superina for her Spanish translation of this
chapters abstract and for many excellent conversations
about the xenarthran world.
22
Maintenance of Xenarthra in captivity
Mariella Superina, Flvia Miranda, and Tinka Plese
tive reproduction because wild animals were deemed

abundant. Todays zoological gardens are aware of the
fact that they can no longer afford to be consumers of
wildlife. On the contrary, they have recognized their
responsibility and the important role they play in the
conservation of wildlife.
The history of captive maintenance of Xenarthra
outside their area of distribution dates back to the
second half of the nineteenth century. The first live giant anteater (Myrmecophaga tridactyla) was brought
to Hamburg by Carl Hagenbeck in 1864 (Bartmann
1983), while the first lesser anteater (Tamandua tetradactyla) was exhibited at London Zoo in 1871 (Encke
1978). Numerous Xenarthra have been kept in different
parts of the world since then and have awoken interest
and curiosity in generations of visitors and researchers. Important progress has been made recently in
understanding the specific needs of several Xenarthra
species. Population management plans that coordinate the breeding strategies of different institutions in
North America have been initiated for M. tridactyla
and Tolypeutes matacus and are planned for Choloepus
spp. (J. Gramieri personal communication). European
breeding programs exist for Choloepus didactylus, C.
hoffmanni, M. tridactyla and T. tetradactyla, but not for
armadillos (European Association of Zoos and Aquaria
[EAZA] 2005).
Although captive conditions for Xenarthra have improved considerably, many questions regarding appropriate husbandry still remain unanswered and breeding success in captivity is low for many species. It is
important to realize that behavior as well as dietary and
environmental needs vary considerably across species
within a taxon. Husbandry protocols that have been
elaborated for one species are therefore of only limited
Resumen
Los xenartros se han mantenido en cautiverio por ms
de un siglo y han despertado el inters y la curiosidad de
generaciones de visitantes. En los ltimos aos se han
logrado importantes avances respecto al conocimiento
de los requerimientos especficos de varias especies. A
pesar de ello, la reproduccin exitosa todava es poco
comn en la mayora de las especies. En este captulo
se exploran las dificultades ms comunes que estn
asociadas con el mantenimiento y la reproduccin en
cautiverio de armadillos, osos hormigueros y perezosos y se provee una breve sntesis de las condiciones
apropiadas.
Resumo
Os xenarthras tm sido mantidos em cativeiro por mais
de um sculo e tm despertado interesse e curiosidade
em geraes de visitantes. Nos ltimos anos tm-se
conquistado bastante avano a respeito dos conhecimentos e requerimentos especficos de varias espcies.
No obstante, o sucesso reprodutivo pouco comum
na maioria das espcies. Este captulo revisa as dificuldades comumente encontradas que esto associadas ao
manejo e reproduo em cativeiro de tatus, tamandus
e preguias e d uma breve sntese das condies apropriadas.
Introduction
In the early days of zoological gardens, the principal
goal of menageries was to acquire and exhibit to the
public as many exotic animal species as possible. Little
regard was paid to long-term survivorship and cap232
233
Figure 22.1. Captive

Priodontes maximus
feeding on a termite
mound. This species
is maintained in only
a few zoos within its
natural area of distribution. Photograph by
M. Superina.
use when designing enclosures or developing diets for

other, even closely related species. A description of the
specific needs of each species would go beyond the
scope of this book. This chapter therefore provides just
a brief overview of appropriate husbandry conditions.
Given that important interspecific differences can occur, we have tried to avoid making sweeping generalizations about captive management. Nonetheless, given
the limitations of space, some generalizations were unavoidable.
Armadillos
Species kept in captive conditions
Most armadillo species have been or are currently represented in zoological collections. While Chaetophractus villosus is the most common armadillo species in
European zoos, Dasypus novemcinctus and Tolypeutes
matacus are kept mainly in North America (International Species Information System [ISIS] 2005). A few
species are rarely kept. For example, Chlamyphorus
truncatus is difficult to maintain in captivity, as it seems
to be very sensitive to changes in environmental temperature and humidity (Chebez 1994; Bertonatti and
Aprile 1999). Additionally, the design of an enclosure
that is both appropriate for this fossorial species and
the visitors wishing to observe it is especially challenging. Thus, few attempts to exhibit pink fairy armadillos
have been made in the past.
The case of the giant armadillo, Priodontes maximus,
illustrates many of the problems associated with main-
taining armadillos in captivity. Unlike the fairy armadillos, countless attempts to keep and breed P. maximus
were carried out in many parts of the world until its
inclusion in CITES Appendix 1 in 1973 (e.g., Cully 1939;
Ostenrath 1974). No reproductive success was ever reported (Chebez 1994). In the past decade, giant armadillos have been maintained in only a few zoos within
their natural area of distribution (figure 22.1). Most of
these individuals share a similar past: they were usually captured in the wild by locals and kept in unsuitable conditions and on inappropriate diets before they
were seized by law enforcement agencies and passed on
to certified zoos. The problems faced in such cases are
the lack of appropriate enclosures, as the zoos had no
plans to keep giant armadillos, and the zoo staff s lack
of experience in the maintenance of P. maximus (Superina 2000). There is almost no published information
on the maintenance of giant armadillos, or their basic
ecology, that caretakers can refer to in attempting to
improve existing facilities or work out a suitable diet. If
the animals do not die of injuries or malnutrition, they
are usually set free by the zoos after a short recovery period. More information on the adequate maintenance
of giant armadillos is needed to increase the survival
rate of these often debilitated animals. Such data can
be obtained only through increased efforts to keep and
study Priodontes in captivity, and intensified fieldwork
to obtain basic knowledge on the ecology of this poorly
known species.
Similar considerations apply to many other armadillo species, making captive maintenance a challenging problem. The following sections review some basic
issues.
234
M. Superina, F. Miranda, and T. Plese
Enclosures
Lacking information on the spatial needs and territoriality of most armadillo species makes it difficult to
determine adequate enclosure size and group composition. In many zoos, armadillos seem to be kept in any
unoccupied space, instead of being housed in a specifically planned enclosure that is appropriate to the species (Superina 2000). Small enclosures and resulting
lack of hiding places can lead to stress and aggression
between conspecifics (Roberts et al. 1982). A survey
carried out in 1998 showed that individually kept armadillos were maintained in enclosures averaging 5.7 m2
(range: 0.418 m2). The average area available to each
animal when kept with conspecifics was only slightly
larger (Superina 2000). The American Zoo and Aquarium Association recommends an area of 0.7 m2 per kg
of body weight per armadillo (Flint 1997). Ratajszczak
and Trzesowska (1997) proposed keeping pairs or small
groups of Cha. villosus in a space of at least 10 m2, while
four animals of the same species shared an area of 49.5
m2 at the National Zoo in Washington, D.C. (Roberts
et al. 1982).
Outdoor facilities should have wall footers that extend underground for more than 1 m and into the exhibit at least 60 cm to keep the armadillos from escaping through digging (Flint 1997). Water ditches are no
obstacle to most armadillo species, with the possible
exception of Cabassous and Tolypeutes that reportedly
do not swim (Krumbiegel 1940). Low walls and mesh
fences are easily climbed. However, armadillos are not
able to climb down mesh fences (Taber 1945), so they
can injure themselves or even die from falling (Merrett
1983). Intestinal torsions have been observed in animals
that repeatedly climbed up and fell down a wire mesh
fence (C. Leizagoyen personal communication). Both
concrete and wire mesh enclosure sides pose the risk of
abrasion and require regular clinical examination of the
armadillos to prevent injuries.
Unsealed wooden floors retain moisture and are difficult to clean and thus should be avoided (Wampler
1969). Concrete floors prevent armadillos from escaping but increase the risk of injuries, therefore they need
to be covered by soft material, such as soil, stringbark
mulch or wood shavings (Superina 2000, figure 22.2).
Sawdust and shredded paper should be avoided as covering materials, as intestinal obstructions have been
reported in armadillos that ingested them (Anderson
and Benirschke 1966). While Tolypeutes seems to prefer
fresh straw as bedding material, straw has been linked
with hyperactivity in some individuals, possibly because

of an allergic response (J. Gramieri personal communication). Eye irritations and drying out of the carapace
have been observed in Cabassous kept on fresh straw
(Meritt 1976b).
The analysis of clinical case histories and necropsy
reports of 352 captive armadillos showed that injuries,
especially on feet and tails, were by far the most common pathologies (107 cases, Superina 2000; see figure
22.2). In addition to the medical treatment of such lesions, it is essential to review and improve the conditions in which the animals are kept. Similarly, pneumonias are among the most frequent diseases observed in
captive armadillos (Superina 2000). Stress is an important factor that predisposes captive animals to infections of the respiratory tract. It can be caused by failure
to adapt to captive conditions, inappropriate handling,
or inadequate husbandry, that is, deficient ventilation
or suboptimal environmental temperatures (Resoagli et
al. 1985).
Nutrition
The last 15 years have brought significant advances in
armadillo nutrition. The analysis of armadillo diets has
led to the development of taxon-specific formulas, some
of which are now commercially produced in the United
States. Unfortunately, these studies are limited to the
most commonly kept species, commercial diets are not
Figure 22.2. Chaetophractus villosus kept in an inadequate

enclosure. Concrete floors prevent the armadillos from
burrowing and increase the risk of injuries. Tail lesions are
among the most frequent pathologies in captive armadillos
and can require amputation. Photograph by M. Superina.
235
acclimatization to the artificial diet seems to occur frequently (Superina 2000).

Reproduction
Figure 22.3. Chaetophractus vellerosus with vitamin A deficiency and pyodermia. Photograph by G. Prez Jimeno.
available in South America, and formulas often contain

ingredients that are not available or too expensive in
the armadillos natural area of distribution. As a consequence, South American zoos receiving confiscated
armadillos that are uncommon in North American institutions are obliged to feed them diets that are based
on experience and tradition rather than on scientific
data. The most common diet consists of a mixture of
beef or dog food, fruits, vegetables, eggs, and vitamin
and mineral supplements. More than 50% of all zoos
regularly add insects, especially mealworms, and 55%
supplement the diet with vitamin K (Superina 2000).
The diet of captive armadillos is usually uniform over
time, although field studies on wild individuals suggest
seasonal variation in food quantity and composition
(McDonough and Loughry this volume). Moreover,
most species are fed similar diets in spite of their different food habits in the wild.
A wide array of pathologies is associated with inappropriate nutrition. The natural diet of armadillos contains low levels of protein and fat (Coppo et al. 1979)
and high levels of ashes and raw fibers (Ramsey et al.
1981). The high prevalence of obesity, renal failure, and
diseases of the digestive tract in captive armadillos suggests that the composition of the artificial diet is often
not well balanced (Superina 2000). Liquid or finely
chopped food does not require the animals to chew
and can lead to mouth or tooth problems because the
continuously growing teeth are not worn down sufficiently. Hypovitaminoses and hypocalcaemia have been
observed in several zoos (figure 22.3). These problems
can be avoided by addition of appropriate vitamin or
mineral preparations, but they indicate that armadillos
may require higher doses than found in current artificial diets. The high prevalence of enteritis, 9.7% of all
necropsies and clinical case histories, most likely has its
origin in nutritional problems and stress. Insufficient
Reports on successful breeding are relatively scarce

although armadillos have been kept in captivity for
decades. According to the available literature, species
such as P. maximus, Chl. truncatus, and Chl. retusus
have never reproduced in captivity (Chebez 1994). D.
novemcinctus has been used in biomedical research
for more than 30 years (e.g., Storrs et al. 1974; Truman
this volume), but most attempts to breed this species
in laboratory conditions have failed. To date, no reliable breeding protocol has been developed, in spite of
intensive research on the reproductive cycle and the
causes of reproductive failure in captivity (e.g., Peppler
and Canale 1980; Peppler and Stone 1981; Rideout et al.
1985). Because of this, many laboratories do not even
try to breed this species. On the other hand, many zoos
do not consider captive propagation a priority because
of the abundance of this species in the wild.
A survey carried out in 1998 revealed that only 16 of
62 zoos had a breeding program for their armadillos.
The group composition of armadillos suggests that reproduction could have been possible in 33 of these zoos.
Nevertheless, only nine zoos reported breeding success
that year (Superina 2000). Recent data (ISIS 2005) suggest that breeding success has not increased since. Low
breeding success might be caused by deficiencies in the
husbandry of armadillos that impair their reproductive
ability (Superina 2000). It should be noted, however,
that some institutions opt not to breed their armadillos
because they keep them for educational programs only
and consider that pregnancy and rearing of offspring
would limit availability for educational work (J. Gramieri personal communication).
Armadillo dams seem to be highly sensitive to external disturbances. Hands-off management and a quiet
environment may ensure a higher survival rate of offspring. Male armadillos may react very aggressively
to the presence of newborns. In T. matacus, juvenile
mortality can be up to 100% in the presence of males
(J. Gramieri personal communication). Similar experiences have been reported from other armadillo species. It is therefore crucial to separate all males from
pregnant females prior to the estimated date of birth.
Enclosure size seems to be another critical factor for
successful rearing, but whether a large or small enclosure is more appropriate for lactating females may be
species specific. For example, nursing females of T.
236
matacus sometimes lost track of their infants in large

enclosures and breeding success was higher in small
enclosures (J. Gramieri personal communication). In
contrast, Ratajszczak and Trzesowska (1997) reported
that nursing Cha. villosus kept in small cages were
prone to destroying youngsters. Mothers may consider
the presence of humans more of a threat to their young
in small enclosures. In some cases, mothers will carry
their offspring around constantly, apparently attempting to hide them from potential predators, eventually
inflicting injuries to the young or even killing them.
Other stressed females continuously dig new burrows
and thus neglect their offspring. If the animals are not
kept in an outdoor enclosure with natural flooring that
allows them to dig a burrow, it is essential to provide
the females with a nesting box with hay into which they
can retreat (Superina 2000).
An analysis of the behavior of wild armadillos suggests some other factors that might be responsible for
the low reproductive success in captivity: cleaning of
enclosures may remove scent marks that delimit the
territory of male armadillos and indicate their reproductive status. The lack of such olfactory information
might inhibit males from breeding (McDonough 1997).
Additionally, it is common to keep breeding groups together all year long, even though wild armadillos are
primarily solitary (McDonough 1997). Housing males
and females together just for the breeding season might
improve reproductive success. It is interesting to note
that most litters of captive nine-banded armadillos
were born in open enclosures. The animals were kept
in large enclosures with a natural substrate in which
they could dig burrows and forage, and were exposed
to natural light cycles and seasonal changes in environmental temperature and precipitation (Job et al. 1984;
Truman and Sanchez 1993; Carvalho et al. 1997; Superina 2000).
Coordinated breeding programs for armadillos are
difficult to develop because of the lack of knowledge
about the reproductive cycle of most species. What little we know can be summarized as follows. No external
signs of estrus are mentioned in the literature; research
on ovulation has been carried out through progesterone analyses (e.g., Peppler and Stone 1976), laparotomy,
or necropsies (e.g., Enders and Buchanan 1959; Enders
1966; DAddamio et al. 1977). A preovulatory genital
hemorrhage that could be used to determine female
reproductive status seems to occur in Cha. villosus and
T. matacus, but not in D. novemcinctus. Contrary to
Enders and Buchanan (1959), DAddamio et al. (1977)
and Peppler and Stone (1980b) were able to determine

cytological changes in urogenital smears during the estrous cycle of nine-banded armadillos. However, this
could not be done in urogenital smears of Cha. villosus
(P. Carmanchahi personal communication).
Behavior
About 50% of the zoos surveyed in 1998 kept armadillos individually and almost 60% maintained them in
groups with several conspecifics. While most surveyed
institutions kept one male with one female, the largest group consisted of 10 animals (Superina 2000). It is
recommended to introduce conspecifics to each other
as juveniles to avoid aggression. Regardless, to prevent
injuries it is important to habituate conspecifics slowly
and monitor them closely. Sufficient hiding places
should be provided to allow subordinate individuals to
retreat from dominants (Roberts et al. 1982; S. McPhee
personal communication). Sick or injured armadillos
need to be separated, as open wounds and fresh blood
are licked by conspecifics and can lead to cannibalism
(Superina 2000).
The first report on stereotypic behavior was in a captive giant armadillo (Cully 1939). In 1998, behavioral
disorders were reported from 50% of surveyed zoos.
The most common behavioral disorders are excessive
digging or scratching on the walls, walking in circles,
pacing in corners, cannibalism, and automutilations
(Superina 2000). Some captive armadillos are hyperactive, while others become apathetic. All eight surveyed
institutions keeping T. matacus reported abnormal behavior in their animals. Similarly, 12/28 zoos keeping
D. novemcinctus and 6/10 zoos keeping Cha. villosus
reported behavioral disorders (Superina 2000). These
observations suggest that there are still deficits in the
maintenance of armadillos. Factors leading to behavioral disorders may include: lack of space and inappropriate surfaces that restrict the ability to roam and dig,
limited stimuli in unchanging environments or, alternatively, an excess of stimuli, diets that are not appropriate to the species, or a lack of consideration of their
natural social structure (Superina 2000).
Anteaters
The great attractiveness of anteaters has stimulated
zoos to intensify their efforts to maintain Vermilingua
in enclosures that imitate their natural habitat and to
breed them in captivity. Many aspects of anteater ecology and maintenance remain to be studied (Rodrigues
237
Figure 22.4. An appropriate enclosure for Tamandua should contain abundant branches to climb.
Photograph by F. Miranda.
et al. this volume). However, a recent increase in scientific research on wild and captive individuals has led
to significant improvements in many aspects of their
husbandry.
Enclosures
The spatial needs of anteaters are considerably larger
than those of armadillos. Montgomery (1985b) estimated that wild giant anteaters (Myrmecophaga tridactyla) have home ranges of 2,500 ha and walk approximately 11 km per day while foraging. There seems to be a
large difference between the home ranges of Tamandua
mexicana and T. tetradactyla, the former covering 25 ha
and the latter 375 ha (Montgomery 1985b). These spatial requirements need to be taken into account when
designing enclosures for anteaters. Flint (1997) suggests
keeping single giant anteaters in a space of at least 28 m2
and increasing it by 50% for each additional animal. The
Brazilian Institute of Environment and Natural Renewable Resources (Instituto Brasileiro do Meio Ambiente
e dos Recursos Naturais Renovveis, IBAMA) recommends an open enclosure of 80 m2, containing soil and
arboreal vegetation, for a maximum of two giant anteaters (IBAMA 2005). Soil flooring is preferred to allow
the animals to wallow in mud. The animals should have
access to a pool of water and a shelter of approximately
2 m2. In areas with low environmental temperatures, it

is essential to provide the animals with heated shelters.
Giant anteaters can swim, so water is not an effective
barrier (Flint 1997). Smooth walls at least 1.5 m high
are appropriate boundaries. As this species can climb,
mesh fences should be avoided and the trunks of trees
inside their enclosures covered with metal sheets.
Wild giant anteaters are solitary animals except during the mating season (Bartmann 1983) and may interact aggressively when kept in groups. No more than
two giant anteaters share an enclosure in most zoos, but
occasionally more individuals have been kept together
(e.g., Bartmann 1983). Enclosures of at least 15 m2 allow
subordinate animals to retreat from dominant anteaters.
Conspecifics should be habituated to each other over
an extended period. Additional space to separate males
from pregnant females should be planned, as males often react aggressively to the presence of neonates and
could kill them. Some males, however, seem to tolerate
offspring until they are fully weaned (J. Gramieri personal communication).
Lesser anteaters (Tamandua spp.) are predominantly
arboreal but also move, feed, and rest on the ground
(Montgomery 1985b). An appropriate enclosure should
therefore provide both soil to walk on and branches to
climb (figure 22.4). Enclosures should have an area of
238
15 m2, be at least 3 m high, and contain one elevated

shelter and one feeding station per individual (IBAMA
2005). Smooth walls or glass will prevent lesser anteaters from escaping. As with giant anteaters, wire fences
and water barriers are ineffective (Flint 1997).
Cyclopes didactylus has strictly arboreal, nocturnal
habits and a highly selective diet (Montgomery 1985b).
The difficulties in keeping this species in captivity are
reflected in the low number of reports describing successful maintenance and the lack of recommendations
for appropriate husbandry.
Injuries are relatively common in captive anteaters
and represented 15.5% of the pathologies observed at
So Paulo Zoo over a period of 23 years (Diniz et al.
1995). Aggression between conspecifics and capture
and restraint are common causes of injuries. Wounds
caused by loose wires, rough walls or floors, or inappropriate equipment can be avoided through regular
inspection of the enclosure. Finally, claw injuries are
frequent in anteaters kept on hard floors and can lead
to the loss of the affected claw as well as serious inflammations.
Nutrition
Wild anteaters feed mainly on different species and developmental stages of ants and termites (Montgomery
1985b). For obvious reasons, it is not possible to provide
captive animals with a sufficient amount of these insects, and composing a diet having similar nutritional
properties is not easy. Determination of appropriate
vitamin doses for captive anteaters seems to be especially challenging. The high prevalence of spontaneous
hemorrhages observed in the past decades could be
reduced through regular supplementation of the diet
with vitamin K or probiotics. The latter stimulate the
growth of bacterial flora in the small intestine, which
in turn synthesize vitamin K. Vertebral hyperostosis,
caused by ingestion of excessive amounts of vitamin A
or D (Crawshaw and Oyarzun 1996), has been linked
to diets based on commercial cat food. Because of this,
several zoos have begun using moistened dog food as
part of their anteater diet because it contains less of
these vitamins. However, this diet may have problems
of its own. Recently, several giant anteaters have died of
heart problems associated with severe cardiomyopathy
(Wilson et al. 2003; figure 22.5). A similar syndrome
in domestic felines and canines is associated with taurine deficiency, and commercial dog foods seem to
have considerably lower taurine levels than cat food.
Figure 22.5. Thoracic radiographs of a healthy giant anteater

(above) and a giant anteater with cardiomyopathy (below).
The heart of a healthy anteater (center of radiograph) spans
fewer than 3 sternebrae, while the enlarged heart of the
taurine-deficient animal spans more than 5 sternebrae and
the trachea appears dorsally displaced by the enlarged
heart. Photographs by R. Aguilar.
These findings suggest that dilated cardiomyopathy

in giant anteaters is associated with a nutritional deficiency (Aguilar et al. 2002). Supplementation of anteater diets with taurine, pinky mice, meat, and insects
is recommended to increase taurine levels and prevent
cardiomyopathies. At present, the most recommended
anteater diet consists of a commercially available insectivore diet supplemented with yogurt, fruits, and
vegetables. Pinky mice, rotten logs, and insects can be
added for behavioral enrichment (R. Aguilar personal
communication). Because the insectivore formula is
not available in South America, institutions within the
natural area of distribution of Vermilingua usually feed
their animals a diet of milk, eggs, ground meat, yogurt,
commercial dog or cat food, and vitamin and mineral
supplements (Trusk et al. 1992).
As anteaters lack teeth and therefore cannot chew
their food, the diet needs to have a pasty or granular
consistency. Food stations should be easy to clean and

have a heavy concrete base to prevent the animals from
seizing them with their strong claws and turning them
over. In accordance with their arboreal habits, lesser
anteaters should be fed on elevated platforms, for example, feeding stations attached to trees. It is important
to institute any changes to diet composition or feeding
time gradually, especially in wild-caught animals.
Twenty-three percent of deaths recorded in vermilinguans at So Paulo Zoo over a period of 40 years were
of nutritional origin or due to cardiopulmonary failure
(Miranda et al. 2004). Many of them might have been
related to an inappropriate composition or consistency
of the diet. Salmonella outbreaks have been observed
in M. tridactyla fed a milk-based diet (J. Gramieri personal communication). Strangulations of the tongue
and degloving injuries are a frequent problem in captive anteaters because tendons or nerves included in the
food become entangled around the tongue (Flint 1997).
Symptoms of this disorder are not specific, but include
inanition and anorexia. Nonetheless, failure to diagnose
the problem can lead to necrosis of the tongue, cachexy,
and eventually death of the affected animal. As preventive measures it is crucial to control food consistency,
remove nerve tissue, finely grind meat, and include inspection of oral cavities in regular physical examinations. Chronic diarrhea often affects anteaters that are
fed imbalanced or liquid diets, but also can be caused
by endoparasites. Rectal prolapse has been reported
frequently in anteaters and can be a complication of
chronic constipation or diarrhea.
Reproduction
Reports on successful breeding are more frequent for
giant than for lesser anteaters. Births of giant anteaters
were recorded as early as 1895, but only one of seven
offspring born prior to 1900 survived (Bartmann 1983).
There are several factors responsible for low reproductive success. Incorrect sexing of anteaters is a common
cause of reproductive failure. In giant anteaters, reproductive management is further hindered by the absence
of visible behavioral or morphological signs of estrus.
Although preovulatory genital hemorrhages have
been reported in lesser anteaters, they are not always
recognized. Pregnancy determination is difficult and
mounting may occur even a few days before delivery,
thus gestations often go unnoticed. As a consequence,
it is common to keep breeding pairs together over long
periods of time to increase the chances of successful
239
breeding. However, this may lower reproductive success by impacting juvenile survival. Survival rates of
juveniles have been estimated at 50%. The low rate of
juvenile survival is caused by stillbirths, incorrect maternal behavior and male aggression. Recent observations suggest that male giant anteaters injure the newborn during attempts to mount the female shortly after
delivery, but it is unknown whether these attempts are
related to a postpartum estrus (Patzl et al. 1998).
Behavior
Captive anteaters spend a large portion of the day sleeping if they are not regularly stimulated to explore new
items in their enclosure, such as rotten logs, pieces of
termite mounds, or food items hidden in artificial termite mounds (figure 22.6). Keeping giant anteaters in
large, open enclosures together with other animal species has been considered beneficial for their well-being
(Bartmann 1983). The absence of such variable stimuli
can lead to behavioral stereotypies, such as pacing and
walking in circles.
Sloths
Bradypus and Choloepus sloths are Neotropical arboreal
mammals that live in the canopy of tropical forests and
have very species-specific dietary, environmental, and
social requirements. These unique animals are appreciated by zoo staff, but are reputedly difficult to maintain
in captivity. It is interesting to note that while Bradypus
represents approximately 40% of the total fauna biomass in selected American rainforests and Choloepus
only 12% (Eisenberg and Thorington 1973), the latter is
most commonly represented in zoological collections.
This apparently is a result of the difficulty in acclimating
Bradypus, an obligate folivore, to an artificial diet (Herbig-Sandreuter 1964; Montgomery and Sunquist 1978).
In fact, even Bradypus maintained on a natural leafy
diet have demonstrated poor longevity (Beebe 1926). It
is possible that the more docile and shy nature of Bradypus also plays a role. A vast majority of pathologies in
captive sloths (96.4%) occur within the first six months
of captivity (Diniz and Oliveira 1999; figure 22.7) and
are usually caused by failure to acclimatize to captive
conditions and/or the artificial diet.
Enclosures
Enclosures should consider space requirements, behavior, preferences for shelter, and exposure to sunlight,
240
Figure 22.6. Anteater enclosures

should be enriched with various
items to stimulate the animals to
explore and forage. Photograph
by F. Miranda.
Figure 22.7. Juvenile Bradypus variegatus with fungal infection. This disease is common in captive Bradypus and may
be caused by stress, inappropriate environmental conditions, or imbalanced diets leading to immunosuppression.
Photograph by S. Moreno, Fundacin UNAU.
moonlight, and rain. Even though Bradypus and Choloepus species differ in their behavior and habits, they
have similar requirements in terms of enclosures. Conventional small areas, or cages made mostly of concrete
or wire mesh, are easy to clean but completely inadequate for these arboreal folivores. Sloths kept as pets in
a conventional house or cage environment frequently
become anorexic and depressed, and die within a short
period of time.
Sloths are well adapted to life in the forest canopy
and therefore require large spaces for captive management. The particularly shy nature of these camouflage
experts requires a rich environment with abundant
trees and branches, plus lianas for movement between

trees and to the ground (figure 22.8). The latter are essential to allow sloths to descend every two to four days
to defecate and urinate. The enclosures should provide
the animals with an area to sunbathe and sites for nesting, such as hammocks placed high above the ground
(figure 22.8). Refuges and hiding spots, simulating
conditions of the natural forest, should be placed in
the enclosure to provide opportunities to retreat when
threatened. Natural materials for enclosure enrichment
are the best option.
Habitat floors should consist of soil and be devoid of
rocks or other hard or sharp objects to prevent injuries
in case of an accidental fall. Sloths will use a small pond
or running water for drinking, bathing, or defecation.
Temperature and humidity should be kept within the
ranges of their natural environment. Sloths are particularly susceptible to low temperatures, and prolonged
exposure may cause irreversible hypothermia (Plese
and Moreno 2005).
In nature, Bradypus and Choloepus are sympatric
(Wetzel 1985a). It is, however, recommended to maintain them in separate enclosures. Choloepus are very active at dusk and sunrise (Chiarello this volume); on the
contrary, at this time of the day Bradypus are generally
sleeping and disruptions are not welcome. Choloepus
may react violently if disturbed while sleeping.
Nutrition
An appropriate diet that satisfies their natural requirements is a crucial factor in the captive maintenance of
241
Figure 22.8. Branches,

lianas, and foliage allow
captive sloths to move
and hide, while baskets
are used to sunbathe and
as nesting sites. Photograph by S. Moreno,
Fundacin UNAU.
sloths. This is not an easy task, as the high percentage

of nutritional and digestive problems (45.7% and 12.3%,
respectively) observed at So Paulo Zoo suggests (Diniz
and Oliveira 1999). The most frequent digestive problems are diarrhea, constipation, and tympany. Uremia,
urolithiasis, and renal failures can appear in sloths receiving diets with excessive protein or calcium content
over an extended period (Rappaport and Hochman
1988; L. S. Brennan personal communication; J. Greathouse personal communication).
Wild Bradypus feed on a large variety of plant species (Chiarello this volume), although most evidence
emphasizes Cecropia spp. as favorites. They feed primarily on leaves, flowers, and seed buds. Wild Choloepus sloths further consume Quercus humboldtii,
Heydyosmium goudotianum, Brunellia sibundoya and
Miconia tonduzii. The natural diet of all sloths contains
low levels of energy and soluble carbohydrates and is
rich in fiber (Beebe 1926; McCrane 1966; Montgomery and Sunquist 1975, 1978; Alvarez 2004; Plese and
Moreno 2005). In Bradypus, there may be considerable
variability among members of any single population in
terms of food preferences. Individual sloths concentrate
large proportions of their time using a single modal tree
species (Montgomery and Sunquist 1975; Chiarello this
volume). Jeuniaux (1962) believed there was individual
variability in the cellulytic power of sloth stomachs. This
fact, coupled with the discovery that certain individuals

become strongly attached to specific tree species, may
mean that there are large individual differences in digestive efficiency over the range of food eaten. In Bradypus, it may be useful to know the individual animals
modal tree or diet regime when acclimating it to a captive diet.
Acclimatization of Choloepus to captive diets has
been more successful than for Bradypus, probably
because they have more generalized food preferences
(Beebe 1926; Eisentraut 19321933; Montgomery and
Sunquist 1975). In addition to fruits and vegetables,
some zoos offer their two-toed sloths animal protein.
For all sloths, it is important to offer a variety of food
items. Food preparation should include items cut
lengthwise to facilitate handling. Fresh, hydrated food
should be offered at least twice a day, ideally early in
the morning and at sunset, and sprinkled with mineral
and vitamin supplements once a day (figure 22.9). Daily
food intake is about 1015% of body weight. In accordance with their arboreal habits, food should be offered
on elevated platforms. Fresh water should be provided
even though sloths can extract considerable liquid from
leaves and other food items. Automatic water dispensers, such as those used in commercial chicken farms,
allow the sloths to ingest water much as they do in the
wild when they lick water off leaves after rainfall.
242
Figure 22.9. Juvenile Bradypus variegatus feeding on

fresh, hydrated leaves. Food
should be offered hanging from branches to allow
sloths to feed in a natural
way. Photograph by S.
Moreno, Fundacin UNAU.
Reproduction
There have been several attempts to study the reproductive biology of sloths. Despite accounts of mating in
the wild and in captivity, reproduction is still the subject of debate. Successful breeding has been reported
for captive Choloepus, but not in Bradypus. Bradypus
are reported to live in high-density populations, with
seasonal reproduction and shorter durations of pregnancy and parental care than in Choloepus (Gilmore et
al. 2000; Taube et al. 2001). Field studies and database
information gathered by Foundation UNAU support
these observations. In Bradypus, seasonality of breeding varies according to climatic differences between locations and between years, and off-season births occur
occasionally. Choloepus live in low-density populations,
reproduction is slightly seasonal, and the duration of
pregnancy and parental care is almost twice as long as
in Bradypus (Gilmore et al. 2000; Taube et al. 2001). Extensive studies carried out on C. hoffmanni at Lincoln
Park Zoo, Chicago, and field work in Panama suggested
that females in estrus actively initiate mating (Meritt
1985a).
Captive breeding of Choloepus is complicated by
the morphological similarity of the two species, C.
hoffmanni and C. didactylus. Many of the wild-caught
founders of captive populations in the United States
appear to have two different geographic origins. The
animals were caught in one country, but exported to
another before final export to the United States. Such a

move caused many of these animals to be misidentified.
Many individuals were classified as C. hoffmanni or C.
didactylus on the basis of the common species in their
country of export, instead of taking their area of origin
into account. Genetic validation is necessary to avoid
possible hybridization (Jorge and Pereira this volume)
that may be suppressing reproductive success.
Infant sloths are completely dependent on the dam
during the first six months and mortality rates are
extremely high (80%). Newly born captive sloths are
often rejected by the mother or fall from her shortly
after birth (McCrane 1966; W. Winstanley personal
communication). Captive male Choloepus can interact
negatively with infants; thus they may need to be separated from females prior to birth. Once past the critical
weaning period, juvenile mortality drops significantly
to about 30%.
Behavior
Sloths live in dense forests with extensive tree cover,
spending the majority of their time among high
branches. Long and light limbs equipped with powerful
claws allow them silent and inconspicuous movements
in the canopy. If kept in enclosures that are appropriate
to the species, that is, which provide sufficient trees and
branches, visitors are challenged to spot and observe
sloths. Their inconspicuousness is enhanced by the fact
that sloths sleep for as many as 20 hours a day, hanging

on a tree branch where they resemble dry leaves, a wasp
nest, or a termite mound.
Bradypus form social groups that share food, shelter,
and even juvenile care, both in nature and in captivity.
Choloepus are more solitary and aggressive than Bradypus, but accept company in captivity. Care should be
taken, however, when introducing new members to an
existing group, as adult males may show their dominance by displaying signs of aggression, such as biting
the nails and toes of opponents. These encounters are
generally short, but may cause permanent injuries if
not treated.
Juvenile Choloepus may show dominance over, or
competition with, smaller sloths by sitting on top of
them, licking their ears, or biting their legs. In some
cases, separation of individuals of different age or size
is necessary to avoid injuries. Bradypus are more docile
and fights among them do not occur frequently. Juveniles rather like to cling to a branch with their hind legs
and play by themselves or with conspecifics in an upside-down position.
Conclusions
The increased scientific interest and intensified research on Xenarthra has not only helped in the better
understanding of the ecology of these peculiar animals,
but also contributed invaluably to improving the husbandry conditions of captive and semicaptive individuals. Several zoos already exhibit captive Xenarthra in
243
enclosures that imitate their natural habitat. These exhibits are often used to increase awareness in the general public of conservation problems and to stimulate
in situ conservation. Promising efforts to initiate ex situ
conservation programs, as well as national and international breeding programs for Xenarthra, are now under
way that will eventually help in preserving wild populations of these fascinating animals.
Acknowledgments
The information about captive maintenance of armadillos stems from research conducted at Zurich University, Switzerland (Superina 2000). M. S. thanks Ewald
Isenbgel for his advice. F. M. is indebted to Marcela
Uhart and the members of Projeto Tamandu at So
Paulo Zoo for their support, and to Neide M. Tanaka
for her continuous encouragement. Experiences in captive management of sloths were the result of the rehabilitation and liberation program of Foundation UNAU
in Colombia. T. P. extends her special thanks to Sergio
Moreno, who helped write the section on sloth maintenance, and all the members of the foundation. Thanks
to Amparo Arango, Rocio Arango, Andrea Arango,
Mary Choperena, and Diego Perez, medical doctors
and veterinarians who have supported this effort with
their knowledge, work, dedication, investigation, and
encouragement. Special thanks are due from all coauthors to Roberto Aguilar for his constant help and encouragement.
23
Exploitation of xenarthrans by the Guaran-Isoseo
indigenous people of the Bolivian Chaco
Comparisons with hunting by other indigenous groups in Latin America, and
implications for conservation
Andrew J. Noss, Rosa Leny Cullar, and Erika Cullar
Resumen
Resumo
Una investigacin participativa a largo plazo con los cazadores Guaran-Isoseos, tierra indgena Isoso, Gran
Chaco boliviano, resalta la importancia de los armadillos para fines de subsistencia. Cinco especies constituyen casi el 50% de mamferos cazados, aunque menos del 10% de la biomasa consumida dado el tamao
menor de estos animales con relacin a los ungulados.
Los usos medicinales y artesanales son secundarios.
Los osos hormigueros no son presas importantes,
mientras que los perezosos no se encuentran en la zona.
Los dems pueblos indgenas de Bolivia en general consumen muchos armadillos tambin, mientras que los
osos hormigueros son menos importantes y los perezosos raras veces son cazados dentro de su rea de distribucin. La presin de cacera sobre xenartros por parte
de pueblos indgenas de la regin Neotropical parece
ser sostenible, porque se relaciona generalmente con la
diversidad y abundancia de las especies de xenartros
disponibles dentro de las zonas de cacera indgenas.
La excepcin notable es el quirquincho Chaetophractus nationi, amenazado de extincin como producto de
una cacera de subsistencia en combinacin con el uso
tradicional de su caparazn para instrumentos musicales, lo que genera un comercio en esta especie y otras
del mismo gnero. Los Xenarthra ms grandes, el armadillo gigante Priodontes maximus y el oso bandera
Myrmecophaga tridactyla, tambin son vulnerables a la
sobre-cacera por parte de cazadores indgenas, y deberan ser protegidos especficamente en el marco de
planes de manejo de fauna en tierras indgenas.
Uma investigao participativa em longo prazo com os

caadores Guaran-Isoseos, rea indgena Isoso, no
Grande Chaco Boliviano, ressalta a importncia dos tatus para fins de subsistncia. Cinco espcies constituem
quase 50% dos mamferos caados, ainda que menos
de 10% da biomassa consumida, pelo tamanho menor
desses animais em relao aos ungulados. Os usos
medicinal e artesanal so secundrios. Os tamandusbandeira no so presas importantes, enquanto que as
preguias no encontram-se na rea. Os demais povos
indgenas da Bolvia, em geral, consumem muitos tatus
tambm, enquanto que os tamanadus-bandeira so
menos importantes e as preguias raras vezes so caadas dentro de sua distribuio. A presso da caa sobre
os Xenarthra por parte dos povos indgenas da regio
Neotropical parece ser sustentvel, porque relaciona-se,
geralmente,com a diversidade e abundncia das espcies de Xenarthra disponveis dentro das reas de caa
indgenas. A notvel exceo o tatu Chaetophractus
nationi, ameaado de extino como produto de caa
de subsistncia em combinao com o uso tradicional
de sua carapaa como instrumentos musicais, o que
gera um comrcio dessa espcie e de outras do mesmo
gnero. Os Xenarthra maiores, o tatu gigante Priodontes
maximus, e o tamandu-bandeira Myrmecophaga tridactyla, tambm so vulnerveis sobre-caa por parte
dos caadores indgenas, e deveriam ser protegidos especificamente em programas de planos de manejo de
fauna em reas indgenas.
244
Exploitation of xenarthrans by indigenous people
Introduction
Across the geographic distribution of xenarthrans,
anthropologists and biologists have taken note of the
relationships between these species and local indigenous populations. Generally, anthropologists focus on
the people themselves and on cultural manifestations
of this relationship, for example, hunting practices,
dietary taboos, traditional beliefs and folk tales (McDonald 1977; Hames and Vickers 1982; Posey 1983b;
Vickers 1984; Hurtado et al. 1985; March 1987; Leeuwenberg 1994; Arenas 2003; Smith 2005). Biologists in
turn tend to emphasize species other than xenarthrans,
particularly ungulates and primates, in part because
they are larger, more visible, easier to study, and more
vulnerable to hunting and other threats. However, the
latter studies do not ignore xenarthrans and their exploitation altogether, describing a range of uses from
subsistence to medicinal to artisanal to commercial,
comparing the exploitation of xenarthrans with that of
other wildlife species, and even evaluating the effects of
exploitation on the conservation status of xenarthrans
(Redford and Stearman 1989; Alvard et al. 1997; Ayala
1997; Baur 1998; E. Cullar 1999; Fang et al. 1999; Puertas 1999; Barbarn 2000; Cabrera et al. 2000; Rumiz et
al. 2001; Zapata Ros 2001; Arispe and Rumiz 2002; Polanco-Ochoa 2003; Copa and Townsend 2004).
The conservation aspect is particularly important
given the extensive and still expanding lands claimed
by and legally titled to indigenous peoples across the
Neotropics. These lands are important complements
to protected areas in assuring the long-term conservation of ecosystems and wildlife. Subsistence hunting is
a traditional right assumed by the indigenous owners
of these lands, a key benefit for the local communities,
and yet possibly a key, long-term threat to wildlife. Sustainable use, through hunting and other land-use practices, together with certain management and conservation measures, can in theory assure that xenarthrans
on indigenous lands can continue to be exploited
without being endangered or eliminated (Vickers 1994;
Townsend 1995, 1996; Ulloa et al. 1996; Robinson and
Bennett 2000; Aquino et al. 2001; Silvius et al. 2004).
These issues are being addressed in the Gran Chaco
of Bolivia. The vast Kaa-Iya del Gran Chaco National
Park, covering 34,400 km2, was created in 1995 as a result of a proposal by three indigenous groups who live
around its boundaries and historically occupied parts
or all of the area: the Isoseo-Guaran, the Chiquitano,
245
and the Ayoreode. Currently none of the groups, except

for an uncontacted band of Ayoreode still living within
the park, use the park lands for settlement, farming, or
hunting. The indigenous group with the largest local
population and strongest organizational structure, the
Isoseo-Guaran, were charged with the responsibility
of co-administering the park, together with the national
park service (Taber et al. 1997). Subsequently, the Isoseos claimed and are in the middle of a titling process
for the 19,000 km2 Isoso territory bordering the park
to the west (figure 23.1). The New York-based Wildlife
Conservation Society (WCS) worked with the Isoseos to prepare the park proposal. Since 1996 WCS has
worked with several hundred hunters in the 25 Isoseo
communities to describe and monitor hunting practices
and to develop management plans, both for subsistence
hunting in particular, and for the indigenous territory
as a whole (Painter and Noss 2000; Noss and Painter
2004).
With respect to xenarthrans, armadillos may reach
their peak of diversity in the Chaco. Six species are
present in the Isoso alone, with five of them frequently
hunted and eaten by indigenous residents. The two anteaters present are also hunted, though less frequently.
This chapter focuses on a case study of subsistence
hunting of xenarthrans by the Isoseo-Guaran indigenous people of the Bolivian Chaco. We describe the
types of uses, evaluate the importance of xenarthrans
for local human consumption relative to other sources
of wild animal meat, and discuss the sustainability of
their exploitation. Finally, we compare this case study
with data from other indigenous groups in Bolivia as
well as across the Neotropics.
Methods
One of the principal methods for compiling hunting
data, employed in this study (and others cited below),
is hunter self-monitoring. Beginning in 1996, we developed data sheets with the hunters that they could carry
on hunts and complete information on the species captured, location, and date, among other things (see Noss
et al. 2003; Noss, Cullar, and Cullar 2004). In 1997, we
hired 11 half-time wildlife monitors, each responsible
for one or two communities, to visit hunters periodically and collect the written data. All hunters participated on a voluntary basis, according to each individuals willingness to contribute information. Hunting
monitors also visited hunting locations named by the
246
A. J. Noss, R. L. Cullar, and E. Cullar
Figure 23.1. Kaa-Iya National Park and Isoso indigenous territory, Bolivia.
hunters to record coordinates using a hand-held GPS

receiver (Garmin 12 XL). Information on other uses of
xenarthrans was compiled through direct observation
and interviews with local residents.
In order to derive total offtakes and evaluate hunting sustainability (Noss 2000) from the voluntary selfmonitoring, we developed a complementary survey to
record the monthly activity (active hunter providing
data, active hunter not providing data, not a hunter, absent from the community) of all potential hunters (men
and boys ages 15 and up). This allowed us to determine
the proportion of hunters participating by community
and by time period (see Noss et al. 2003, 2005; Noss,
Cullar and Cullar 2004). More than 700 hunters have
provided hunting data, with around two-thirds of active hunters participating. We applied standard models
to estimate hunting sustainability, though precise data
on the numerous factors to be included in the model
(population density, age of first and last reproduction,
gestations per female per year) remain only estimates
from inconclusive research at the field site and/or figures from other sites where environmental conditions
may differ significantly.
Comparative data for other indigenous groups across
Latin America are drawn principally from the literature

on wildlife management, including biannual regional
conferences on the topic. The issue of conservation
through sustainable wildlife management, particularly
on indigenous lands, brings together anthropologists
and biologists in evaluating current practices (what
people do, which may not coincide with what they say
or believe) and proposing long-term solutions.
Results
The only xenarthran not used by the Isoseos is the tiny
and rarely encountered fairy armadillo Chlamyphorus
retusus. However, the Isoseos consider sightings of
this animal, possibly because of its unusual physical appearance and/or its rarity, to be an evil omen portending a death in the family. Therefore any fairy armadillo
that does appear is hurriedly dispatched in the hopes of
warding off the curse (Cullar 2001). Similar beliefs are
held by peoples inhabiting the Argentine (Arenas 2003)
and Paraguayan (Meritt this volume) Chaco.
The five species of armadillos in the Isoso are the
most important group, by number of individuals hunted
(though not biomass), of mammals exploited for sub-
247
Table 23.1. Subsistence Hunting of Mammals in Isoso, 19962004

Species
Xenarthrans





Tolypeutes matacus


Other mammals

Catagonus wagneri

Dasyprocta variegata

Mazama gouazoubira

Tapirus terrestris

Tayassu tajacu

Tayassu pecari

Carnivores (spp. unknown)

Other rodents

Other mammals
Total
Number
Biomass (kg)
% total number
% total biomass
322
1,057
1,904
405
1,988
25
12
363
2,162
4,138
1,421
2,300
475
36
3%
9%
16%
3%
17%
<1%
<1%
<1%
2%
3%
1%
2%
<1%
<1%
25
214
3,069
180
1,779
418
86
127
34
11,645
704
762
51,620
19,633
31,595
11,236
1,282
324
135
128,188
<1%
2%
26%
2%
15%
4%
1%
1%
<1%
1%
1%
40%
15%
25%
9%
1%
<1%
<1%
Source: Data from hunter self-monitoring.
sistence use (table 23.1). The meat of the three-banded

armadillo, Tolypeutes matacus, is preferred above all
other game. Some Isoseos do not eat the yellow armadillo, Euphractus sexcinctus, because of the flavor of
its meat (see also Meritt this volume). Most armadillo
hunting takes place at night, using dogs to track them.
The nine-banded armadillo, Dasypus novemcinctus,
generally takes refuge in a burrow and is caught then
by digging it out with a machete or shovel. The fat or
carapace of these three species is also used by the Isoseos to cure a variety of conditions, such as diarrhea,
bronchitis, asthma, snakebites, cuts, and eye irritations.
They are sold when the opportunity presents itself,
though no established markets exist within Isoso nor
are there means of exporting game outside the Isoso
(Cullar 2000).
Considering estimated total offtakes by Isoseo
hunters over the entire hunting range of approximately
4,000 km2, current use of all the species, except for T.
matacus and, possibly, Chaetophractus vellerosus, appears to be sustainable (Noss 2000). All five hunted
species continue to be found within 5 km of the Isoso
communities, with only a few important hunting locations 1020 km away, and very little hunting inside the
Kaa-Iya National Park itself (figure 23.2). The threebanded armadillo, with only one young per gestation,
has the lowest productivity of any of the armadillos.
Unfortunately its defense mechanism of rolling into a
ball, excellent against jaguars or foxes, leaves it at the
mercy of human hunters. The diurnal armadillos E. sexcinctus and Chaetophractus villosus may be less vulnerable to Isoseo hunters because they are relatively good
runners and diggers, and dogs and hunters also avoid
the hotter hours of the day. The nine-banded armadillo, with its identical quadruplets in every gestation
(Prodhl et al. this volume), is much more productive
and, as such, resistant to hunting pressure.
Giant anteaters, Myrmecophaga tridactyla, are rarely
encountered in the Isoso hunting areas, perhaps because of low natural densities in the Chaco, which may
be exacerbated by hunting pressure. Isoso hunters do
kill and eat them when discovered. The tamandua, Tamandua tetradacyla, is encountered more frequently,
but not actively pursued, as Isoseos do not eat it. It
is killed occasionally by dogs, and its tongue at least is
saved as a good luck charm.
Table 23.2 presents comparative hunting information
for other lowland indigenous peoples in Bolivia, and
table 23.3 indicates which species are hunted and for
what purposes, including the Andean armadillo, Chaetophractus nationi. Sloths do not occur in the Isoso, and
are only rarely hunted by other indigenous groups in
other areas of Bolivia where they do occur. As in the
Isoso, other indigenous groups hunt the giant anteater
and the tamandua, but these two species represent no
more than 5% of hunted mammals and 9% of mammalian biomass. Armadillos are relatively unimportant
to the Tsimane and Tacana, but represent 2550% of
248
Figure 23.2. Isoseo hunting of

armadillos (19962004).
the mammals hunted by the other indigenous groups

recorded, and 513% of mammalian biomass.
At least two species of armadillos appear at risk because of human impacts. Chaetophractus nationi in the
Bolivian altiplano is severely endangered as a result of
subsistence hunting combined with the traditional use
of its carapace for the guitar-like musical instrument
called the charango and the percussion instrument
called the matraca (H. Gmez personal communication; Torrico et al. 2005), and as a fortune-telling talisman (Romero-Muoz and Prez-Zubieta 2005). The
latter two uses have fostered commercial markets in
urban areas, and Bolivia has begun to import carapaces of other Chaetophractus species from Argentina
because of the scarcity. In 1997, Bolivia proposed that
Cha. nationi be transferred from CITES Appendix II
to Appendix I, and a zero export quota has been established. It is ranked as vulnerable on the IUCN Red List
of Threatened Species (see Aguiar 2004; Aguiar and
Fonseca this volume). The giant armadillo Priodontes
maximus does not occur in the Isoso, but is also vul-
nerable to subsistence hunting across its range. This is

the only armadillo species currently in CITES Appendix I and is considered vulnerable by the IUCN (Aguiar
2004; Aguiar and Fonseca this volume).
In Bolivia there is some sale of armadillo meat in
rural and urban markets and restaurants, but this pressure appears to focus on nine-banded armadillos from
partially degraded habitats and is probably not a threat
to the persistence of the species.
Across the Neotropics, all indigenous groups for
which data are available use one or more of the xenarthrans present in their hunting areas (table 23.4).
Armadillos are generally the most frequently hunted,
anteaters by a smaller number of indigenous groups,
and sloths by only a few. Similar to the Isoseo-Guaran, armadillos are the most important group of species
hunted by the Ach of Mbaracay Reserve in Paraguay
(Hill et al. 1997; Hill and Padwe 2000), by the Chiquitano in Bolivia (Guinart 1997), and by local residents
throughout the Argentine Chaco (Zuleta and Bolkovic
1994; Bolkovic 1999). The Siona-Secoya of Ecuador evi-
249
Table 23.2. Hunting of Xenarthra and Other Principal Mammal Taxa by Indigenous Peoples in Bolivia
Group and location
Ayoreode, Tobit

Species

% Na

% Bb
Chiquitano, Lomero

Species

% N

% B
Chiquitano, Lomero

Species

% N

% B
Guaran, Akae

Species

% N

% B
Guaran, Isoso

Species

% N

% B
Sirion, Ibiato

Species

% N

% B
Tacana, Ixiamas

Species

% N

% B
Tsimane, Santa Rosita

Species

% N

% B
Tsimane, Tacuaral

Species

% N

% B
Sloths
Anteaters
Armadillos
Primates
Ungulates Carnivores Rodents Reference
0
0
0
0
0
0
5
26
5
??
7
0
4
57
89
??
7
6
1
2
0
0
0
nac
1
1
na
2
53
na
??
2
na
5
31
na
??
16
na
3+
17
na
1
0
na
1
1
na
4
43
na
??
2
na
5
27
na
??
14
na
3+
14
na
0
0
0
2
3
9
2
26
7
1
2
0
2
31
62
??
8
13
1+
23
9
0
0
0
2
0
0
5
49
8
0
0
0
5
47
90
8
1
1
3+
3
1
0
0
0
2
2
1
3
38
13
??
2
0
6
26
74
??
14
4
2+
18
8
1
0
0
3
0
0
2
4
3
3+
42
16
4
28
70
4+
14
6
3+
11
6
0
0
0
2
5
4
1
1
0
4
52
23
3
24
61
1
11
5
3+
7
7
0
0
0
2
5
3
3
2
2
4
32
10
3
35
73
1
12
5
3+
12
6
Sources: (1) Townsend 1995, 1996, 2000, (2) Ayala 1997, (3) R. L. Cullar 1997, 1999, (4) Guinart 1997, (5) R. L. Cullar 2000, (6) Rumiz et
al. 2001, (7) Copa and Townsend 2004, (8) Gmez personal communication.
a. % N = proportion of total number of mammals harvested.
b. % B = proportion of total biomass of mammals harvested.
c. na = data not available.
dently do not prefer armadillos but may use them as a

fallback when other sources are unavailable or decline
(Vickers 1991).
Table 23.5 details recorded uses by species. In the
Ach case, armadillos are also the most important
group of mammals in terms of biomass (Hill and Padwe
2000). Medicinal uses are generally secondary benefits
of animals hunted for subsistence. The most important
medicinal products are the hair, tongue, and claws of
anteaters; hair of sloths; and the carapace, claws, and fat
of armadillos (Cullar 2000; Medinaceli 2004). With respect to commercial use, no meat trade in xenarthrans
was reported that originated from indigenous peoples.

Only two studies note a trade in pets: for silky anteaters
(Aquino et al. 2001) in Peru, and for giant armadillos
and giant anteaters (Bolkovic 1999) in Argentina.
Several researchers also highlight taboos on eating
the following meats by children or by women/parents
before and/or after the birth of a child: nine-banded
armadillos (KayapMcDonald 1977; XavanteLeeuwenberg and Robinson 2000), and giant anteaters
(Kayap, Yanomam, Sirion, Tenetehara, Tapirap,
TimbiraMcDonald 1977). Among the Toba and
Wich of Argentina, Arenas (2003) reports taboos for
Table 23.3. Uses of Xenarthra by Indigenous Groups in Bolivia

Subsistence
Armadillos

3,8

Chaetophractus nationi
14

4,5,8

5

Dasypus kappleri
8,15

18,10,11,15,16

2,3,5,710,16

Priodontes maximus
1,2,3,8,10,11,15,16

Tolypeutes matacus
2,5,8,9,16
Anteaters

Cyclopes didactylus

1,2,4,5,8,10,15,16

1,3,4,7,8,10,15,16
Sloths

Bradypus variegatus
3,8,15
Medicinal
Artisanal
Commercial
14
12,13,14
12,13,14
8,15
15
5,8,11,15
15,16
8,9,16
8,11,15
8,15
5,8
9,16
15
5,8,15
8
15
5,8
8
5,8,11,15
5,8,15
15
8,15
15
15
15
12,15
11
15
Sources (indigenous group in parentheses): (1) Townsend 1995, 1996, 2000 (Sirion), (2) Ayala 1997 (Ayoreo), (3) Guinart 1997 (Chiquitano), (4) R. L. Cullar 1997, 1999 (Guaran), (5) Noss 1998, E. Cullar 1999, R. L. Cullar 2000 (Guaran-Isoseo), (6) Aguirre et al.
1999 (Tsimane), (7) Rumiz et al. 2001 (Chiquitano), (8) Arispe and Rumiz 2002 (Chiquitano), (9) Dosapey 2003 (Ayoreo), (10) Copa and
Townsend 2004 (Tsimane), (11) Medinaceli 2004 (Tsimane), (12) Romero-Muoz and Prez-Zubieta 2005 (Aymara), (13) Torrico et al.
2005 (Aymara), (14) Gmez personal communication (Aymara), (15) Gmez personal communication (Tacana), (16) Pomeroy personal
communication (Ayoreo).
Table 23.4. Hunting of Xenarthra and Other Principal Mammal Taxa by Indigenous Peoples Outside of Bolivia
Group and location
Ach, Paraguay
Species
% N
% B
Andoke, Colombia
Species
% N
% B
Bugl, Panama
Species
% N
% B
Ember, Colombia B
Species
% N
% B
Ember, Colombia J
Species
% N
% B
Ember, Colombia SM
Species
% N
% B
Ember, Colombia UC
Species
% N
% B
Huaorani, Ecuador
Species
% N
% B
Sloths
Anteaters
Armadillos
Primates
Ungulates Carnivores
Rodents
Reference
0
0
0
2
0
1
3
44
35
2
30
14
4
4
27
1
9
6
1
12
16
0
0
na
1
0
na
3
3
na
5
4
na
5
37
na
??
2
na
3+
54
na
20
2
8
10
0
0
0
1
15
13
3
7
13
3
3
24
??
6
7
2+
50
31
21
1
1
0
0
0
0
1
12
6
??
4
1
2
30
63
0
0
0
2+
53
30
0
0
0
0
0
0
1
10
6
??
2
1
2
23
57
??
2
1
2+
63
35
0
0
0
0
0
0
1
13
6
??
2
1
2
26
65
??
1
0
2+
58
25
0
0
0
1
0
1
2
14
5
??
2
0
2
27
68
??
1
1
2+
57
25
0
0
na
0
0
na
1
1
na
??
72
na
4
17
na
1
1
na
2+
9
na
15
continued
Table 23.4.Continued
Group and location
Sloths
Huaorani, Ecuador
Species
0
% N
0
% B
0
Lacandones, Mexico
Species
0
% N
0
% B
0
Maya, Mexico
Species
0
% N
0
% B
na
Maya, Mexico
Species
0
% N
0
% B
0
Maya, Guatemala
Species
0
% N
0
% B
na
Maya, Guatemala
Species
0
% N
0
% B
0
Pacaya Samiria, Peru A
Species
0
% N
0
% B
na
Pacaya Samiria, Peru B
Species
0
% N
0
% B
na
Pacaya Samiria, Peru C
Species
0
% N
0
% B
na
Quichua, Ecuador
Species
0
% N
0
% B
0
Quichua, Ecuador
Species
2
% N
0
% B
0
Shuar, Ecuador
Species
0
% N
0
% B
0
Siona Secoya, Ecuador
Species
0
% N
0
% B
0
Tamshiyacu Tahuayo, Peru A
Species
0
% N
0
% B
na
Tamshiyacu Tahuayo, Peru B
Species
0
% N
0
% B
na
Anteaters
Armadillos
Primates
Rodents
Reference
2
0
1
2
1
1
7
54
37
4
14
44
??
4
3
3+
27
15
12
0
0
0
1
14
5
2
4
1
5
32
66
1
2
1
1
47
26
19
0
0
na
1
1
na
0
0
na
3
75
na
1
10
na
1
6
na
0
0
0
1
17
5
0
0
0
4
40
80
1
17
3
2
26
11
13
0
0
na
1
9
na
0
0
na
4
53
na
0
0
na
1
39
na
11
0
0
0
1
28
8
0
0
0
5
32
75
1
6
2
1
34
14
0
0
na
2
3
na
??
59
na
4
12
na
??
5
na
3
21
na
14
0
0
na
2
2
na
??
25
na
4
57
na
??
3
na
2
13
na
14
0
0
na
2
3
na
??
42
na
4
27
na
??
4
na
2
23
na
14
0
0
0
2
7
4
??
33
17
4
18
60
??
3
1
2+
38
17
16
1
0
0
1
0
0
??
16
8
4
20
47
??
14
5
4
50
39
17
0
0
0
1
4
2
??
31
19
3
20
56
??
3
2
3+
39
21
16
0
0
0
2
2
2
7
38
16
5
45
78
0
0
0
5
11
2
2
1
na
1
1
na
??
5
na
5
59
na
??
1
na
3+
33
na
2
3
na
2
3
na
??
18
na
5
42
na
??
6
na
2+
28
na
8
continued
252
Table 23.4.Continued
Group and location

Ticuna, Colombia
Species
% N
% B
Waimiri Atroari, Brazil
Species
% N
% B
Xavante, Brazil
Species
% N
% B
Yanesha, Peru
Species
% N
% B
Yanomam, Venezuela
Species
% N
% B
Yekwana, Venezuela
Species
% N
% B
Sloths
Anteaters
Armadillos
Primates
Rodents
Reference
1
1
0
3
2
2
2
5
2
6
9
1
5
43
87
??
2
0
3+
34
7
0
0
0
2
<1
<1
3
4
2
4
39
11
5
46
85
2+
<1
<1
3
10
2
0
0
0
2
22
11
2
9
3
0
0
0
7
69
86
0
0
0
0
0
0
10
1
0
0
0
0
0
1
24
11
2
3
3
4
8
34
??
12
5
3+
53
47
18
1
1
1
2
11
22
2
12
8
??
24
8
4
16
52
0
0
0
??
31
9
0
0
0
1
1
0
1
12
5
??
33
10
5
22
71
??
0
0
??
26
14
Sources: (1) Hames and Vickers 1982, (2) Vickers 1994, (3) Ulloa et al. 1996, (4) Hill et al. 1997, (5) Baur 1998, (6) Hernndez-Betancourt
et al. 1998, (7) Gaitn 1999, (8) Puertas 1999, (9) de Souza-Mazurek 2000, (10) Leeuwenberg and Robinson 2000, (11) McNab 2000, (12)
Mena Valenzuela et al. 2000, (13) Weber 2000, (14) Aquino et al. 2001, (15) Franzen 2001, (16) Mena Valenzuela and Cueva 2001, (17)
Zapata Ros 2001, (18) Gonzlez 2003, (19) Guerra and Naranjo 2003, (20) Rodrguez and van der Hammen 2003, (21) Smith 2005.
Notes: See Table 23.2 for definitions of table entries.
children, youths, and women against eating the meat of

Chaetophractus vellerosus, Cha. villosus, and Tolypeutes
matacus. All anteaters and sloths are tabooed as food by
the Siona-Secoya and do not appear in hunting records
(Vickers 1991). In other cases, as with the Isoseos, no
taboo is stated, but the Piro of Peru avoid hunting and
eating the tamandua (Mitchell and Raez Luna 1990),
and the Lacandn of Mexico avoid both the tamandua
and the silky anteater (March 1987). In most of these
cases, traditional taboos are being forgotten or abandoned.
How sustainable is the hunting of xenarthrans? Researchers agree that hunting of the nine-banded armadillo is sustainable, although populations may be
reduced near settlements, and this species is therefore
suitable for management and continued exploitation
(Hames and Vickers 1982; Vickers 1994; Guinart 1997;
Hill et al. 1997; Baur 1998; Aguirre et al. 1999; Hill and
Padwe 2000; Townsend 2000; Gonzlez 2003; KoppValdivia 2005). In contrast, just as in the Isoso, the
three-banded armadillo is reported to be declining as a
result of hunting pressure in the Argentine Chaco (Barbarn 2000). In two notable cases, anteaters provide ap-
proximately 10% and 20% of biomass to Xavante and

Yanomam hunters, respectively (Hames and Vickers
1982; Leeuwenberg and Robinson 2000). In the only
evaluation for the species, Leeuwenberg (1994) considers hunting of the giant anteater to be sustainable
within the Xavante reserve.
Discussion
Research on free-ranging Xenarthra remains constrained by the lack of appropriate research methods
(see chapters by Chiarello, McDonough and Loughry,
and Rodrigues et al. this volume). In particular we lack
suitable methods to count animals, yet the basis of sustainable use programs requires information about population densities, both under natural conditions and
where hunting occurs. Line transects and camera trap
surveys provide some data but are not sufficient for estimating population density for most species. However,
systematic camera trapping has been applied to giant
armadillos (Noss, Pea, and Rumiz 2004) and shows
promise for giant anteaters where the latter are sufficiently abundant. A method developed by K. Hill and
253
Table 23.5. Uses of Xenarthrans by Indigenous Peoples Across the Neotropics

Armadillos

Cabassous centralis

Cabassous chacoensis

Cabassous tatouay




Dasypus kappleri


Dasypus septemcinctus

Dasypus yepesi


Priodontes maximus

Tolypeutes matacus
Anteaters

Cyclopes didactylus


Sloths

Bradypus variegatus

Choloepus hoffmani

Choloepus didactylus
Subsistence
Medicinal
8
12
9,17
17
5
23,25
12,25
6,16,20,28
114,1620,22,2427,29
8,17
5,9,12,17,28
17
23
5,9,17,23,25
17
1,4,5,13,14,16,18,21,22,25
7,12,15,23,25
Artisanal
25
25
1,13
1,4,5,8,9,13,14,16,18,25,28
1,5,8,9,13,14,16,18,24,25
8
8
8
17,25
17
1,8,24,26
8,13
24
8
8
8
8
Sources (country in parentheses): (1) Hames and Vickers 1982 (Venezuela), (2) Hurtado et al. 1985 (Paraguay), (3) March 1987 (Mexico),
(4) Mitchell and Raez Luna 1990 (Peru), (5) Leeuwenberg 1994, Leeuwenberg and Robinson 2000 (Brazil), (6) Vickers 1994 (Ecuador),
(7) Zuleta and Bolkovic 1994 (Argentina), (8) Ulloa et al. 1996 (Colombia), (9) Hill et al. 1997, Hill and Padwe 2000 (Paraguay), (10) Baur
1998, McNab 2000 (Guatemala), (11) Hernndez Betancourt et al. 1998 (Mexico), (12) Bolkovic 1999 (Argentina), (13) Gaitn 1999 (Colombia), (14) Puertas 1999 (Peru), (15) Barbarn 2000 (Argentina), (16) de Souza-Mazurek et al. 2000 (Brazil), (17) Fernandes-Pinto and
Krger 2000 (Brazil), (18) Mena Valenzuela et al. 2000 (Ecuador), (19) Weber 2000 (Mexico), (20) Aquino et al. 2001 (Peru), (21) Franzen
2001 (Ecuador), (22) Mena Valenzuela and Cueva 2001 (Ecuador), (23) Vizcano and Giallombardo 2001 (Argentina), (24) Zapata Ros
2001 (Ecuador), (25) Arenas 2003 (Argentina), (26) Gonzlez 2003 (Peru), (27) Guerra and Naranjo 2003 (Mexico), (28) Rodrguez and
van der Hammen 2003 (Colombia), (29) Smith 2005 (Panama).
Note: Hunting studies do not always confirm how each species is used, and therefore records may be inappropriately classified in the
subsistence column.
colleagues in the Mbaracay forests involves counting

burrows occupied by D. novemcinctus (Hill et al. 1997).
This method is derived from traditional hunting practices and depends on the experience of Ach hunters in
knowing which burrows are occupied. Unfortunately,
this method is not broadly applicable because hunters
in other areas do not have the same experience, plus,
in some areas such as the Isoso, burrows are dug and
occupied by more than one species (e.g., all five species
of armadillos, tuco tuco rodents [Ctenomys spp.], and
tegu lizards [Tupinambis spp]). Cullar (2002) tested
a survey method derived from another local hunting
practice, that is, tracking with dogs, but the difficulty
remains in converting encounter rates to a density estimate. Telemetry and spool-and-thread studies of T.
matacus have provided some initial information on
ranging patterns, but the animals shed their spools
after five days or fewer, and radio transmitters affixed

with glue lasted no more than 60 days (Perrogn 2001;
Barrientos 2004; Barrientos and Cullar 2004; Cullar
this volume).
The effects of exploitation by indigenous peoples on
xenarthrans remain poorly described. Certainly the
largest and most conspicuous, the giant armadillo and
the giant anteater, would appear to be protected within
large national parks in Bolivia and across their range.
But scanty evidence suggests that subsistence hunting
can greatly reduce or eliminate them from indigenous
territories and other lands outside fully protected areas
(Aguiar and Fonseca this volume; Meritt this volume).
In general, populations of the smaller species appear to
be stable even where they are hunted, particularly those
that reproduce the most quickly and are most adaptable, such as the nine-banded armadillo. The status of
254
the three-banded armadillo T. matacus is troubling,

particularly considering the fate of its congener T.
tricinctus, which is considered vulnerable as a result of
hunting and habitat destruction (Aguiar 2004; Aguiar
and Fonseca this volume). If T. matacus is to persist on
indigenous lands, management measures are required
to bring hunting offtakes down to sustainable levels,
and to preserve adequate habitat. Small communal and
private protected areas can be valuable conservation
elements as well, since the species is not wide ranging.
By shifting hunting pressure to other, more resilient
species, including D. novemcinctus in particular, indigenous peoples across the Neotropics can continue
to generate benefits from the sustainable use of xenarthrans.
Traditional indigenous wildlife management practicesincluding hunting zone rotation, reserve systems
or sacred areas, taboos, and restrictions imposed by
shamanshistorically favored conservation of xenarthrans and other hunted wildlife (Taylor 1972; ReichelDolmatoff 1976; McDonald 1977; Ross 1978; Posey 1983a;
Hames 1991; Ulloa et al. 1996; Robinson and Bennett
2000; Arenas 2003; Silvius et al. 2004). Whether or not
they originated as conscious conservation measures, the
effectiveness of these practices is increasingly reduced

by human population growth, sedentarization, and integration in market economies (Moran 1981; Berlin and
Berlin 1983; Redford and Mansour 1996). Nevertheless, they should not be ignored, but rather integrated
with new wildlife management measures, developed
in the context of formally recognized and legally titled
indigenous lands such as the Isoso and others across
the Neotropics.
Acknowledgments
This publication was made possible in part by financial
support from the United States Agency for International Development (USAID, Cooperative Agreement
No. 511-A-00-01-00005). The opinions expressed are
those of the authors and do not necessarily reflect the
criteria of USAID. We particularly thank the people of
the Isoso for their participation and support. We also
thank the Capitana de Alto y Bajo Isoso, the Kaa-Iya
del Gran Chaco National Park, the Sistema Nacional de
reas Protegidas, and the Direccin General de Biodiversidad for institutional support and permissions.
5
Living Xenarthra
Ecology and Behavior
24
Anteater behavior and ecology
Flvio H. G. Rodrigues, sis M. Medri, Guilherme H. B. de Miranda,
Constana Camilo-Alves, and Guilherme Mouro
dad de los animales. Existe una gran superposicin de

reas entre diferentes individuos del mismo sexo y de
sexos opuestos. El principal riesgo para las poblaciones
de osos hormigueros es la fragmentacin de hbitats,
aunque estos animales (a excepcin de C. didactylus)
poseen la capacidad de atravesar ambientes alterados
por el hombre.
Resumen
En este captulo son consideradas las caractersticas
ecolgicas, comportamentales y reproductivas de las
cuatro especies actuales de osos hormigueros: el oso
hormiguero gigante (Myrmecophaga tridactyla), dos
especies de osos meleros (Tamandua mexicana y Tamandua tetradactyla) y el oso hormiguero arborcola
(Cyclopes didactylus), con nfasis en el oso hormiguero
gigante. Los osos hormigueros habitan en una gran
variedad de hbitats y slo el menor de ellos (C. didactylus) es exclusivamente arborcola. Las dos especies
del gnero Tamandua son trepadoras y habitan principalmente ambientes forestados. Por su parte, el oso
hormiguero gigante habita en reas forestadas y tambin en reas abiertas, siendo ms abundante en estas
ltimas debido a su limitada capacidad para trepar. Los
osos hormigueros son especialistas que se alimentan
principalmente de hormigas y termitas y la proporcin
de estos dos grupos consumidos vara de una especie
a otra, entre regiones y estacionalmente. Las presas
de estos animales son un recurso relativamente abundante. Sin embargo, la obtencin de alimento no es
fcil y la ecologa trfica est fuertemente influenciada
por el tipo de defensa desarrollado por cada presa. Los
osos hormigueros son normalmente solitarios aunque
pueden ser vistos en pareja durante la estacin reproductiva. El tamao del mbito del hogar calculado para
estos animales, y principalmente para el oso gigante,
es muy variable y depende de diferentes factores como
la regin de estudio, el mtodo utilizado para estimar
el rea de vida, el grado de intervencin humana, y
otras caractersticas de la regin tales como la temperatura, un factor que puede interferir directamente en
el patrn de movimientos y en los horarios de activi-
Resumo
Neste captulo ns consideramos as caractersticas
ecolgicas, comportamentais e reprodutivas das quatro
espcies atuais de tamandus: o tamandu-bandeira
(Myrmecophaga tridactyla), duas espcies de tamandus-mirins (Tamandua mexicana e Tamandua tetradactyla) e o tamandua (Cyclopes didactylus), com nfase no tamandu-bandeira. Os tamandus ocorrem em
uma variedade grande de habitats, sendo que o tamandu o nico essencialmente florestal e arborcola. As
duas espcies do gnero Tamandua so escansoriais e
habitam principalmente ambientes florestais, enquanto
que o tamandu-bandeira tem limitada capacidade de
escalar e ocorre tanto em ambientes florestais quanto
em reas abertas, sendo geralmente mais abundantes
nestas. Os tamandus so especialistas em se alimentar
de cupins e formigas, porm a proporo de cupins e
formigas varia de uma espcie para outra, entre indivduos da mesma espcie, diariamente, sazonalmente e
entre regies. Apesar de suas presas serem um recurso
abundante, a obteno de alimento por tamandus
no to simples e a ecologia alimentar fortemente
influenciada pelo tipo de defesa apresentado por suas
presas. Tamandus so normalmente solitrios, mas
podem ser vistos aos pares na estao reprodutiva. O
tamanho das reas de vida calculadas para tamandus,
257
258
F. H. G. Rodrigues et al.
em especial o tamandu-bandeira, em diferentes locais

varivel, de acordo com o mtodo de estimativa, grau
de interferncia humana e caractersticas dos habitats,
como a temperatura, que pode interferir diretamente
no padro de movimentao e no horrio de atividade
dos animais. H grande sobreposio de reas entre indivduos, tanto de mesmo sexo quanto de sexos opostos. O principal risco para populaes de tamandus a
fragmentao de habitats, apesar de possurem (exceto
o tamandu) relativa capacidade de atravessar matriz
de habitats alterados pelo homem.
Introduction
The ecology of species, such as their patterns of movement and activity and the occurrence of intraspecific
interactions, is affected by their feeding behavior. Most
insectivorous mammals consume a wide range of insect
types and even other invertebrates. Some mammals,
however, specialize in feeding on social insects such as
ants and termites. A unique feature of such prey is that
they are concentrated and localized. Moreover, termites
and ants are an abundant and easily found source of
food. However, the number of mammals specializing,
to a greater or lesser extent, in the consumption of ants
and termites is relatively small. These mammals include
the aardwolf (Proteles cristatus), pangolins (Manis spp.)
and aardvarks (Orycteropus afer) of Africa, the numbat
(Myrmecobius fasciatus) and echidnas (Tachyglossus
and Zaglossus) of Oceania, and the hoary fox (Pseudalopex vetulus), some armadillos (e.g., the giant armadillo
[Priodontes maximus] and the three banded armadillo
[Tolypeutes tricinctus]), and all the anteaters (families
Myrmecophagidae and Cyclopedidae) of South America.
Most termitophagous and myrmecophagous species
are highly specialized. For example, anteaters evolved
a morphology highly adapted to the consumption of
termites and ants, possessing characteristics such as
strong forelegs equipped with powerful claws, a long,
sticky and protractile tongue (with insertion in the sternum, a convergence with pangolins), a small mouth,
long snout and well developed salivary glands, as well
as a highly refined sense of smell (Nowak 1999; see also
McDonald et al. this volume). Among the Xenarthra,
they are the only ones completely devoid of teeth. This
specialization is directly related to the anteaters way of
life, which has been documented since the pioneering
work of Montgomery and Lubin (1977). Although several studies have focused on the ecology and behavior
of anteaters since then (e.g., Lubin and Montgomery

1981; Lubin 1983; Montgomery 1985b,c; Redford 1985b;
Shaw et al. 1985, 1987; Drumond 1992; Rodrigues et al.
2001; Medri 2002; Mouro and Medri 2002; CamiloAlves 2003; Miranda 2004; Medri and Mouro 2005a;
Camilo-Alves and Mouro 2006), many aspects, such
as ecophysiology, sociobiology, and population dynamics, remain little explored.
Four living species of anteaters occur from Mexico
to Argentina: the giant anteater (Myrmecophaga tridactyla), the southern tamandua (Tamandua tetradactyla),
the northern tamandua (T. mexicana), and the silky
anteater (Cyclopes didactylus). Current distributions
and conservation status for each species can be found
in the chapter by Aguiar and Fonseca in this book.
Size varies enormously within the group. The giant
anteater can exceed 35 kg in weight, more than one
hundred times the mass of the silky anteater (~300
g) and about seven times the mass of southern and
northern tamanduas (~5 kg). Most of the information
available on anteaters refers to the giant anteater. Few
studies have been dedicated to the northern and southern tamanduas and almost nothing is known about the
silky anteater, possibly because of its cryptic nocturnal
and arboreal habits.
Our objective in this chapter is to provide updated
information on the ecology of anteaters. Our main focus is the giant anteater, and we present the most recent
information obtained by two research groups working
with this species in areas of flooded savanna (Pantanal)
and savanna (Emas National Park) in central Brazil.
Feeding behavior
Although todays four anteater species specialize in foraging for termites and ants (table 24.1), the proportion
of ants and termites in the diet varies among species,
individuals, regions, and seasons of the year. The silky
anteater specializes in ants (Best and Harada 1985; Lubin 1983; Montgomery 1985c), although beetles have
also been recorded in minor proportions (Best and Harada 1985). The other species feed extensively on ants
and termites (Wetzel 1982), but there seems to be wide
variation in the relative proportions each occupies in
the diet in different localities. This fact is particularly
well documented for the giant anteater (Redford 1985b,
1986; Drumond 1992; Medri et al. 2003; table 24.1).
Some studies have found that the giant anteater
consumes almost exclusively ants (Montgomery 1979;
Lubin 1983; Redford 1985b), or a substantially greater
259
Table 24.1. Proportions of Termites and Ants in the Diet of Anteaters

Species and location
% ants
% termites
Giant anteater

Amap, Brazil

Braslia, Brazil

Cear, Brazil

Emas National Park, Gois, Brazil

Pantanal, Mato Grosso do Sul, Brazil

Serra da Canastra National Park, Minas

Gerais, Brazil

Serra da Canastra National Park, Minas

Gerais, Brazil

Paraguay

Llanos, Venezuela

Llanos, Venezuela
Silky anteater

Panama

Manaus, Brazil
Northern tamandua

Panama
Southern tamandua

Llanos, Venezuela
Sample size
(sample type)
Source
0
15
100
11
81
88
100
85
0
89
19
12
*
*
* (sc)
691 (fs)
50 (fs)
58 (fs)
3
3
3
3
6
4
45
55
359 (fs)
100
99
96
0
1
4
*
* (fs)
28 (fp,sc)
3
3
2
100
97
0
0
42 (fp,sc)
4 (fp)
2
1
33
77
29 (fp,sc)
49
51
19 (fp,sc)
Sources: (1) Best and Harada 1985, (2) Montgomery 1985b, (3) Redford 1985b, 1986, (4) Shaw et al. 1985, (5) Drumond 1992, (6) Medri et
al. 2003.
Note: Sample type refers to the number of feeding sites (fs), fecal pellets (fp), or stomach contents (sc) examined; * = not reported.
proportion of ants (Shaw et al. 1985; Medri et al. 2003).

Other studies have revealed termites to be the predominant prey, or even the only type of food consumed
(Redford 1985b, 1986; Drumond 1992). This variation is
likely related to the availability of prey in the environment, but this has not been tested yet.
In the Pantanal wetlands of Mato Grosso do Sul,
where the soils are predominantly sandy, giant anteaters consume primarily ants, which they find mainly on
the shores of lakes and in native pasturelands (Medri et
al. 2003). In this part of the Pantanal ground-dwelling
termites are scarce because of annual floods and they
occur only in flood-free areas (Medri et al. 2003). Tree
termites are abundant, but may not be easily accessible
to giant anteaters even though they do have some treeclimbing abilities (Rummel 1988; Young et al. 2003).
The situation is similar in the Venezuelan llanos (grassy
plains) where, once again, giant anteaters in these seasonally flooded areas consume few termites and have
a much greater proportion of ants in the diet (Montgomery and Lubin 1977; Montgomery 1985b). In Emas
National Park, on the other hand, the consumption of
termites is much higher (ca. 88%, Redford 1985b, 1986).
This area consists of grasslands with a high density of
termite mounds, especially Cornitermes cumulans, pro-
viding a plentiful source of food (C. cumulans or other

termite tenants in the nests of this species).
The opportunistic nature of the giant anteaters choice
of prey has also been evidenced in studies conducted at
different times in the same area. In Serra da Canastra
National Park, Shaw et al. (1985) reported the diet was
87% ants and 12% termites. Working at the same site,
however, Drumond (1992) found a larger proportion
of termites (55%) than of ants (45%). This variation
was likely related to the study period: Shaw et al. conducted their study in just two months (February and
March), while Drumonds study extended from March
to October. In fact, Drumonds findings for March were
very similar to those of Shaw et al., with 82% of ants
recorded in the diet that month. This suggests that seasonality affected diet, with more ants being consumed
in the rainy season (October to March) and more termites during the dry season (May to September). At
Emas National Park, a larger proportion of termites in
the diet was also recorded during the dry season (Redford 1985b). Winged nymphs are highly nutritious, with
a high percentage of fat (Redford and Dorea 1984). Termite colonies begin preparing winged nymphs months
before the actual flight, which occurs at the onset of the
rainy season (Brando 1996). Therefore, the growth of
260
winged nymphs occurs during the dry season, coinciding with the period of greatest consumption of termites
in both the Cerrado and the Pantanal (Redford 1985b;
Shaw et al. 1985; Drumond 1992; Medri et al. 2003).
Interestingly, both southern and northern tamanduas
display longer attack times on termite nests when they
contain winged forms (Lubin and Montgomery 1981).
Thus, termite nests are presumably under greater attack
during specific portions of the year, resulting in a seasonal dietary pattern.
Termites and ants are among the most common insects in the world, and are therefore an abundant and
easily found source of food, particularly in tropical regions. Yet, as stated above, few mammals specialize in
the consumption of these species. From a nutritional
standpoint, compared with other groups of terrestrial
invertebrates, nonwinged termites tend to contain
higher proportions of ash, similar amounts of water
and nitrogen, but lower proportions of fats (Redford
and Dorea 1984). However, these differences are small
and may not be sufficient to explain prey choices. Insectivorous species may consume their prey based mainly
on availability or other characteristics of the preys biology (Redford and Dorea 1984).
One important characteristic of ant and termite
prey concerns the many defense mechanisms they have
evolved along their ancient evolutionary path. Such
mechanisms range from powerful mandibles, prongs,
and stingers, to chemical defenses (Lubin and Montgomery 1981; Lubin 1983; Redford 1985b), all of which
hamper the action of predators. This may be why so
few species have specialized in feeding on these social
insects. It also suggests that myrmecophagous predators represent the few species able to evolve means of
overcoming these defenses.
The soldiers of some termite genera have a structure
(the nasus) that can emit volatile terpenoids, substances
that irritate the mucosa of anteaters and other mammals and render the insects unpalatable (Lubin and
Montgomery 1981; Redford 1985b). Once a colony is
attacked, the soldiers rapidly group around the damaged area to protect it. Lubin and Montgomery (1981)
evaluated the behavior of Tamandua spp. when feeding on termites of the genus Nasutitermes. Nasutitermes
soldiers were present in greater numbers in nests and
less so on foraging trails and in the decomposing wood
on which the termites fed. Southern and northern tamanduas adjusted their behavior so as to avoid large
concentrations of soldiers, foraging on Nasutitermes
predominantly at sources farther away from the nest,
or attacking the nest only in the presence of more attractive food, such as fatty ants or nymphs (Lubin and
Montgomery 1981).
Another way of avoiding large concentrations of soldiers is to limit the time of each feeding bout, so that
when the density of soldiers is very high, the anteater
stops feeding at the site and looks for another nest to
exploit (Lubin and Montgomery 1981; Redford 1985b).
Feeding bout times of free-ranging giant anteaters averaged 38 seconds per ant or termite nest, after which
they moved on in search of another nest (n = 194 feeding bouts, Drumond 1992). Limiting the duration of
each feeding bout may also be an efficient foraging
strategy for species with little or no chemical defenses,
but which build stronger mounds. For example, in tests
in captivity, giant anteaters preferred termite species
without chemical defenses, even though species with
chemical defenses were most commonly consumed by
free-ranging anteaters (Redford 1985b). Termites lacking chemical defenses normally flee the region being
attacked by an anteater. Their strong nests ensure sufficient time for many of the termites to escape while
the anteater is busy attempting to create an opening.
As a result, the density of termites decreases rapidly to
a point where it is no longer worth foraging at the site,
so the anteater goes off in search of another colony. In
captivity, without the protection of a strong nest, these
termites are easy prey for anteaters, which is why they
are among the preferred species. In contrast, when portions of termite nests from species that possess chemical defenses are offered to anteaters in captivity, soldiers
swarm around the damaged areas of the nest so that
practically only soldiers of low palatability are present
on the surface. Hence, the anteaters avoid these species
(Redford 1985b).
If free-ranging anteaters contend with prey defense
mechanisms by sampling large numbers of colonies,
this may result in their having to travel large distances
daily to obtain sufficient food. In fact, the foraging
strategy of anteaters has been compared with that of
some herbivores, which prey on immobile sources that
can regenerate (e.g., trees, Drumond 1992).
Although the giant anteater is considered one of the
mammals most specialized for feeding on termites and
ants (Redford 1985b), it has been reported to feed on
other insects as well, including beetle larvae (Silveira
1969) and European bees (Apis mellifera) that have built
hives inside termite mounds. In foraging on the latter,
anteaters probably consume both insects and honey
(Miranda et al. 2003). Bee products are rich in nutri-
ents, so anteaters may preferentially seek out termite

mounds that have associated beehives in order to consume bees and their products (Miranda et al. 2003).
Home-range size and overlap

Anteaters are usually solitary, but can be found in pairs
during the mating season or when the mother is still
carrying her young. There is still a paucity of data on
patterns of movement and home range. Silveira (1969)
made a theoretical estimate of the giant anteaters
home range based exclusively on the species ecological characteristics, and concluded that home-range size
would be at least 9 km2. He did not, however, collect
field data to support his estimate. At Hato Masaguaral,
in the llanos of central Venezuela, the giant anteaters
home range was estimated at about 25 km2 (table 24.2).
In Brazil, estimated ranges are highly variable. Studies conducted by conventional VHF radiotelemetry at
two sites in Brazil, the Pantanal of Nhecolndia (Medri
and Mouro 2005a) and Emas National Park (Miranda
2004), found average home ranges of 5.711.9 km2 (table 24.2). The smallest home ranges ever reported for
this species were also based on VHF radiotelemetry,
from Serra da Canastra National Park (Shaw et al. 1987;
table 24.2). Unfortunately, these authors only reported
261
an overall average; the range of values and sizes of individual home ranges were not presented.
Mouro and Medri (2002) developed a new, lowcost way of using a modified GPS radio collar (a conventional GPS receiver connected to an external power
source and fixed in a standard radio collar) to monitor,
over short time periods, medium-sized to large vertebrates living in open areas. Medri and Mouro (2005a)
tested the device on two giant anteaters, a male and a
female, who had previously been monitored by VHF
radiotelemetry. The home range of the male during 9
days of recordings was greater than that estimated from
locations obtained by VHF radiotelemetry during 252
days of monitoring. The females GPS home range was
smaller, albeit close to the value obtained by VHF radiotelemetry (figure 24.1).
Camilo-Alves (2003) also used these modified GPS
radio collars to monitor giant anteaters. In the Pantanal of Nhecolndia, 10 animals monitored for 412 days
had average ranges of 2.8 km2 (SD = 1.1 km2) per day,
with no difference between the sexes.
The use of the modified GPS radio collar allowed
home range estimates to be made in about 7 days that
were similar to those obtained in 612 months of monitoring by conventional VHF telemetry. The graphic
analysis in figure 24.1 suggests that about 10 days of
Table 24.2. Home Ranges of Giant Anteaters, Myrmecophaga Tridactyla, in Five Independent Studies

Study sitea
Monitoring
methodb
Masaguaral
VHF

Canastra
VHF

Pantanal
GPS

Emas
VHF

GPS

Pantanal
VHF

GPS

Sampling dates
(number of days monitored)
nd
Sexe
AprJun 1974 and FebApr 1975
(~120)
Nov 1979Jun 1980
(~200)
MarSep 2002
(412)
Aug 2001Oct 2002
(77420)
JanApr 2003
(m = 26; f = 7)
MarDec 2001
(m = 165252; f = 259)
MarDec 2001
(m = 9; f = 9)
Home range (km2)c

Mean
Range
Source
25
4
4
7
3
9
3
4
2
4
1
1
1
m
f
m
f
m
f
m
f
m
f
m
f
2.7
3.7
5.3
3.6
10.8
6.9
0.8
2.7
5.7
11.9
7.3
9.5
-
1.49.3
1.67.5
3.025.5
3.98.7
0.11.7
1.63.8
4.07.5
-
2
3
4
Sources:(1) Montgomery and Lubin 1977, (2) Shaw et al. 1987, (3) Camilo-Alves 2003, (4) Miranda 2004, (5) Medri and Mouro 2005a.
a. Masaguaral = Hato Masaguaral (834'N, 6735'W), Venezuela; Canastra = Serra da Canastra National Park (2000'S, 4615'W), Minas
Gerais, Brazil; Pantanal = Nhumirim ranch (1859'S, 5639'W), Mato Grosso do Sul, Brazil; Emas = Emas National Park (1815'S,
5253'W), Gois, Brazil.
b. VHS = conventional radiotelemetry; GPS = modified GPS radio collar (Mouro and Medri 2002).
c. Estimated by the 100% minimum convex polygon method (Mohr 1947).
d. Number of individuals
e. m = male, f = female, * = not reported.
262
Figure 24.1. Home ranges of giant anteaters,

Myrmecophaga tridactyla, (estimated by the
100% minimum convex polygon method, Mohr
1947) as a function of monitoring time using
modified GPS radio collars (circles) or conventional radiotelemetry (rhombi). Solid black
rhombi indicate that the asymptote of the home
range was reached. Dashed lines connect data
from the same individuals monitored by both
methods. Data taken from Camilo-Alves (2003),
Miranda (2004), and Medri and Mouro (2005a).
intensive GPS monitoring (recording locations at 10minute intervals) are required to obtain a home range
of the same dimensions as the asymptotic home range
estimated from daily VHF locations.
Differences in home-range sizes of giant anteaters
may be attributed not only to different study protocols,
but also to the degree of anthropic interference and to
characteristics intrinsic to each site, such as type of habitat, temperature, population density, and availability of
resources.
All the studies of giant anteater home ranges have
been conducted in regions with savannalike vegetation
containing a few forest fragments. Home-range sizes
of giant anteaters in predominantly forested environments are still unknown. The llanos of Venezuela and
the Pantanal of Nhecolndia are similar insofar as they
are seasonally flooded. Flooding may be a factor that alters the availability of resources for giant anteaters, and
this may be reflected in increasing home-range size as
the animals search for resources. So far, studies of giant
anteater home ranges have not included quantitative
data on the availability of food. However, as mentioned
in the discussion of diet composition above, we have
found that seasonally flooded environments, such as the
Venezuelan llanos and the Brazilian Pantanal, contain
fewer ground-level termite nests than do nonfloodable
areas like Serra da Canastra and Emas National Parks.
Nevertheless, despite the abundance of termite mounds
in Emas, giant anteaters had larger home ranges there
than in Serra da Canastra (table 24.2).
Estimates of giant anteater population density were

obtained in Brazil at Serra da Canastra (12 animals/
km2, Shaw et al. 1985, and 1.3 animals/km2, Shaw et al.
1987), the Pantanal (0.040.15 animals/km2, Coutinho
et al. 1997) and Emas (0.100.40 animals/km2, Miranda
2004), although the methods were not directly comparable. The high density of giant anteaters in Serra da
Canastra may be a consequence of the wide availability of favorable resources such as termite and ant nests,
combined with the absence of predators, such as the
jaguar Panthera onca, and other negative effects such
as hunting and road kills (Miranda 2004). It should be
noted that jaguars are also absent from large areas of
the Pantanal of Nhecolndia, although this predator
does occur with giant anteaters in other parts of the
Pantanal biome.
One might assume that the high density of giant anteaters at Serra da Canastra was responsible for the very
small home ranges recorded there. However, the studies of Shaw et al. (1987), Camilo-Alves (2003), Miranda
(2004), and Medri and Mouro (2005a) indicate that
giant anteater home ranges can overlap considerably.
This would argue against density as the predominant
factor producing small ranges. For example, one giant
anteater was found to have 91.4% of his home range
completely overlapped by three others in the Pantanal of Nhecolndia (Medri and Mouro 2005a). In the
same region, Camilo-Alves (2003) found 91% of the
home range of one male overlapped by another, with a
total of 6.5 km2 shared between the two males. In Emas
National Park, the home range of each pair of 11 monitored individuals showed an overlap of about 20.3% (SD
= 24.4%, Miranda 2004). Giant anteaters thus seem to
tolerate their conspecific neighbors, although agonistic
encounters have been observed at Serra da Canastra
(Shaw et al. 1987; see below for more about this subject)
and Nhecolndia (Fabiana Rocha personal communication).
Interspecific overlap between giant anteater and
southern tamandua home ranges has been reported in
Venezuela (Montgomery 1985b) and also occurs frequently in the Pantanal; agonistic encounters between
these species were never observed. Although the giant
anteater is not a good tree climber, Tamandua and Cyclopes anteaters are skillful climbers equipped with a
prehensile tail. Their excellent tree climbing skills enable southern, northern, and silky anteaters to exploit
other sources of food and shelter not available to the
giant anteater and may facilitate interspecific overlap in
home ranges.
In the Venezuelan llanos, Montgomery and Lubin
(1977) observed two southern tamanduas whose home
ranges averaged 3.75 km2. In this same study, but during
different monitoring intervals, the authors determined
an average home range of 0.65 km2 for two northern
tamanduas on Barro Colorado Island, Panama. At
the same location in Panama, Montgomery (1985b)
reported a home range of 0.25 km2 for the northern
tamandua. He attributed the considerable interspecific
difference in home-range size to the fact that the population of southern tamanduas under study lived in palm
savanna, while the northern tamanduas lived in moist
lowland forest. Because the palm savanna has little
vertical structure and a lower diversity of plant species
than the lowland forest, it probably contains a lower diversity of ants and termites, which are the primary food
source. Thus, the southern tamandua would probably
need to roam farther to obtain each days ration of food.
The only Brazilian study of the southern tamanduas
home range, conducted in the Serra da Mesa, Gois,
showed an average home range of 1 km2 for 7 individuals monitored after being translocated (Rodrigues et al.
2001). This study was carried out from December 1996
to February 1998, but most of the animals were monitored for only one or two months, so their home ranges
may have been larger than those recorded.
The silky anteater is the smallest, most nocturnal, and
most arboreal of all anteaters (Meritt 1971a). The silky
anteaters home range on Barro Colorado Island was
found to be 0.05 km2 and individuals were completely
263
arboreal (Montgomery 1985b). At the same locale, the

home range of an adult male was 0.11 km2, and this area
was overlapped by the home ranges of two adult females, although there was no overlap with an adjacent
adult male. The average home range of four adult females was smaller (0.028 km2), and three females were
able to live within the home range of an adult male,
though there was no overlap between home ranges of
adjacent females (Montgomery 1985b). Home-range
overlap between silky anteaters and northern tamanduas was also found in Panama. Sometimes, northern
tamanduas were observed feeding in trees where silky
anteaters were radio-located (Montgomery 1985b).
During the monitoring of translocated silky anteaters
in Palmas, Tocantins, Brazil, home ranges for two individuals were found to be 0.008 and 0.1 km2 (Silva et al.
unpublished data), similar to the findings of Montgomery (1985b).
The estimates for anteater home-range sizes can be
considered minimum values for several reasons. First,
many studies do not mention cumulative home range
(e.g., Montgomery and Lubin 1977; Montgomery 1985b;
Shaw et al. 1987). Second, some studies do not reach
the asymptote on the cumulative home-range curve, indicating that the animals were not monitored for long
enough to stabilize home-range size (Medri and Mouro
2005a). Third, in some cases most of the animals were
monitored for short periods of time (Rodrigues et al.
2001; Camilo-Alves 2003; Silva et al. unpublished data).
Finally, a comparison of studies involving home-range
estimates obtained by conventional radiotelemetry
with studies in which estimates were obtained through
intensive data collection protocols indicated that conventional radiotelemetry tends to underestimate home
ranges (Ballard et al. 1998). In the absence of studies
reporting more realistic anteater home ranges, the values we report here must serve as the basis for current
management and conservation decisions about these
little-known species. However, managers should be advised that underestimated home ranges, when used to
obtain data on densities, could render higher densities
than actually occur.
Timing of activity, habitat use, and rest sites

Anteaters can be found in various habitats ranging
from tropical rainforests to dry savannas, dry forests
and thorn scrub. These habitats can be seasonally very
dry (e.g., savanna) or very moist (some forests), and
can occur at low to moderate altitudes. As described
264
above, in many areas species coexist, sharing parts of

their habitat (Montgomery 1985b). Anteaters may use
several habitats within their home ranges and some
studies have detected divergent preferences for resting
versus foraging sites.
The habitat distribution of the southern tamandua in
Venezuela included thorn forests (44%), evergreen forests (30%), savanna or cropland (21%), and deciduous
forests (5%, Handley 1976). Giant anteaters and southern tamanduas studied in the llanos of central Venezuela lived mainly in seasonally flooded palm savanna, but
used higher areas of sand dunes as well, particularly in
the wet season (Montgomery 1985b). Within the llanos
both species utilized several habitats, for example, large
areas of grassland with scattered palm trees (Copernicia tectorum) and smaller areas of a low, single-layered,
closed canopy of fig trees (Ficus spp.), palms, and a few
other tree species (Montgomery and Lubin 1977).
The northern tamandua occurs at low elevations in
a variety of habitats from gallery forests to multistratal
tropical evergreen forests (Eisenberg and Redford
1999). On Barro Colorado Island, there are reports of
northern tamanduas in tropical monsoon forest with
a closed canopy, but with a lower stratum consisting
of lianas and interlaced tree crowns that form arboreal
pathways (Montgomery and Lubin 1977). The highly arboreal silky anteater occurs in the same area in Panama
(Montgomery 1985c). It is generally confined to moist
multistratal tropical evergreen forests and is absent
from savanna areas (Eisenberg and Redford 1999).
Little is known about habitat use by anteaters; the
most detailed studies have been conducted in Brazil
on giant anteaters. Giant anteaters monitored by VHF
radiotelemetry in the Pantanal of Nhecolndia generally used habitat types in the same frequency as they
occurred in each home range. The study area covered a
mosaic of vegetation comprising 30.1% scrub savanna,
25.2% grasslands, 24.6% forest patches, 14.5% savanna,
and 5.6% lakes. However, there was some evidence of
habitat preferences. Giant anteaters rested in forest
patches and savanna more often than expected and
were more active in grasslands. In the same study, two
giant anteaters monitored with modified GPS radio collars preferred savanna and grasslands and avoided lakes
and scrub savanna, although the pattern of habitat use
varied between individuals (Medri and Mouro 2005a).
In Brazils Cerrado biome, some of the giant anteaters
monitored by GPS spent much of their time roaming
in soybean plantations, occasionally visiting wet fields
and meadows near drainage ditches; others spent more
time in the grasslands or wet fields and meadows of

Emas National Park, and a third group stayed exclusively in the grasslands of the park (Miranda 2004).
Further studies focusing on anteater habitat requirements would provide crucial information to underpin
the development of management strategies for this species.
Anteaters are usually nocturnal, although they occasionally engage in diurnal activity. In some regions and
during some periods, they display much more diurnal
than nocturnal activity. The exception to this finding
is the silky anteater, which occurs only in Neotropical
canopy forests, roaming through thick masses of lianas
at night.
Few studies have examined activity and habitat use in
the Vermilingua. Studies conducted on giant anteaters
in different regions have reported various differences: in
Brazils Cerrado, individuals rested more often in scrub
savanna, within a home range that included gallery forests and grasslands (Shaw et al. 1987). In the Pantanal
wetlands, animals preferred to rest in forest patches
and savanna, displaying greater activity in grasslands
(Medri and Mouro 2005a). Other studies carried out
in the same area of the Pantanal reported that, in some
instances, giant anteaters showed a preference for resting in open areas (Camilo-Alves and Mouro 2006).
Shaw et al. (1987) reported that even during cool and
rainy periods, anteaters were commonly found bedding
down in open grassland.
With regard to daily activity, giant anteaters engaged
in both nocturnal (Montgomery and Lubin 1977) and
diurnal activity (Shaw et al. 1987), even displaying
changes in daily activity in the same region from one
season to the next (Camilo-Alves and Mouro 2006).
Southern and northern tamanduas are two allopatric species differing little morphologically, behaviorally, or ecologically, and this is reflected in their activity
patterns. When not active, southern tamanduas rest
in hollow trees, in the burrows of other animals, or in
other natural shelters (Emmons and Feer 1990). Southern tamanduas in Venezuela were nocturnal and rested
in trees and the crowns of palm trees (Montgomery
1985b). In Brazils Cerrado, southern tamanduas were
also nocturnal (with just one observation of activity
during daylight), but the animals always rested on the
ground, usually (90%, n = 19) inside armadillo burrows
(Euphractus sexcinctus, Rodrigues and Marinho-Filho
2003). Other studies have reported nocturnal, crepuscular, or diurnal activity for both southern and northern tamanduas, with considerable variation between
individuals. In studies conducted by Montgomery and

Lubin (1977), both species exhibited individual differences in the timing of activity, in their use of different
microhabitats, and in their degree of arboreality. Another study conducted in the same areas (Montgomery
1985b) reported that individual differences were more
pronounced in the northern than the southern tamandua. It is interesting to note that differences in daily
activity and habitat use occurred not only between regions or at different times within the same region, but
also in the same region, at the same time, among different individuals (Montgomery 1985b).
Thermoregulation
Anteaters, sloths, and armadillos share in common,
along with structural features, lower metabolic rates
than would be expected for their body size, according
to Kleibers rule (Kleiber 1932 as cited in McNab 1984).
This peculiar metabolism permits foraging on exclusively low caloric prey. However, a low energy intake
limits the expenditure of energy required to maintain
high basal rates of metabolism. As a consequence, anteaters have a low body temperature, which may vary
with ambient temperature. Although the normal body
temperature of a mammal lies between 36 and 38C,
that of giant and silky anteaters averages 33C, while the
average for southern and northern tamanduas is 34.5C
(McNab 1984). Furthermore, the body temperature of
giant, southern, and northern tamanduas may vary
by about 4C. The silky anteaters body temperature
is even more variable, dropping as low as 16C when
ambient temperatures reached 3C (McNab 1984).
These energy constraints, along with diminishing food
resources in temperate environments, are believed to
restrict Xenarthra in their distribution to tropical environments (McNab 1985). However, living in the tropics
is no guarantee that ambient temperatures will always
lie within the limits of thermoneutrality. When facing
temperatures below their lower limit of thermoneutrality, mammals with conventional metabolic rates increase their metabolism, along with voluntary behavioral responses. Because anteaters have some difficulty
in maintaining a constant body temperature, a logical
prediction is that they should display marked behavioral responses to changes in ambient temperature, thus
avoiding heat loss or overheating. This all leads to the
question: is variation in activity and habitat use a response to climatic variation?
One study that attempted to answer this question
265
examined giant anteaters in the Pantanal of Nhecolndia (Camilo-Alves and Mouro 2006). The weather in
Brazils Pantanal wetlands is partly humid tropical, with
an average annual temperature of 25C. Giant anteaters
monitored here were usually nocturnal, engaging in activity throughout the night and into the early morning
until the ambient temperature rose, resting thereafter in
sheltered habitats. This behavior enabled them to avoid
overheating by reducing both their exposure to the suns
rays and their activity during the hottest hours of the
day. However, on cool nights during the dry season, giant anteaters began their activity earlier and ended before dawn. As the days became colder, anteaters began
and ended their activity progressively earlier and preferred resting in more open areas, though their habitat
preferences for foraging did not change. On the coldest winter days (with daily mean temperatures below
15C), they were sometimes even exclusively diurnal.
This behavior benefits thermoregulation in two ways.
Reduced nocturnal activity helps preserve body heat,
while increased diurnal activity and resting in open areas increases exposure to sunlight and, hence, absorption of solar radiation before and during activity. Similar seasonal changes in activity as a function of ambient
temperatures have been reported for sloths (Chiarello
this volume) and armadillos (Layne and Glover 1978,
1985).
Giant anteaters usually sleep with their bushy tails
covering their bodies. But on a particularly cool sunny
morning in the Pantanal wetlands, when the ambient temperature was about 17C, an animal was found
sleeping with its tail stretched out flat on the ground,
exposing the full length of its body to the suns rays,
suggesting that the giant anteater can use its tail to aid
its thermoregulation (Medri and Mouro 2005b).
Many other mammals display thermoregulatory
behavior, reducing their activity during the hottest or
coldest hours of the day and protecting themselves in
suitable habitats. However, the interesting thing about
anteaters is that they change their daily activity and
habitat use considerably in order to regulate their body
temperature. Giant anteaters not only exhibited seasonal differences in timing of activity and choice of rest
sites, but also differences from one day to another, according to variations in daily mean temperatures.
Northern and southern tamanduas also exhibit considerable variability in their schedules of activity and
use of particular resting sites. Given that these species face much the same problems as giant anteaters, it
seems likely that this variation is due to the same rea-
266
sons outlined above. However, even though the same

arguments should apply, silky anteaters do not exhibit
the same pattern. Montgomery (1985b) reported that
silky anteaters in Panamanian forests displayed consistent patterns of behavior, without changes from day to
day, season to season, or individual to individual. Thus,
there may be other constraints that prevent silky anteaters from behaving like the other species.
All four species of anteaters undoubtedly have multiple constraints on activity patterns and habitat preferences, such as prey habits and distribution, and avoidance of predation and/or competition. However, giant
anteaters have few predators (jaguars and pumas primarily, Silveira 1969, 2004) and can prey on several species of ants and termites, so these are probably not very
restrictive constraints. A more important constraint
may be high levels of anthropic activity, which may
trigger more nocturnal activity. Southern and northern tamanduas probably have more predators, but they
can also prey on several species of ants and termites. In
contrast, silky anteaters are conservative in food preferences and probably have many more predators because of their very small size. In view of this, it seems
reasonable to wonder whether the silky anteaters small
size, combined with the other aforementioned factors,
restricts its thermoregulatory behavior, habitat choices,
and activity patterns. Silky anteaters face both physiological and behavioral constraints in keeping their
body temperature constant. Their temperature can vary
far more than that of other anteaters and their body
temperature remains stable in relation to that of the environment only because of their low thermal conductance (McNab 1984). Perhaps because of this the silky
anteater occurs only in regions close to the equator,
where temperatures do not vary as significantly as they
do in the areas where other anteater species are found.
Reproduction and development

All anteater species are normally solitary except during the mating season and in the case of mother-young
pairs (Shaw et al. 1987). Female anteaters usually produce only one young at a time, and parental care is
highly developed (Eisenberg and Redford 1999). The
offspring rides on its mothers back during much of its
first year of life. Because the mother cannot pick up her
young, the juvenile must climb onto her back by itself
(Emmons and Feer 1990).
Among silky anteaters, the male sometimes carries the baby on his back. A mother does not carry
her young during nocturnal foraging expeditions; instead she leaves it in the tree where she spent the day.
A mother has been observed placing her baby on a nest
of dry leaves in a hole in a tree trunk (Nowak 1999).
Mothers leave their young alone for about eight hours
each night (Montgomery 1983). Southern and northern
tamanduas may place their offspring on a branch near a
preferred feeding location, or leave them for short periods in a leafy nest.
The courting behavior of free-ranging giant anteaters has not been well documented. Several observations
of agonistic interactions, generally between males, have
been reported (Shaw et al. 1985, 1987; Fabiana Rocha
personal communication). These interactions varied
from slow circling to chases and serious fighting; injuries to one or both males can occur (Nowak 1999).
Although territoriality in giant anteaters might seem
unlikely given the considerable home-range overlap
that can occur (see above), agonistic encounters between males may indicate that giant anteaters defend
their territories to some extent.
Reproductive status may be communicated chemically. The giant anteater has a well-developed olfactory
system. Its olfactory discrimination is about 40 times
more accurate than a humans and it can detect and
respond to the odor of its own saliva (McAdam and
Way 1967). Giant anteaters and tamanduas can produce strong-smelling secretions from their anal glands
(Naples 2001). These secretions could be used to mark
paths and maybe advertise the presence, status, or sexual condition of the marking individual. Silky anteaters
have a facial gland, but its function is unknown (Naples
2001).
For giant anteaters, breeding can occur throughout
the year, both in captivity and in the wild (Bartmann
1994 as cited in Patzl et al. 1998). There may be anatomical (Bartmann et al. 1991) or behavioral cues that indicate readiness to breed. All male anteaters have intraabdominal testicles (cryptorchids). Because their body
temperature is low, they do not require extra-abdominal
(exposed) testicles to avoid damage to sperm. The penis is short and boneless (no baculum) and is equipped
with a glans. When sexually inactive, the dimensions
of the genital organs shrink. Patzl et al. (1998) stated
that female giant anteaters do not display any specific
behavior during estrus. However, Shaw et al. (1987)
reported vaginal discharge associated with estrus and
that during courtship the female raises her tail and the
male follows closely behind, occasionally pawing and
sniffing her. The pair continue to forage while courting,
even feeding together on the same insect nest (Shaw et

al. 1987). When mating, the male (usually larger than
the female) crouches over the female, sometimes pressing her also laterally to the soil as she lies on her side.
Copulation occurs several times over approximately
three days (Bartmann 1983).
Gestation in giant anteaters lasts 180190 days (Bartmann 1983; Shaw et al. 1987). The decidual, discoid,
and lobulated placenta is hemochorial (Silveira 1968;
Benirschke this volume). The female delivers her young
in the bipedal position (Nill 1907 as cited in Bartmann
1983). Immediately after birth, the newborn climbs
through its mothers fur and she licks the baby clean.
Nursing lasts about six months (Bartmann 1983).
In the initial months of their life, young giant anteaters may emit a trilling sound when abandoned by the
mother (Bartmann 1983). The youngster is carried on
the mothers back for up to a year and remains with
the mother until she again becomes pregnant (Nowak
1999). Thereafter, at more than a year old, the young
leaves its mother to live alone (Silveira 1969). Sexual
maturity is attained at 2.54 years (Nowak 1999).
Reproduction of giant anteaters is common in
zoos, but newborn mortality is high, and the successful breeding rate in captivity is only 30% (Bartmann
1983; Superina et al. this volume). Newborn mortality
is caused by stillbirth, incorrect maternal behavior, or,
quite often, males injuring or killing offspring immediately after birth (Patzl et al. 1998). At Dortmund Park
zoo, there is a record of a newborn weighing 1.1 kg at
birth. This animal, which was created by artificial insemination, weighed 2.2 kg at the age of one month, 11
kg at eight months, and 31 kg at 12 months (Bartmann
1985). In Brazils Serra da Canastra National Park, yearling giant anteaters weigh approximately half as much
as adults, indicating that growth rates in the wild are
slightly lower than in captivity (Shaw et al. 1987). Yearlings display rapid growth rates during the second half
of their second year, and take at least three years to
reach full size in the wild (Shaw et al. 1987).
Silveira (1968) reports that southern tamanduas
mate mostly in tree hollows and other natural depressions. Females emit low moans when in estrus (Silveira
1968). Mating takes place in the dorsal-ventral position,
usually at dusk or night, though it may also occur during the day. The penis is short, so the male must stand
in an almost vertical position with his back curved to
reach the vulva. Sexual intercourse may take more than
30 minutes, during which the females behavior is passive.
267
The southern tamanduas sexual cycle appears to be

polyestrous and uniparous, with one young per delivery and two births per year, each involving a 35-week
rest period. Gestation lasts 130150 days (Silveira 1968),
or 160 days (Eisenberg and Redford 1999). Like other
anteaters, after birth young southern tamanduas ride
on their mothers back for a long time, accompanying
their mothers almost until adulthood (Silveira 1968).
Gestation in northern tamanduas is either 130150
days or 160190 days, and the young remains with its
mother for about one year before dispersing (Nowak
1999). It appears the low metabolism of xenarthrans
slows embryonic development, thus increasing gestation length (Silveira 1968).
Estrus in the silky anteater lasts from December to
January (Grass 1955 as cited in Silveira 1968), and gestation lasts 120150 days (Messias-Costa et al. 2001).
The young is fed semidigested ants that are regurgitated
by both parents. Moreover, the offspring may be carried by either parent (Nowak 1999). Little else is known
about the reproductive biology of this species.
Anteater conservation
General issues regarding anteater conservation are
covered in this volume in the chapters by Aguiar and
Fonseca and Noss et al. Here, we focus on some of our
own recent work.
The deterioration and reduction of suitable habitat
are believed to be main contributing factors for the decline of giant anteater populations (Fonseca et al. 1999).
The resulting habitat fragmentation can isolate populations, resulting in low genetic variability. The type
of matrix and extent of habitat fragments determine
whether giant anteaters can use this matrix. Miranda
(2004) showed that giant anteaters living on the edges
of Emas National Park foraged in soybean and corn
fields only if these plantations contained natural habitats (wet grasslands). Soybean and other crops can be
a permeable matrix when natural habitats are present.
Pasturelands can maintain some of the characteristics
of the original habitat, including large numbers of termite mounds. Pasturelands are probably suitable habitats for giant anteaters, or at least offer a more permeable matrix than do monocultures. On the other hand,
Camilo-Alves and Mouro (2006) have found that, although giant anteaters are usually associated with open
vegetation, their thermoregulatory needs still require
covered habitats. This is particularly critical during periods of extreme temperatures, because of their limited
268
thermoregulatory abilities (see above). Therefore, any

matrix must contain habitats that provide sufficient
shelter.
In Serra da Canastra National Park, the giant anteater population has adapted to a habitat that is highly
disturbed by grazing and prescribed burning (Shaw
et al. 1987). However, at other sites fire can be a major
threat. For example, the inability to move quickly probably makes giant anteaters more susceptible to brush
fires. In Emas National Park, at least 340 individuals
(representing about two-thirds of the population) died
in 1994 as a result of brush fires (Silveira et al. 1999).
This is of great concern genetically, because periodic
burning may repeatedly reduce the population to low
numbers. These successive population bottlenecks
might then lead to an eventual loss of evolutionary potential.
Although protected by law, there can be no doubt
that giant anteaters will survive only in protected areas.
Most giant anteaters living in zoos were captured in the
wild. As stated earlier, reproductive rates in captivity
are low. Reproduction and survival rates of young in
captivity must be improved through better management techniques and greater knowledge of reproductive physiology. Some obvious areas of research include
means of diagnosing estrus, development of methods
for artificial insemination and semen conservation, and
the artificial creation of young (Bartmann 1985).
What can be done to conserve anteaters? The first
step is to conduct more in-depth studies on their biology, covering all the topics reviewed in this chap-
ter as well as other aspects of behavior. Based on this

knowledge, new protected areas must be created in the
regions best suited for the conservation of these species. The creation of ecological galleries connecting
these protected areas to other fragments will be crucial
in preventing populations from becoming genetically
isolated. Programs for breeding anteaters in captivity
may offer a viable alternative for the reintroduction of
species into areas from which they have disappeared
and to replenish zoo stocks without removing animals
from the wild. In areas where anteaters experience
pressure from subsistence hunting, the implementation of sustainable development alternatives for local
populations is needed. Environmental education programs are essential to prevent the deliberate killing of
anteaters on roads and in farming communities. While
all of these initiatives may help, it remains true that the
main causal factor for the disappearance of anteaters
is loss of habitat. Therefore, the most critical measure
that must be adopted (and legally enforced) to ensure
the conservation of these species is to protect the areas
where they occur.
Acknowledgments
We thank the Conselho Nacional de Desenvolvimento
Cientfico e Tecnolgico (CNPq), Brazil (Process PELD
#520056/98-1) for financial support, Jim Loughry and
Sergio Vizcano for comments on the manuscript, Beatrice Allain for reviewing the English text, and Paula
Lara-Ruiz for translating the abstract to Spanish.
25
Sloth ecology
An overview of field studies
Adriano Garcia Chiarello
noche, aunque se registra una gran variacin individual

y poblacional. Un cuerpo creciente de evidencias indica
que Bradypus tiene un ritmo de actividad circadiano.
Resumen
En este captulo se revisa la ecologa de las seis especies vivientes de perezosos. A pesar de los profundos
estudios de campo iniciados a principios de la dcada
de 1970, varios aspectos de la biologa de estos animales
son an poco comprendidos, particularmente para los
perezosos de dos dedos (Choloepus spp.). Los perezosos
de tres dedos (Bradypus spp.) han sido ms estudiados,
aunque la mayora de la informacin proviene solamente de un manojo de trabajos de campo. Los datos
sobre alimentacin confirman la idea anterior de que
los perezosos de tres dedos son folvoros selectivos ya
que slo 5 a 8 especies vegetales constituyen ms del
50% de su dieta a lo largo del ao. Sin embargo, individuos de una localidad dada consumen diferentes
conjuntos de especies de lianas. Por lo tanto, las poblaciones de perezosos deben ser consideradas como
generalistas, como fue indicado originalmente por
Montgomery y Sunquist (1975, 1978). Los perezosos se
desplazan distancias cortas (2040 m/da) y tienen mbitos del hogar reducidos aunque la variacin individual es alta (0,810,8 ha/perezoso), lo que quizs refleja
mejor particularidades de la estructura del bosque que
diferencias interespecficas. La densidad poblacional
puede variar en dos rdenes de magnitud (0,088,46
perezosos/ha) y, aunque las comparaciones son difciles
debido a la inconsistencia de las tcnicas de muestreo
empleadas, los perezosos de tres dedos parecen ser ms
abundantes que los de dos dedos. Se sabe muy poco
acerca de la proporcin de sexos en Bradypus y an
menos en Choloepus, debido a que la determinacin de
sexo es ms difcil en stos ltimos. Al contrario que
Choloepus, cuyos individuos son estrictamente nocturnos, los de Bradypus son activos tanto de da como de
Resumo
Neste captulo a ecologia das seis espcies atuais de
preguias revisada. Embora estudos de campo aprofundados tenham comeado no incio da dcada de
1970, vrios aspectos da histria natural continuam
pouco conhecidos, particularmente a respeito das
preguias-de-dois-dedos (Choloepus spp). As preguiasde-trs-dedos (Bradypus spp.) so melhor conhecidas,
mas as informaes disponveis so provenientes de
um nmero restrito de estudos. Dados sobre dieta confirmam a suposio inicial de que as preguias-de-trsdedos so folhvoras seletivas, j que ~58 espcies de
plantas perfazem >50% de sua dieta anual. Entretanto,
os indivduos de uma mesma localidade consomem
conjuntos diferentes de espcies de rvores e lianas.
Conseqentemente, as populaes de preguias devem ser consideradas generalistas, como inicialmente
proposto por Montgomery e Sunquist (1975, 1978).
Preguias percorrem distncias curtas (2040 m/dia)
e possuem pequenas reas de vida, embora a variao
individual seja grande (0,810,8 ha/preguia), um resultado que deve refletir mais as particularidades da
estrutura e composio florestal do que possveis diferenas interespecficas. Densidades populacionais apresentam variao de duas ordens de magnitude (0,08
8,46 preguias/ha), mas comparaes so dificultadas
pela ausncia de tcnicas consistentes de amostragem.
Mesmo assim, as preguias-de-trs-dedos parecem ser
mais abundantes do que as de dois dedos. Pouco conhecido a respeito da razo sexual em Bradypus e menos
269
270
A. G. Chiarello
ainda em Choloepus tendo em vista que a determinao

do sexo mais difcil para esta ltima. Ao contrrio de
Choloepus, que estritamente noturna, Bradypus ativa
tanto durante o dia quanto noite, mas grande variao
encontrada entre indivduos, mesmo dentro de uma
mesma populao. Um conjunto crescente de evidncias indica que Bradypus tem um ritmo circadiano de
atividade.
Introduction
For centuries sloths have attracted the attention of scientists because of several interesting characteristics associated with their incredibly slow movements (Beebe
1926; Britton and Kline 1939; Lundy 1952). This interest
has led to the production of an impressive number of
papers on anatomy, morphology, physiology, neural
and circulatory systems, metabolism, and sensory systems, among others (Kredel 1928; Goffart 1971; McNab
1978; Gilmore et al. 2000). An excellent review of early
knowledge on these and related subjects is available
in Goffart (1971) and, more recently, in Gilmore et al.
(2000, this volume). However, as will become clear
throughout this chapter, there are far fewer studies on
ecological aspects of free-ranging sloths, particularly so
for the two-toed species (Choloepus spp.). Undoubtedly,
this is due mainly to difficulties associated with finding
and observing a sloth in the wild. As most readers of
this book probably know, sloths, both two- and threetoed (Bradypus spp.), are strictly arboreal mammals,
spending most of their time well hidden in the high
canopy of Neotropical forests. Detection is further hindered because sloths are solitary, have a relatively small
body mass, and are extremely silent during their rare
and careful movements. To complicate matters even
more, the two-toed sloths and at least some individual
three-toed sloths are nocturnal (see below). Even when
sloths can be found and observed, challenges still arise
because many activities, such as eating, are difficult to
record accurately even by a trained observer (personal
observations). These features explain why the overwhelming majority of studies utilized captive individuals (mainly Choloepus, which can be kept in captivity
for long periods, see Superina et al. this volume) or
examined aspects of sloths that did not require direct
observation of wild animals wandering freely in their
natural environment.
Extant sloth species are divided into two families,
Bradypodidae (genus Bradypus), which has three digits in the forelimbs, and Megalonychidae (= Choloepi-
dae; genus Choloepus), which has only two. These two

families differ in several other anatomical characters,
including teeth, cranium, and postcranium (Wetzel and
Avila-Pires 1980; Wetzel 1985a; see also McDonald and
De Iuliis this volume), but externally the number of
fingers is perhaps the most obvious, at least for nonexperts. Important analyses of phylogenetic relationships
between the extant families and between these and fossil forms can be found elsewhere (Webb 1985a; Wetzel
1985a; Delsuc et al. 2001; Barros et al. 2003; Delsuc and
Douzery this volume; Gaudin and McDonald this volume).
There are four extant species of three-toed sloths: B.
variegatus (Central and South America), B. tridactylus
(Amazon, northern South America), B. torquatus (Atlantic forest, southeast Brazil) and the recently discovered B. pygmaeus, the smallest of all sloths, restricted
to Isla Escudo off the Atlantic coast of Panama (Anderson and Handley 2001). Two of these are threatened with extinction: the maned sloth (B. torquatus)
is listed as endangered as a result of forest destruction,
fragmentation, and isolation of populations (Chiarello
et al. 2004); the pygmy sloth (B. pygmaeus) is listed as
critically endangered (Aguiar and Fonseca this volume) owing to its extremely small geographic distribution (restricted to the mangroves of the 4.3 km Isla
Escudo, Anderson and Handley 2001). The other extant
sloths are two Choloepus species: C. hoffmanni (Central America and northwestern South America) and C.
didactylus (Amazon, northern South America). Both
are widespread and locally common, and therefore not
endangered (Wetzel 1985a; see also Aguiar and Fonseca
this volume).
The objective of this chapter is to review information
on diet, ranging and movement patterns, and other aspects of the ecology of three- and two-toed sloths. Unfortunately, since the pioneering works of Mel Sunquist
and Gene Montgomery (Sunquist and Montgomery
1973; Montgomery and Sunquist 1975, 1978), few longterm and in-depth studies have been conducted (table
25.1). Worse still, all have focused on Bradypus species.
Thus, although significant advances in technology have
occurred since the early 1970s, particularly regarding
radiotelemetry, viewing and recording equipment (infrared binoculars, night-recording cameras and videos),
and related devices (lightweight, long-lasting, and powerful headlamps and torches, GPS, etc.), several aspects
of the ecology and behavior of free-ranging two-toed
sloths and, to a lesser extent, three-toed sloths, are still
unknown.
Sloth ecology: An overview of field studies
271
Table 25.1. Features of Sites Where In-depth Studies on the Ecology and Behavior of Sloths Have Been Conducted

Species
Locality
Altitude (m)
26171a
Annual
Annual
rainfall (cm) temperature (C)
260b
B. variegatus
Barro Colorado Island, Panama (BCI)
and C. hoffmanni
B. variegatus
Mamirau, Amazonia, Brazil (Mm)
80a
170

B. variegatus
Valena, Rio de Janeiro, Brazil (Va)
5411,296
130150
B. tridactylus
Manaus, Amazonia, Brazil (Ma)
44c
209d
e
B. torquatus
Santa Teresa, Espirito Santo, Brazil (ST)
600900
140187e

B. torquatus
Silva Jardim, Rio de Janeiro, Brazil (SJ)
20200
118f

Ref.
27.0b
1,2,3
28.2
(23.533.2)
26.0
26.6d
19.7e
(14.226.1)
23.5f
(19.027.0)
6
11
9
7,8,10
4,5
Sources: (1) Sunquist and Montgomery 1973, (2) Montgomery and Sunquist 1975, (3) Montgomery and Sunquist 1978, (4) Pinder 1985,
(5) Pinder 1993, (6) Queiroz 1995, (7) Chiarello 1998a, (8) Chiarello 1998b, (9) Carmo 2002, (10) Chiarello et al. 2004, (11) Consentino
2004.
a. UNEP/WCMC at http://sea.unep-wcmc.org/sites/pa/0309q.htm.
b. Leigh and Wright (1990).
c. http://www.explore.co.uk.
d. http://www.worldclimate.com.
e. Thomaz and Monteiro (1997) and Mendes and Padovan (2000).
f. http://www.inmet.gov.br.
The main field studies

Amazingly, in-depth studies of free-ranging sloths are
restricted to just three main sites: Barro Colorado Island (BCI) in Panama, the Mamirau Sustained Development Reserve in the flooded forest of the Brazilian
Amazon, and the Atlantic forest region of Santa Teresa
municipality, in southeastern Brazil (table 25.1). Other
studies of shorter duration have been conducted in a
few other places (table 25.1).
Diet and feeding strategies

Three of the four extant species of Bradypus have been
studied regarding their feeding habits (table 25.2). In
studies where food items were quantified, leaves constituted 9499%, fruit, about 1%, and flowers, another
1% of the diet of sloths. Other items can be occasionally important, though. Queiroz (1995), for example,
observed extensive consumption of wood from a dead
tree (Luehea cymulosa) by three B. variegatus individuals, particularly by a pregnant female just days before
giving birth.
All Bradypus species that have been studied, although occurring in widely separated places, seem to
have similar diets (table 25.2). There is a ubiquitous
preference for young leaves, although mature leaves can
be heavily consumed at times. This preference is prob-
ably due to the higher digestibility and nutritional quality of young leaves. Montgomery and Sunquist (1978)
observed that young leaves were digested more easily
than mature leaves in an experiment carried out with
gastric fistulas in B. variegatus. Previous studies have
shown that young leaves have lower proportions of cell
wall contents (cellulose and lignin), and higher concentrations of water, sugars, and crude proteins than
mature leaves (Milton 1979; Davies et al. 1988).
Unfortunately, there are no published studies on the
natural diet of two-toed species. Reports from captive
individuals indicate, however, that both species of Choloepus consume a much more varied diet than threetoed sloths, including vegetal matter such as fruits, legumes, and processed food (bread), and even animal
matter (raw and cooked meat, eggs, fish, etc., see Goffart 1971 and authors cited therein). This perhaps is the
main reason why Choloepus, unlike Bradypus, is successfully kept, sometimes for years, in captivity (Meritt
1976a, 1985a; Superina et al. this volume). These findings suggest that in nature two-toed sloths might feed
more heavily than three-toeds on fruits, flowers, and,
possibly, even bird eggs or nestlings. This prediction,
however, is not corroborated by their gut anatomy. The
stomach of Choloepus is voluminous and complex, allowing bacterial fermentation of vegetal matter (Grass
1955; Goffart 1971). Their low rate of metabolism and
low levels of activity (Sunquist and Montgomery 1973;
272
A. G. Chiarello
Table 25.2. Composition of Diets of Three Sloth Species Studied in the Wild

Species
Locality
B. variegatus
B. variegatus
B. variegatus
B. tridactylus
B. torquatus
B. torquatus
BCI
Mm
Va
Ma
ST
ST
Sampling
effort
(months)
17
12
9
7
14
36
Sampling % Diet composition

method
LE
FR
FL
OT
ad libituma
Focal/scan
Focal/?
Focal/scan
Focal/scan
Focal/scan

94.4

97.2
99.2

1.0

1.4
0.1

0.0

1.4
0.1

4.6c

% Young/mature Diet diversity

leaves
Species Families Ref.

54.5/37.0
Young > mature
59.7/37.5
68.2/7.2

31b
16
15
14
21
30d
14
13
9
11
12
> 12
Sources: Numbered references are the same as those in Table 25.1.

Notes: LE = leaves, FR = fruits, FL = flowers, OT = other items.
Abbreviations of localities are the same as in Table 25.1.
a. Based on frequency of encounters with sloths in feeding trees, but sloths were rarely seen feeding.
b. Includes 26 species observed to be eaten plus 5 species identified in stomach contents.
c. Includes dead wood of Luehea cymulosa.
d. Includes several morphospecies and three years of observations.
see below) further indicate that Choloepus, like threetoed sloths, must base their natural diet heavily on a
low-energy food base, that is, leaves of trees and lianas
(McNab 1978).
Three-toed sloths feed on a relatively small number
of plant species (tables 25.2 and 25.3). Available data
document consumption of 1431 plant species from
913 families (table 25.2). This includes both tree and
liana species, but tree species predominate in all studies thus far (Montgomey and Sunquist 1978; Queiroz
1995; Chiarello 1998a; Carmo 2002; Consentino 2004).
As summarized in table 25.3, the top five genera consumed (accounting for 58 plant species) make up
more than 50% of the diet. Diet composition varies,
with the genus Ficus (Moraceae) the only one present
among the top five in all four studies. Moraceae was the
plant family most consumed by sloths at BCI, in Amazon flooded forest, and also in the Atlantic forest (table
25.3). This strong preference for Moraceae in general,
and Ficus spp. in particular, is likely due to several factors. Montgomery and Sunquist (1978) observed, for
example, higher levels of digestibility of Ficus trigonata
(one of the preferred feeding trees) versus other tree
species tested during in vivo experiments. Moreover,
high-ranking trees in terms of feeding selection also
were digested at the highest rates (Montgomery and
Sunquist 1978). Other likely reasons are the widespread
distribution of this family and the high number of existing species (Gentry 1990). Another important factor
is that Moraceae trees tend to be asynchronous in the
production of new leaves (and other seasonal items,
such as flowers and fruits), even when individuals in
the same locality are considered (Milton 1991; Queiroz
1995). This guarantees a more or less constant supply of

new leaves to sloths year round.
These plant species represent a very small proportion of the total number of tree and liana species available to sloths in their native forests. At BCI, Montgomery and Sunquist (1978) listed 91 tree species (>15 cm
DBH) in a sample plot of 0.77 ha. By including trees >1
cm DBH, Foster and Hubbell (1990) found an average
of 176 tree species in single ha plots, also at BCI. Even
higher tree diversity was found in the Atlantic forest
near Santa Teresa, where an astonishing 443 tree species (>6.4 cm DBH) were identified in an area of 1.02 ha
(Thomaz and Monteiro 1997). Thus B. variegatus consumed <18% and B. torquatus <7% of the tree species
present in their respective forests. Actual percentages
are probably much smaller if larger tracts of forest are
considered (home ranges of three-toed sloths can be
larger than 1 ha; see below).
An even smaller number of species are consumed
when individual sloths are examined. In an early study,
I observed that radio-collared maned sloths consumed
712 plant species each on an annual basis (Chiarello
1998a), a total that increased to 19 species/sloth/year
in a later, more intensive study (Chiarello et al. 2004).
In the latter study, the top five plant species consumed
by each sloth (n = 5 radio-collared individuals) made
up 7284% of the annual diet. Although differences
in individual diets have not been published for other
sloth species, equivalent levels of variation can be inferred. Montgomery and Sunquist (1978) observed
that modal trees (those in which sloths were most
frequently found or radio-located) were few and differed among monitored individuals. Only two of nine
2
6
11
9
7
10
273
Table 25.3. Trees and Lianas Consumed by Three-Toed Sloths at Four Study Sites
Bradypus variegatus
BCIa
Rank
1
2
3
4
5
Total
Remaining genera (spp)
Genus
Cecropia (2)
Lacmellea (1)
Poulsenia (1)
Ficus (3)
Anacardium (1)
8 species
16 (17)
Bradypus tridactylus
Mm
%
14.3
11.7
10.2
9.0
8.3
53.5
46.5
Ma
Genus
Coussapoa (1)
Pseudobombax (1)
Piranhea (1)
Hippocrateaceaeb (1)
Ficus (1)
5 species
11 (11)
%
38.4
12.8
6.9
6.8
6.0
70.9
28.6
Genus
Clitoria (1)
Cecropia (1)
Ficus (1)
Guatteria (1)
Cochlosperma (1)
5 species
6 (6)
Bradypus torquatus
ST
%
39.8
17.2
10.8
6.5
6.5
80.8
19.2
Genus
Ficus (2)
Micropholis (1)
Mandevilla (1)
Prunus (1)
Apocinaceaeb (1)
6 species
12 (12)
Sources: BCI (Montgomery Sunquist 1975), Mm (Queiroz 1995), Ma (Carmo 2002), ST (Chiarello 1998a).
Notes: Values in parentheses are number of species in each genus. See Table 25.1 for abbreviations of localities.
a. Considering only species that were observed being eaten by sloths.
b. Identified at the family level only.
radio-collared sloths chose the same tree species as

the top-ranking modal tree (Anacardium excelsum,
Montgomery and Sunquist 1978). On the basis of these
and related results, Montgomery and Sunquist (1978)
believed that preference for each set of tree species is
passed down from generation to generation through
each mother sloth to her young, thus each genealogy
of three-toed sloths tends to have a different set of tree
species preferences from other genealogies of sloths
which live in the same area. Although likely, this hypothesis still awaits testing. Similarly, modal trees differed among the 11 three-toed sloths most frequently
observed by Queiroz (1995). One particular species of
Euphorbiacea (Piranhea trifoliata) was the top modal
tree for four individuals; second was a fig tree (Ficus
trigonata, Moraceae), which ranked first in preference
for three other individuals; four other tree species were
top ranking for the remaining sloths.
From the data summarized in tables 25.2 and 25.3
one might conclude that sloths are highly selective folivores. In a sense, this is true, but only if diets of individual sloths are considered. Montgomery and Sunquist (1978, p. 333) remarked that three-toed sloths,
when considered as a species rather than as individuals, were generalists with respect to the large number
of trees they used. However, as individuals, these animals specialized on just a few species that differed from
individual to individual (Montgomery and Sunquist
1978). Montgomery and Sunquist arrived at these conclusions because of the relatively high number of tree
species used by the sloths: their radio-collared sloths
were found in 40 of 91 tree species available to them
in the sample area. In other words, their study animals
used a relatively high percentage (44%) of the existing
tree diversity. It is important to mention, however, that

these authors considered as used all those tree species
where sloths were radio-located, and not, as one might
expect, those trees actually used as sources of food. In
fact, they rarely observed sloths feeding (Montgomery
and Sunquist 1975). Therefore, the terms generalist
and specialist, used by Montgomery and Sunquist
(1978) and also by Queiroz (1995), refer more appropriately to habitat selection than to diet selection. Nevertheless, recent studies on maned sloths have confirmed
that individual diets are indeed composed of a small
number of plant species (Chiarello 1998a; Chiarello et
al. 2004). Thus, the available data do seem to indicate
that Montgomery and Sunquist (1975, 1978) were right
in their statements about feeding selection in sloths.
As individuals, sloths are feeding specialists, but, given
that diets of individuals vary, sloth populations should
be regarded as generalist folivores.
This reasoning makes sense in adaptive terms. Neotropical forests are notoriously rich in tree species, but
most species are rare (Gentry and Terborgh 1990; Hubbell and Foster 1990). In fact, the general pattern for
a given site is to present a small number of abundant
species (several individuals per hectare), while the majority of tree species are represented by a few individuals per hectare, and some are even scarcer (Thomaz
and Monteiro 1997). One can reason that in such an
environment a selective folivore would be forced to
range widely, perhaps tens of hectares or more, as only
then would enough individuals of a given tree species
be available to guarantee an annual supply of young
leaves. In such a scenario, small home ranges would be
possible only if folivores chose to feed on the few most
abundant plant species. A corollary of this is that diet
%
19.5
16.4
13.7
12.2
6.8
68.6
31.4
274
A. G. Chiarello
diversity would have to be low among individuals, as

all individual folivores would feed mostly on the few,
highly abundant, species of trees.
As described above, this is not the case for threetoed sloths. First, different individuals chose different
subsets of tree species in their diet (or as modal trees).
Second, the most consumed tree species were, in most
cases, not the most abundant in the forest. In the Atlantic forest, for example, the preferred tree species
for maned sloths (Microphilus venulosa, Sapotaceae;
16.4% of overall diet for three individuals) occurred at
a density of 12 trees/ha (Thomaz 1996). In fact, this tree
was not included among the ten most abundant of the
region (range: 22149 trees/ha; Thomaz and Monteiro
1997). On the other hand, the least important tree species for these sloths (Ocotea aciphyla, Myrtaceae, 0.02%
of diet; Chiarello 1998a) was six times more abundant
(75 trees/ha) than M. venulosa (Thomaz and Monteiro
1997). A similar situation was observed at BCI: the top
modal trees were not the most abundant species in the
two areas sampled by Montgomery and Sunquist (1978).
In Amazonian flooded forest, however, the most consumed tree species were relatively abundant (Queiroz
1995). Although the preference rank of feeding trees
was not related with the rank of tree frequency, the
five most consumed tree species were among the top
ten species in terms of number of individuals (Queiroz
1995).
Thus, in contrast to other populations, the threetoed sloths of Mamirau (Amazon forest) based their
diet on abundant species of trees, such as Coussapoa
spp., Pseudobombax munguba, and Piranhea trifoliata
(ranked 1st3rd in diet and 9th, 7th, and 1st, respectively, in number of individuals; Queiroz 1995). Perhaps this difference is due to the much lower number
of tree species present in the flooded forest: 37 and
68 tree species (>10 cm DBH) were identified in two
one-ha plots sampled in chavascal and restinga baixa
forests, respectively. Together, these two plots totaled
80 tree species (Queiroz 1995). Although this figure is
not entirely comparable with those of BCI and Santa
Teresa because of different choices of DBH classes and
other idiosyncrasies of sampling, it does suggest that
the flooded forest of Mamirau is significantly poorer
in tree diversity than other sites. Consequently, sloths
there may have many fewer tree species available as
sources of food. The small Escudo island, where B. pygmaeus is found, seems to be another (extreme) example
of this situation, as the pygmy sloth eats primarilyif
not exclusivelythe leaves of mangrove trees (Rhi-
zophora mangle, Anderson and Handley 2002). Interestingly, the small size of this sloth might have been an
adaptive response (dwarfism) to this low quality food
source (Anderson and Handley 2002).
Home range and ranging patterns

Studies on sloth home ranges are few and restricted
to two Bradypus species (B. torquatus and B. variegatus) and Choloepus hoffmanni (table 25.4). Although
the studies are not homogeneous in regard to sample
sizes and methods, results suggest there are no obvious
differences in home-range size between sloth genera.
Available estimates are relatively small (range: 0.510.8
ha/sloth) and observed variation is as large within as
between studies (both in the range of one order of magnitude).
This variation among individuals is understandable
if one considers the heterogeneous nature of Neotropical forests (changing tree composition and structure
from place to place, Hubbell and Foster 1990; Milton
et al. 1994) and corresponding changes in availability
of feeding trees. Changes in tree composition among
sampling plots were detected in the three main study
sites: BCI (Montgomery and Sunquist 1978), Mamirau
(Queiroz 1995), and Santa Teresa (Thomaz and Monteiro 1997). In Santa Teresa, the most species-rich site of
the three, Thomaz and Monteiro (1997) observed that
the majority of sampled trees (57%) were found in just
one of three sampled plots, 31% occurred in two plots,
and only 12% of the tree species were common to all
three plots. Amazingly, these plots were located within
an altitudinal range of just 200 m (650855 m) and
less than 1 km apart. Likewise, in Mamirau, Queiroz
(1995) observed that only 31% of tree species were present in both of two areas inventoried by him (both of
which were used by sloths). Moreover, 1554% of tree
species were exclusive to one or the other site. At BCI,
two plots of similar size (0.34 and 0.43 ha) presented
similar numbers of tree species (61 and 57 species, respectively), but many of these were exclusive to one or
the other site because together they totaled 91 species
(Montgomery and Sunquist 1978). So, home-range size
of a given individual sloth may result from tree composition and abundance in the forest where it lives, rather
than from differences between species or even genera
of sloths. It would be interesting, however, if we could
test this hypothesis, perhaps by comparing densities of
preferred tree species among a range of different-sized
275
Table 25.4. Home-Range Size and Day-Range Length of Three Sloth Species at Five Localities in the Neotropics

Species
Locality
B. variegatus
B. variegatus
B. variegatus
B. torquatus
B. torquatus
C. hoffmanni
BCI
Mm
Va
ST
SJ
BCI
Number of
individuals
Mean home-range
Mean day-range
(months)
Method
size (range), ha
length (range)
9 (17)
19 (12)
9 (9)
5 (1036)
2 (6)
6 (17)
hoca
ad
ad hocb
-
MCPd
cell gride
ad hoca
1.6 (0.53.7)
0.9 (0.91.4)c
-
5.4 (0.810.8)
5.6 (5.55.8)
2.0 (0.43.9)
< 38 m/24 h in 89% of cases

23 m/24 h (5.349.5 m/24 h)
37.9 m/24 h (1133.3 m/24 h)
23.8 m/24 h (0306 m/24 h)
7597 m/week
> 38 m/24 h in 54% of cases
Reference
2
6
11
8,10
4
2

Notes: Abbreviations of localities are the same as in Table 25.1.
a. Home range estimated by the minimum-size rectangle enclosing all trees in which the sloth was located.
b. Home range estimated by calculating the area formed by linking all extreme locations.
c. Average for adult sloths.
d. Minimum convex polygon, using 100% of locations.
e. Home range estimated by counting the area formed by 0.25 ha cells occupied by sloths.
home ranges. Unfortunately, no such data are available

yet.
Sex differences in home ranges are still poorly
known, with only two studies providing relevant data
(Queiroz 1995; Chiarello et al. 2004; Chiarello et al. unpublished data). In the study of Queiroz (1995), B. variegatus adult females had slightly larger home ranges
(1.4 and 1.2 ha) than adult males (0.9 and 1.1 ha) in two
sampled areas (chavascal and restinga baixa, respectively). Pooling data from adult and nonadult individuals, the resulting averages were 0.92 ha for females (n =
10; SD = 0.688 ha) and 0.87 ha (n = 9; SD = 0.550 ha)
for males (Queiroz 1995). These differences are small
and probably not significant (the author did not test
for statistical significance). For maned sloths, adult
females had annual home ranges varying from 0.8 to
4.5 ha; the only adult male monitored ranged over an
area of 7.6 ha (Chiarello et al. 2004; Chiarello et al. unpublished data). The largest home range ever estimated
for a maned sloth was 10.9 ha (Chiarello et al. 2004)
but, unfortunately, accurate sex determination was not
possible for that particular individual (BT9.3; see LaraRuiz and Chiarello 2005 for a discussion on sexual dimorphism). Further information is needed before any
firm conclusions can be drawn on possible sexual differences in home ranges. Assuming sex differences do
occur, perhaps variation between the sexes, just as with
the sizable variation between individuals generally, is
fundamentally driven by forest characteristics, as discussed above.
With such small home ranges and energy-poor diets, it is not surprising that sloths travel small distances
on a daily basis. Available daily averages for Bradypus
spp. are very similar: 2338 m/24 h (table 25.4). Again,

lack of standards prevents an accurate comparison
among studies, but existing data suggest that interspecific variation within Bradypus is in the same range as
intraspecific variation (table 25.4). In contrast, Sunquist
and Montgomery (1973) mentioned that C. hoffmanni
tended to range farther (>38 m/day in 54% of recorded
occasions) than B. variegatus (<38 m/day in 89% of occasions). According to these authors, Bradypus were
frequently seen in the same tree on consecutive days,
while this was rarely observed in Choloepus. Other authors have corroborated that observation in Bradypus
spp. (Queiroz 1995, Chiarello 1998b), but unfortunately
there are no other studies on Choloepus for comparison. Thus, further studies are necessary to confirm a
possible intergeneric difference.
Only one study tested for individual differences in
ranging (Chiarello 1998b), but the observed variation
was not significant (average range: 21.726.3 m/24 h).
As three-toed sloths can be active both day and night
(see below), individual differences were also tested
separately for distances traveled during daytime (average range: 13.217.9 m) and during the night (average
range: 3.15.6 m). Again, individual differences were
not significant (Chiarello 1998b).
Data on seasonal variation in ranging are available
only for maned sloths (Chiarello 1998b). In that study
maned sloths ranged twice as far during the dry season
(average = 36.5 m/24 h; SE = 19.1 m/24h) as during the
wet season (average = 17.3 m/24h; SE = 3.70 m/24h),
but this difference was not significant. This lack of significance was clearly influenced by the huge variation
(high standard error) recorded during the dry season.
276
A. G. Chiarello
The lack of seasonal differences was also maintained

when diurnal and nocturnal ranges were tested separately (Chiarello 1998b).
Population density
Sloth density is difficult to estimate accurately because
these animals succeed extremely well in hiding themselves in the forest canopy. The problem is worse in Choloepus because of its nocturnal pattern of activity (see
below). As a consequence, the use of traditional techniques such as transect sampling (sensu Buckland et al.
1993) are not adequate. Traps cannot be used to capture
them, for obvious reasons. Additionally, because sloths
do not call regularly, they cannot be detected by their
vocalizations in a systematic way, as has been done with
other arboreal mammals such as primates. To overcome
these difficulties, some authors have used alternative
methods (table 25.5). Montgomery and Sunquist (1975),
for example, generated their estimate by counting the
number of sloth fecal masses found while clearing a
small area of the forest litter during a known period
of time. Others have used less controversial data, such
as the number of sloths captured in rescue operations,
for example, during construction of hydroelectric dams
(Taube et al. 1999). In still other situations, researchers have used the conservative approach of dividing the
number of known sloth individuals by the size of the
sample area (Queiroz 1995; Carmo 2002). Finally, in the
absence of more accurate alternatives, one can obtain a
gross estimate by using known home ranges, calculated
using radio-collared individuals, as is advanced here for
the first time for B. torquatus (table 25.5). It is likely that
all these approaches are biased in one way or another
and therefore they should be considered with care.
Keeping this caution in mind, density data are available for five species (table 25.5). In all cases, variation
is high between studies and also within species (ca. one
order of magnitude), which is not surprising given the
limitations described above and the use of differing
methods. Accurate comparisons are thus difficult. Nevertheless, it seems evident that sloths can achieve very
high densities in some places, and not just on islands
such as BCI, or in forest fragments such as the Manaus
site (800 ha; Carmo 2002), where sloth predators may be
absent, but also in continuous forests such as Mamirau
(Queiroz 1995). In some cases amazingly high densities
of 78 Bradypus/ha have been estimated. In general,
Choloepus spp. (range: 0.042.7 individuals/ha) do not
seem as abundant as Bradypus spp. (range: 0.098.46
individuals/ha), and this holds true even when the two

genera are compared within the same studies (BCI,
Mamirau, Manaus, and French Guiana; table 25.5). In
Manaus and Mamirau, the results might have been biased toward Bradypus because the estimates were based
on visual detection of individuals (Bradypus are more
easily seen than Choloepus). However, at the BCI and
French Guiana sites, density estimates were derived
from other methods that did not depend on this. So
available data do indicate that Bradypus are more abundant than Choloepus, a result backed by other studies
on mammal communities in the Neotropics (Eisenberg
and Thorington 1973; Peres 1999). Additionally, these
findings suggest Bradypus is an ecologically adaptable
genus, able to survive and sometimes proliferate in a
range of forest types. Indeed, some studies have found
sloths to constitute the predominant arboreal mammalian biomass in Neotropical forests (Eisenberg and
Thorington 1973; Peres 1999).
The ecological flexibility of Bradypus is confirmed by
information on wild individuals surviving, for years, in
small urban parks and plazas (Luederwaldt 1918; Carvalho 1960). In the 1.68 ha plaza of Valena city in Rio
de Janeiro, for example, located well within an urban
environment, several B. variegatus are said to have lived
there for at least 50 years (Consentino 2004). Currently,
12 sloths are found there (7.14 individuals/ha), although
there is no information on the health of this population
(Consentino 2004). A similar situation is found in another small plaza (~1 ha) located in the center of Tefilo
Otoni city, in Minas Gerais state, where several B. variegatus (16 individuals in the last count) have been living
for decades, according to local inhabitants (personal
observation; see also Gonalves 2003).
Few data on sex ratios are available and these are
restricted to three Bradypus species (table 25.5). The
lack of data on Choloepus is certainly influenced by the
greater difficulty in discriminating sexes in this genus,
as animals lack the secondary dimorphic characters
present in Bradypus (Goffart 1971; Lara-Ruiz and Chiarello 2005). Differences in male and female genitals
in both genera are slight and inconspicuous (Pocock
1924; Goffart 1971), but Bradypus males have either a
speculum (an orange or yellowish patch of hairs surrounded by black hairs in the mid dorsum), present in
both B. variegatus and B. tridactylus, or a black mane,
which is characteristic of B. torquatus (both males and
females have manes, but these are larger, darker, and
more conspicuous in sexually mature males than in females; Lara-Ruiz and Chiarello 2005).
277
Table 25.5. Density and Sex Ratio Estimates of Five Sloth Species at Several Locations in the Neotropics

Species
Locality
B. variegatus
B. variegatus
B. variegatus
B. tridactylus
B. tridactylus
B. torquatus
C. hoffmanni
C. hoffmanni
C. didactylus
C. didactylus
C. didactylus
BCI
BCI
Mm
Ma
SRFG
ST
BCI
BCI
Mm
Ma
SRFG
Density
(individuals/ha)
8.46a
7.70
6.72
2.21d
0.09e
0.091.25g
1.05h
2.70
0.88
0.13
0.04e
Sex ratio
(male:female)
Method
1:1.1

1:1.0
1:1.4

1:1.7f

Censusb
Censusc
Censusc
Censusc
Rescue operatione
Home rangeg
Censusb
Censusc
Censusc
Censusc
Rescue operatione
Sampling
area (ha)
Reference
0.324
1.3
11.8
24
7,100
-
0.324
1.3
11.8
24
7,100
2
2
6
9
12
10,13
2
2
6
9
12
Sources: Numbered references are the same as those in Table 25.1, with the addition of (12) Taube et al. 1999 and (13) Lara-Ruiz and
Chiarello 2005.
Notes: Abbreviations of localities are the same as in Table 25.1, with the following addition: SRFG = Sinnaramy River, French Guiana
a. 2.77 adult males + 2.93 adult females + 0.97 juveniles + 1.79 young/ha.
b. Estimate based on census of fecal masses deposited by sloths in an area of known size during a known period of time.
c. Estimate based on number of sloths found in an area of known size.
d. Or 1.88 adults/ha (= 0.75 males + 1.04 females/ha).
e. Based on number of sloths captured during 17 months of rescue operations in an area to be flooded by a reservoir.
f. Estimate based on 30 adults (19 females and 11 males) captured in the wild between 2002 and 2003 in Espirito Santo, southern Bahia
and northern Rio de Janeiro (see Lara-Ruiz and Chiarello 2005 for details).
g. Estimate based on the size of home ranges of radio-collared sloths monitored for at least one year in the Santa Teresa region (Chiarello
et al. 2004).
h. 0.99 adults + 0.06 juveniles/ha.
The sex ratio is unbiased in B. variegatus populations

studied in Panama (BCI) and in flooded Amazon forest
(Mamirau), but is slightly biased toward females in B.
tridactylus and seems to be even more so in B. torquatus
of southeastern Brazil (table 25.5). The female-biased
sex ratio in B. torquatus might suggest either differential mortality of males or a female-biased sex ratio at
birth, but currently there are no data regarding these
possibilities.
Recent data on maned sloths (Lara-Ruiz and Chiarello 2005) and on B. tridactylus and C. didactylus
(Richard-Hansen et al. 1999) indicate females are significantly heavier than males, contrary to what is observed for most mammalian species (Clutton-Brock
1989). This might suggest, among other things, that
competition for females among males is probably low
in these species (Lara-Ruiz and Chiarello 2005). In fact,
this might also support an alternative view that females
compete among themselves for access to males. If this is
true, we could conclude that the mating system of these
sloths is not polygynous. This would make sloths an
exception, as polygyny is the typical mating system in
most mammals and is characterized by sexual dimorphism where males are usually larger or stronger than
females (Clutton-Brock 1989). Alternatively, female-biased body weights could result from selective pressures
related not to mating systems but to other aspects of

natural history. For example, in some mammals, larger
females may be better mothers than smaller females (see
Ralls 1976 and Schulte-Hostedde et al. 2002 for general
reviews). Additional data are, however, clearly needed
to adequately test these hypotheses. Field data on other
populations are also necessary to verify whether these
biases are indeed representative of the species concerned or just artifacts of limited sampling.
Patterns and rhythms of activity

The incredibly low level of activity characteristic of
sloths has attracted the attention of scientists since the
time of Buffon or even earlier (Britton and Kline 1939;
Lundy 1952; Gould 1996), generating a huge number
of articles focusing, primarily, on the anatomy and
physiology of these animals (see Introduction). Several
early studies provide anecdotal information on sloth
activity patterns (Luederwaldt 1918; Beebe 1926; Lundy
1952; Goffart 1971; Meritt 1985a). Systematic information collected from free-ranging individuals observed
in their natural environment is, however, much scarcer.
Our knowledge on these topics was improved by the
pioneering work of Sunquist and Montgomery (1973)
on BCI, Panama. Currently, data are available for three
278
A. G. Chiarello
Table 25.6. Sampling Period, Main Period of Activity, and Average Percentage of Time Allocated to Main Activities of Four Sloth
Species at Six Locations in the Neotropics

Average % of time in activitya
Species
B. variegatus
B. variegatus
B. variegatus
B. tridactylus
B. torquatus
B. torquatus
C. hoffmanni
Locality
Method
BCI
Mm
Va
Ma
SJ
ST
BCI
Radiotelemetryb
Focal/Scan sampling
Focal/?
Focal/Scan sampling
Focal/Scan sampling
Focal/Scan sampling
Radiotelemetryb
Sampling period Main period of activity

00:0024:00
06:0022:00
00:0024:00
00:0024:00
06:0018:00
06:0018:00
00:0024:00
Day/Nightb
Night
Day
Day/Nightd
Nighte
Dayf
Nightb
RE
FE
MO
GR Reference

82.1
75.8

73.7

2.1
10.9

12.3

9.4c
10.2

11.3

3.4
3.1

2.7

1
6
11
9
4
8
1

Notes: See Table 25.1 for abbreviations of localities.
a. RE = resting, FE = feeding, MO = moving, GR = grooming.
b. Rates of activity determined by recording changes in the amplitude of signal from transmitters worn by free-ranging sloths.
c. Includes moving and traveling.
d. Traveling more accentuated at night and foraging more frequent in daytime and early evening.
e. Sloths were not observed at night but were probably more active then because only minor movements were observed during the day
(Pinder 1985).
f. Sloths were not observed during the night, but day-range lengths were significantly longer than night-range lengths (see text).
Bradypus species (B. variegatus, B. tridactylus, and B.

torquatus) and Choloepus hoffmanni (table 25.6). In
their paper, Sunquist and Montgomery (1973) reported
a strong divergence between the two genera in regard to
the active period: Bradypus did not present a clear pattern, with individuals being active both day and night,
but Choloepus was strictly nocturnal. The nocturnal
nature of the latter had already been suggested by early
authors working on captive or semicaptive animals (see
Meritt 1985a for a review). Unfortunately, there are no
other published data on Choloepus in the wild, perhaps
in part because this genus is indeed mostly nocturnal.
Recent authors have mentioned how difficult it is to
find two-toed sloths in the field (Queiroz 1995; Carmo
2002). This is probably because when not active, Choloepus seem to sleep (during daytime) well hidden within
dense tangles of lianas or other canopy foliage (Montgomery and Sunquist 1978).
Regarding three-toed sloths, recent work indicates
that they can indeed be active both day and night, but
there is much variation between studies (table 25.6). In
B. variegatus, for example, the study in the flooded forest
of Mamirau showed higher levels of activity, particularly moving, grooming, and feeding, during early night
(18:0022:00 h) than during daytime (Queiroz 1995).
This author, however, did not monitor sloths after 22:00
h, so no comparable information is given for the rest
of the night. Such data were available in another study,
in southeastern Brazil, where nine B. variegatus were

observed continuously for several sampling days (Consentino 2004). Feeding, moving and grooming were
observed with the highest frequency during daytime
(morning and afternoon); these activities decreased
strongly during early night (18:0024:00 h), and were
rarest from 24:0006:00 h (Consentino 2004). Sunquist
and Montgomery (1973) did not directly observe their
study animals at BCI; rather, the rates of activity were
determined by recording changes in the amplitude of
signal from transmitters worn by free-ranging sloths.
Nevertheless, a careful examination of figures 1 and 2
of their paper (pp. 948949) suggests a close resemblance to the results of Consentino (2004), particularly the pronounced decrease in activity from 24:00 to
06:00 h.
There is only one study on the congener B. tridactylus (Carmo 2002). This author observed each of her 10
study animals (six males and four females) for one period of 24 h. No overall percentages are given for their
recorded activities, but figures 1114 of her work unambiguously show that this species behaves much the
same as B. variegatus: moving, foraging, and grooming
were observed both day and night. Moving was most
frequent at night, foraging was concentrated during the
day and early night, ceasing almost completely from
24:00 to 06:00 h, and grooming followed a pattern
similar to that of foraging. Carmo (2002) observed that
males were more active than females, but no statistical

tests are mentioned. Apart from that, no information
on individual differences is presented.
Results from maned sloths (B. torquatus) are more
complete but also more complex. Pinder (1985) saw
only minor movements during daytime by two radiocollared individuals monitored in Poo das Antas Biological Reserve. These movements appeared restricted
to changes in position relative to sunlight. No data were
obtained during the night, but, given the low amount
of diurnal activity, the author concluded that the study
animals were nocturnal (Pinder 1985). In another study
with the same species near Santa Teresa (Chiarello
1998b), results indicated otherwise. All three radio-collared sloths were active throughout the day, although
the level of activity varied substantially among them
(range: 14.540.3% of daytime in feeding, moving, or
grooming). These animals were not observed during
the night; the level of activity at that time was inferred
by comparing distances moved during the day (dayrange length) with those during the night (night-range
length; night defined as 18:0006:00 h). This comparison was justified because day-range length correlated
highly with time spent active (Spearman rank correlation: r = 0.761, p < 0.001; Chiarello 1998b). Thus, comparing day and night ranges indicated, in relative terms,
in which of the two periods sloths were most active.
The analysis showed that the average day-range length
(16.6 m/day, range = 13.219.1 m/day) was significantly
longer than night-range length (average = 4.6 m/night,
range = 3.15.6 m/night; t-test; p = 0.006) indicating
that these three individuals were mostly diurnally active (Chiarello 1998b).
Subsequent studies on two additional maned sloths,
carried out in another reserve in the same region (Santa
Teresa, southeastern Brazil), showed a different picture
(Chiarello et al. unpublished data). One animal was
highly active throughout the day, but the other was
incredibly sleepy, being the least active of all maned
sloths monitored thus far (it spent only 6.8% of daytime hours active). This level of activity (or, better, lack
of it) was seven times lower than that of the most active
maned sloth observed (48.1% of daytime hours active).
The behavior of this lethargic sloth (BT13, sex undetermined) is reminiscent of those studied by Pinder in
Rio de Janeiro (only minor movements observed during the day). To summarize, three of seven (42.9%) B.
torquatus observed to date seemed mostly nocturnal,
while the remaining four (57.1%) were predominantly
diurnal. The degree of individual variation in daytime
279
activities, and also in ranging, is therefore huge (figures

25.1 and 25.2).
Individual variation in timing of activity translates
to a lack of synchrony between animals, as noted by
Queiroz (1995) and tested by Chiarello (1998b). Each
sloth has its own rhythm, the causes of which are not
yet well understood (Chiarello et al. 2004). As a consequence, some authors have stated that Bradypus spp.
do not have a circadian rhythm of activity (Sunquist
and Montgomery 1973; Queiroz 1995). More recent
work has suggested this might not be the case. All three
sloths first monitored in Santa Teresa tended to start
feeding and moving at the same time of day (07:00
08:00 h; Chiarello 1998b). More importantly, the time
that elapsed between the start of these activities on
consecutive days was close to 24 h (range: 23.3724.48
h for feeding and 23.1523.37 h for moving; Chiarello
1998b). These findings are backed by recent studies
on free-ranging B. variegatus captured from the wild
and kept in a captive environment subjected to natural light conditions (Duarte et al. 2004). These authors
found the existence of a biphasic circadian fluctuation
in blood pressure with the highest level observed during the light period. According to Duarte et al. (2004),
higher blood pressure is associated with an increase in
activity such as feeding, clearly demonstrating, therefore, the existence of a circadian rhythm in this species.
Nonetheless, the issue is far from being settled, as the
existence of other rhythms (e.g., ultradian or infradian)
cannot be ruled out at present (Duarte et al. 2004).
Concluding remarks
As stressed throughout this chapter, although a lot is
already known about the ecology of sloths, much remains unexplored. For some species, even basic natural history data, for example, feeding behavior and diet
(particularly in two-toed sloths), are virtually absent.
We will need to collect such information in order to
support further comparative analyses. For example,
foraging data on Choloepus spp. would provide opportunities to explore correlations between diet and levels
of activity (and ranging patterns), not only within species, but principally in comparison with Bradypus spp.
The overview in this chapter suggests a host of additional questions. For example, why are Choloepus
and Bradypus so different in their period of activity?
Are these differences fundamentally driven by differing foraging strategies, and are these influenced or constrained by phylogeny? Data relevant to these questions
280
A. G. Chiarello
Figure 25.1. Average percent

of time spent by five radio-collared sloths in resting, feeding,
moving, and grooming at
Santa Lcia Biological Station (sloths BT6.5, BT9.3, and
BT11.3) and at So Loureno
Municipal Park (sloths BT1.5
and BT13), Santa Teresa municipality, southeastern Brazil.
ANOVA results are shown
above each group of bars. See
Chiarello (1998b) and Chiarello
et al. (2004) for details on
methods.
Figure 25.2. Average distances

traveled by five radio-collared
sloths during the day (06:00
18:00 h), night (18:0006:00
h), and 24 h (06:0006:00 h)
at Santa Lcia Biological Station (sloths BT6.5, BT9.3, and
BT11.3) and So Loureno Municipal Park (sloths BT1.5 and
BT13), Santa Teresa municipality, southeastern Brazil. ANOVA
results are shown above each
group of bars. See Chiarello
(1998b) and Chiarello et al.
(2004) for details on methods.
also might help us understand why Choloepus generally

has lower population densities than three-toed sloths.
However, a full understanding of that issue will require
information on mating and social systems. How do
individual sloths monitor and find each other in the
structurally complex environment of Neotropical forests? Do males compete for females or do females compete for access to males? Do males tolerate home-range
overlap with other males? Do females? I hope these interesting questions can stimulate future research, not
only for the sake of knowledge per se, but also to help us
better plan conservation strategies for the endangered
species.
Acknowledgments
I thank the following people who have helped with
data collection for Projeto Preguia, which began
in 1994 and has continued since: Daniel Rios, Clarice
Bassi, Maria Amlia Maciel, Mariel Bazzalo, Leandro S.
Moreira, Rodrigo Lira Meyer, Ana Carolina S. Arajo,
Paula Lara Ruiz, and Oscar Echevery. I thank the editors, Jim Loughry and Sergio Vizcano, for the invitation to contribute a chapter to this book. I also thank
the editors and an anonymous reader for a careful review of an early draft of this chapter.
26
Behavioral ecology of armadillos
Colleen M. McDonough and W. J. Loughry
Chaetophractus vellerosus. Apesar disso, o mais interessante atributo dessa reviso o pouco que sabemos
sobre tantas espcies de tatus, destacando a grande necessidade de mais estudos de campo para fornecerem
informaes para comparaes mais robustas entre os
txons.
Resumen
Este captulo es una revisin de los estudios ecolgicos
y de comportamiento realizados con armadillos desde
el ao 1900. La cobertura taxonmica es extremadamente irregular y slo unas pocas especies reciben la
mayor parte de la atencin. Sin embargo, aun en estos
casos, son raros los estudios de campo a largo plazo y
los objetivos son relativamente estrechos. Hasta la fecha, la mayora de los estudios se enfocan en patrones
espaciales (e.g., estimaciones de rea de accin o mbito del hogar) y dieta. Los trabajos ms recientes son
un poco ms diversos, por ejemplo, sobre la influencia
de la poliembrona en Dasypus novemcinctus y la aplicacin de modelos de forrajeo ptimo en Chaetophractus vellerosus. No obstante, el rasgo ms llamativo de
esta revisin es cun poco conocemos sobre muchas
especies de armadillos, lo que subraya la necesidad de
ms estudios de campo que provean datos para realizar
comparaciones ms robustas entre taxones.
Introduction
The primary framework for studying animal behavior
was formulated by Tinbergen (1963), who argued that
a complete explanation for any behavior required an
understanding of its immediate causation (both internal and external), ontogenetic development, functional
consequences, and evolutionary history. Of course, before one could begin answering such questions, one had
to first compile a catalog of all the various behaviors
an animal exhibited (i.e., create an ethogram). Then,
once one knew what the animal did, one could begin
asking questions about why and how it did it. As a
subdiscipline, behavioral ecology has focused primarily
on functional aspects of behavior, particularly in understanding how an individuals behavior influences its
ability to transmit its genes to succeeding generations
(Krebs and Davies 1984). It is this emphasis we wish to
review here.
The purpose of this chapter is to review what we
know about the behavior of armadillos. The behavior
and ecology of other xenarthrans are discussed in the
chapters by Rodrigues et al. (anteaters) and Chiarello
(sloths). Somewhat arbitrarily, we have opted to review
all published work since 1900. In developing this review,
we began with the papers found in the bibliography
created and maintained by Superina (2000, 2005). We
used online WorldCat and Current Contents searchable
Resumo
Neste captulo revisamos estudos comportamentais e ecolgicos conduzidos com tatus desde 1900. A
abrangncia taxonmica foi extremamente descompensada, com poucas espcies recebendo a maioria da
ateno. Todavia, mesmo nestes casos, estudos baseados em expedies extensas eram raros e o foco foi relativamente estreito. Assim, as maiorias dos estudos focaram-se no padro espacial (por exemplo, estimativas
da distribuio) e dieta. Estudos mais recentes tm sido
um pouco mais diversos, por exemplo, examinando a
influncia da poliembrionia em Dasypus novemcinctus
e a aplicao de modelos de forrageamento timo em
281
282
C. M. McDonough and W. J. Loughry
databases to identify additional sources. Traditionally,

behavioral observations of armadillos often relied on
captive animals. Such data must be viewed cautiously
because of small sample sizes and concerns about how
representative they may be of naturally occurring behavior in the wild (see Superina et al. this volume). Because
of these difficulties, we opted to focus on data collected
in the field. We also excluded from consideration any
accounts that were primarily anecdotal, such as surveys
or brief taxonomic synopses. Thus, we were left with
a handful of studies in which animals were studied in
the wild for a significant amount of time (2 months)
and behavioral or ecological observations were a major
focus of data collection (tables 26.1 and 26.2). However,
we recognize this list may be incomplete: armadillos are
found primarily in Latin America and much of the research on them has been done by inhabitants of Latin
countries. The resulting publications can be difficult to
locate or access, so it is possible that we have omitted
some relevant works. Nonetheless, we believe we have
located the vast majority of the published data on armadillo behavior and ecology, providing a sufficient basis
for the following review.
What do we know?
While not exactly a hundred years of solitude, work
on armadillos since 1900 definitely cannot be considered exhaustive. Remarkably, for most species, we have
not progressed very far in our understanding of their
behavior. In many cases we are still at the descriptive
stage; indeed, in some cases we are not even that far.
Of the 21 extant species of armadillos, we could find
no publications meeting our criteria for 7 (Cabassous
centralis, Ca. chacoensis, Chaetophractus nationi, Chlamyphorus retusus, Dasypus pilosus, D. septemcinctus,
and D. yepesi). Of the remaining 14, 8 (Ca. tatouay, Ca.
unicinctus, Cha. vellerosus, Cha. villosus, D. novemcinctus, Euphractus sexcinctus, Priodontes maximus, and
Tolypeutes matacus) have been the subject of field studies lasting 6 months (table 26.1). However, this is a little misleading because, while table 26.1 provides the beginning and ending dates for each study, in many cases
fieldwork occurred only during a limited portion of this
overall period (e.g., Kalmbach [1943] reports data from
7 years of fieldwork, but the actual time spent in the
field observing animals totaled just 11 months).
It is clear that the most extensively studied species
is the nine-banded armadillo, D. novemcinctus, which
may not be surprising because it has the largest geo-
graphic range (Wetzel 1985b) and, in many locales, is

relatively abundant. But even here there are several
concerns. First, nearly all our data come from North
American populations, a region D. novemcinctus has
colonized only in the last 200 years (Humphrey 1974;
Taulman and Robbins 1996). Behavior can clearly vary
geographically as a result of adaptation to local conditions (Lott 1991; Foster and Endler 1999), so observations in this relatively new environment might not be
representative of that occurring in more ancient parts
of the species range, in conditions under which the
species presumably evolved. Second, although numerous studies have been conducted, the focus has been
relatively narrow. Most work has examined just two
aspects of behavior: home-range size and diet. Finally,
in some cases, behavior was inferred rather than being observed directlyfor example, in studies of activity patterns, where track counts were used (Layne and
Glover 1985), in studies of burrows, where features of
the burrow were measured but the animals inhabiting
those burrows were not observed, and in dietary studies that relied on collection of gut contents from dead
animals (but see Kalmbach 1943; Fitch et al. 1952; Gause
1980; Zimmerman 1982; Breece and Dusi 1985). Thus,
it seems clear that, not only do we need data from species that so far have been overlooked, but we also need
more diverse sampling (both in terms of populations
and aspects of behavior) for those species that are better
known. Bearing in mind these important caveats about
the data available, we now turn to an overview of what
we know about various aspects of armadillo behavior
and ecology.
Diet
Armadillos possess several morphological features that
are associated with feeding strategy. Most armadillos
dig in the soil while foraging. Digging is essential, not
only for procuring soil-dwelling prey such as coleopteran larvae, but also for burrowing to build nest chambers or to escape predators. Armadillos have short,
thick limb bones that can apply large forces against soil
for digging through substrate. Armadillos also have
claws that enable them to dig into compacted soil or
rip into decaying logs and, in some cases, carrion. Cabassous and Priodontes have enlarged specialized claws
for ripping into termite mounds. Most armadillos have
an elongated tongue with a sticky surface enabling
them to gather up small mobile prey. Armadillo teeth
are peglike and not differentiated. Consequently, little
Table 26.1. Field Studies of the Behavior and Ecology of Armadillos, 19002006
Species
Locationa
Ref(s)
Duration
Cabassous spp.
Argentina
24
not provided
Ca. tatouay
Brazil
18,21
Sep. 1979Jun. 1980
Ca. unicinctus
Brazil
18,21
Sep. 1979Jun. 1980
Cha. vellerosus
Argentina
22
Nov. 1971Jul. 1972

Bolivia
44
19962005
Cha. villosus
Argentina
42
19982001

Bolivia
44
19962005
Chl. truncatus
Argentina
25
Spring 1972 (7 weeks)

Argentina
3
not provided
D. hybridus
Uruguay
38
Feb.May 1997
D. novemcinctus
Bolivia
44
19962005

Brazil
19
Apr.1977Sep. 1978

Costa Rica
35
Feb.Mar. 1982

United States

AL
20
Jul. 1972Aug. 1973

FL
6
Feb.Dec. 1951

13,16
Winter 1976Spring 1978

28
12 months, dates not given

9,23
Jul. 1968Jul. 1979

39
Jun. 1992Aug. 2003

FL/GA
14
Jan. 1975Mar. 1978

GA
36,40
Jul. 1987May 1988

43
Jun. 2005Jun. 2006

LA
12
Jun.Aug. 1979

7
Feb.Jun. 1948

26
Mar.Apr. 1986

34
Feb.Apr. 1970

MS
11
Mar.Aug. 1977

MO
31
Jan. 1993May 1994

OK
17,27
Jul. 1980Mar. 1982

TX
5
Oct. 1947Oct. 1948

2
May 1932Jul. 1936

30,33,37
Sep. 1987Jul. 1991

29
Mar. 1990Mar. 1992

4
Aug. 1938Jun. 1939

15
Jun. 1978Jul. 1978
D. sabanicola
Venezuela
10
not provided
E. sexcinctus
Bolivia
44
19962005

Brazil
18,21
Sep. 1979Jun. 1980

Brazil
19
Apr. 1977Sep. 1978
P. maximus
Bolivia
44
19962005

Brazil
18,21
Sep. 1979Jun. 1980
T. matacus
Argentina
8
not provided

Bolivia
41
Jun. 2000Feb. 2001

44
19962005

Brazil
1
2 months, dates not given

19
Apr. 1977Sep. 1978
T. tricinctus
Brazil
32
not provided
Z. pichiy
Argentina
8
not provided

Argentina
45
20012005
Emphasis
Sample sizeb
captive behavior, natural history

activity, burrow use
activity, burrow use, home range
activity, burrows, home range
activity, reproduction, diet
burrows
reproduction, diet
burrows, natural history
burrows
reproduction, activity, diet
density, habitat preferences
activity, burrows
not provided
1
5
91
130
406
not provided
~10
1060
11
7
activity, home range

activity, natural history, burrows
evolution of polyembryony, natural
selection on morphology, home range
home range
home range, activity
Influence of polyembryony, activity,
population biology
some behavioral observations
burrow use, home range
habitat use, home range
behavioral observations
(mostly in captivity)
home range, natural history
leprosy, population density,
home-range overlap
activity, home range, social interactions
burrows, home range
natural history, home range
natural history
home range, social interactions
burrows, home range
activity, natural history
reproduction, activity, diet
reproductive behavior
density
activity, reproduction, diet, home range
natural history
home range
reproductive behavior
natural history, ecology, physiology
51
not provided
365
not provided
47
830
936
16
41
29
13
77
34
189
29
15
11
not provided
212
71
not provided
1
not provided
236
11
19
2
2
not provided
41
686
9
20
20
not provided
100
Sources: (1) Sanborn 1930, (2) Kalmbach 1943, (3) Minoprio 1945, (4) Taber 1945, (5) Clark 1951, (6) Bushnell 1952, (7) Fitch et al. 1952,
(8) Meritt 1973, (9) Layne and Glover 1977, (10) Pacheco and Naranjo 1978, (11) Jacobs 1979, (12) Christensen 1980, (13) Galbreath
1980, (14) Gause 1980, (15) Thomas 1980, (16) Galbreath 1982, (17) Zimmerman 1982, (18) Carter and Encarnaao 1983, (19) Schaller
1983, (20) Breece and Dusi 1985, (21) Carter 1985, (22) Greegor 1985, (23) Layne and Glover 1985, (24) Meritt 1985b, (25) Meritt 1985c,
(26) Stallknecht et al. 1987, (27) Zimmerman 1990, (28) Herbst and Redford 1991, (29) Smith 1992, (30) McDonough 1994, (31) Schell
1994, (32) Guimares 1997, (33) McDonough 1997, (34) Suttkus and Jones 1999, (35) Vaughan and Shoenfelder 1999, (36) Bond et al.
2000, (37) McDonough 2000, (38) Gonzalez et al. 2001, (39) Loughry and McDonough 2001 (and references therein), (40) Bond et al.
2002, (41) Cullar 2002, (42) Abba et al. 2005, (43) Gammons 2006, (44) Cullar this volume, (45) Superina this volume.
a. For sites within the United States, AL = Alabama, FL = Florida, GA = Georgia, LA = Louisiana, MS = Mississippi, MO = Missouri, OK
= Oklahoma, TX = Texas.
b. The total number of animals (living and dead) examined in each study. Blanks indicate studies, i.e., of burrows, where no animals were
sampled.
284
processing of food takes place in the mouth. At least

one species, D. novemcinctus, has chitinase activity occurring in gastric tissue to help digest the exoskeletons
of insects, as is found in other vertebrate insectivores
(Smith et al. 1998).
While there are many studies that describe armadillo diets (see below), there are almost none using
armadillos as model systems to test foraging theories.
One notable exception involved experimental analysis
of spatial foraging patterns in Cha. vellerosus (Cassini
et al. 1990; Cassini 1993). In this study, individuals did
not randomly sample their foraging environment, but
spiraled in a systematic pattern through an artificial
patch (Cassini 1993). In addition, the animals seemed
to monitor their feeding success, using this information
in deciding whether to remain or move on to a new
food patch (Cassini et al. 1990).
In reviewing the remaining studies of armadillo
diets, we have adopted a qualitative approach, partly
because of differences in methodology between studies (see table 26.2). Studies of diet have focused on two
primary aspects: species composition of gut or fecal
contents, and seasonal changes in diet composition
(table 26.2). With regard to diet composition, the summary provided in table 26.2 is not exhaustive. We did
not include every single taxon consumed, but instead
chose to emphasize only those constituting major components of the diet. Similarly, we combined some data
from the original works to get values for groups; for
example, various beetle species are not listed separately
but are grouped as coleopterans. In summarizing seasonal changes in diets, we used the seasonal designations given by the authors. When monthly totals were
provided for North American populations, we combined data into 4 seasons: winter (DecemberFebruary), spring (MarchMay), summer (JuneAugust),
and fall (SeptemberNovember). One study (Wirtz et
al. 1985) had seasonal totals that were off by one month
Table 26.2. Field Studies of Armadillo Diets

Insecta
Season
Plant
Vert. Oligo. Total
Col. Hym.
Iso.
Dip.
Cha. vellerosus
Argentina
8
15 months
% volume

Bolivia
23
na
% dry weight
Su
W
na
48
36
15
22.0
50.7
56.0
27.7
13.9
0.2
0.1
0
na
46.0
25.7
41.0
31.0
20.7
na
0.3
2.8
na
0
0
na
0
0
na
Cha. villosus
Bolivia
23
na
% dry weight
na
33
58.9
3.4
na
35.5
na
na
na
na
D. kappleri

Colombia
Venezuela
12
15
na
na
% total prey
% volume
na
Su-F
1
4
0
0.1
14.3
0
0
0
71.4
44.5
19.0
42.1
19.0
0.8
9.5
tr
0
0.6
8.5
1.1
na
84.9
na
na
na
na
2.3
11.8
1.7
2.0
46.2
65.5
59.5
70.6
tr
0.6
tr
1.1
0.8
na
na
na
na
7.0
17.0
15.0
18.0
2.2
1.1
7.3
23.7
0.13
0.14
0.24
0.97
0.1
1.2
0.7
2.4
2.4
na
na
na
na
na
na
na
na
0.8
0.8
8.9
8.0
2.5
3.7
2.1
3.9
tr
6.4
7.4
15.6b
4.9
4.0
9.0
9.0
4.0
18.0
4.0
28.0
40.0
66.1
53.6
55.2
29.0
na
na
na
na
59.6
56.8
58.6
69.9
78.5
na
na
na
na
na
na
na
na
28.2
19.5
28.7
12.2
2.3
3.2
9.6
7.4
49.3
53.2
6.5
34.8
29.7
80.0
78.0
49.0
51.0
53
6.2
12.9
5.2
21.3
28.3
15.5
4.9
9.4
3.8
0.5
6.6
4.2
1.4
1.5
12.2
15.1
25.0
27.0
32.0
19.0
51.1
36.9
82.4
79.2
6.9
0.3
0.1
0.9
39.1
61.9
22.6
12.3
tr
0.7
tr
na
0.1
na
na
na
na
1.7
0.3
0.2
0
1.9
1.5
0.4
4.2
49.2
31.2
67.2
73.7
1.5
0.65
49.9
tr
13.3
20.0
9.0
11.0
42.0
36.9
55.0
3.1
12.5
Species
Location
References Duration
Measure
D. novemcinctus Bolivia
23
na
% dry weight
na
43

United States

AL
5,13
1 year
% volume
Sp
12

Su
18

F
14

W
6

17
2 years
% volumea
Sp
39

Su
29

F
42

W
34
2 years

AR
16
% volume
Su
29

F
33

W
8

FL
3
11 months
% volume
na
139

6
na
% volumec
All
172

Sp
na

Su
na

F
na

W
na

14
1 year
variabled
Jan-Mar 52

Apr-Jun 49

Jul-Sep
35

Oct-Dec 50
continued
Insecta
Season
Plant
Vert. Oligo. Total
Col. Hym.
Iso.
Dip.

GA
20e
1.25 years
% dry weight

20e
1 year
% dry weight

LA
4
na
% volume

MS
9
3 years
% volume

MO
18
1 year
% volume

OK
10
1 year
% volume

TX
1
1 month
% volumef

2
7 years
% occurrence

Sp
Su
F
W
Sp
Su
F
W
Sp
Su
F
W
W
Sp
Su
F
W
Sp
Su
F
W
Sp
Sp
Su
F
W
8
16
8
12
7
15
6
15
31
27
29
16
36
10
13
9
13
20
13
11
6
25
52
97
5
15
0
27.7
12.7
9.6
0
7.4
11.1
0
9.8
9.9
13.2
8.1
10.3
0
1.4
0
0
7.5
4.3
1.3
3.3
2.6
2.2
8.2
5.9
8.5
1.8
1.0
5.3
3.9
1.5
0.4
0.1
4.4
2.2
3.0
2.2
13.2
2.0
1.8
3.4
1.7
0
1.9
1.3
0.3
7.7
0.7
1.5
0.9
0.8
2.6
2.2
0.1
0.8
3.4
2.1
0.1
0.1
1.4
9.7
1.1
2.7
13
5.4
5.3
6.2
17.9
5.0
16.4
0.7
0.9
2.0
13.6
na
na
na
na
na
na
na
na
na
na
na
na
66.7
72.5
70.1
60.6
72.2
74.3
70.9
72.9
67.1
70.2
89.3
95.6
81.3
77.4
85.2
72.8
86.5
67.3
45.8
24.6
34.0
24.1
22.6
39.8
49.7
19.5
33.6
55.9
56.0
24.7
23.3
32.4
17.1
26.5
28.9
36.3
34.7
76.8
9.7
92.0
52.2
37.8
52.9
25.8
11.9
26.0
11.9
12.0
26.7
22.5
9.7
2.2
5.0
5.8
1.7
4.1
10.5
30.2
28.6
22.6
2.2
7.2
45.8
15.4
0.7
40.0
8.5
20.4
12.9
16.2
na
na
na
na
na
na
na
na
1.5
1.7
0.7
0.5
2.8
na
na
na
na
0.3
tr
tr
0
4.0
2.8
3.0
15.2
0.1
4.8
0.3
0.8
1.9
3.2
0.3
0.9
2.4
12.1
1.0
0.1
1.7
23.9
1.7
12.9
11.1
25.4
7.4
0.2
tr
32.7
4.0
3.2
1.2
0.2
0.9
D. sabanicola

Colombia
Venezuela
12
7
na
na
% total prey
% total prey
na
na
1
na
0
na
0
occ.
0
occ.
99.9
na
1.1
18.0
10.0
22.0
88.8
45.0
0
na
E. sexcinctus

Bolivia
Brazil
Brazil
23
22
11
na
3.25 years
na
% dry weight
% volumeg
% occurrence
na
na
na
14
12
10
68.3
32.6
70.0
12.9
3.9
0
na
na
na
15.8
61.4
na
na
30.7
na
na
22.0
na
na
1.0
na
na
0.08
na
P. maximus

Brazil
Colombia
21
12
3 months
na
% occurrence
% occurrenceh
na
na
25
3
100.0
33.3
0
0
0
0
na
100.0
0
2.8
36
49.9
96
45.1
56
0
T. matacus
Argentina
19
9 months

Bolivia
23
na
% aggregate
weight
% dry weight
na
66
20.2
60.6
27.1
10.0
21.2
na
37
12.3
86.1
na
na
na
na
Species
Location
References Duration
Measure
Sources: (1) Baker 1943, (2) Kalmbach 1943, (3) Bushnell 1952, (4) Fitch et al. 1952, (5) Breece 1974, (6) Nesbitt et al. 1977, (7) Pacheco and Naranjo 1978, (8)
Greegor 1980b, (9) Singletary 1981, (10) Zimmerman 1982, (11) Schaller 1983, (12) Barreto et al. 1985, (13) Breece and Dusi 1985, (14) Wirtz et al. 1985, (15)
Szeplaki et al. 1988, (16) Sikes et al. 1990, (17) White 1992, (18) Lippert 1994, (19) Bolkovic et al. 1995, (20) Osborn et al. 2000, (21) Anacleto and MarinhoFilho 2001, (22) Dalponte and Tavares-Filho 2004, (23) Cullar this volume.
Notes: For prey types, numbers represent the percentages of plants, vertebrates (Vert.), worms (Oligo.) and insects in the diet. Insects are further broken into the
percentages of Coleoptera (Col.), Hymenoptera (Hym.), Isoptera (Iso.), and Diptera (Dip.) consumed.
na = not available, tr = trace, occ = occasional, Sp = spring, Su = summer, F = fall, W = winter.
Location abbreviations within the United States are the same as in Table 26.1.
a. % volume determined for plants, vertebrates, and worms; insect types reported as percentage of total number of insects per stomach.
b. Includes other invertebrates.
c. % volume provided for total sample across all seasons; seasonal data based on % occurrence (presence or absence) of each prey type in stomachs examined
(i.e., in spring 80% of stomachs contained coleopterans, 25% contained hymenopterans, etc.).
d. % occurrence reported for plants and worms, insect types reported as percentage of total number of insects per stomach; percentages not available for vertebrates and total insects as these were recorded as total number of individuals per stomach.
e. Study conducted at 2 sites: first one listed was oak-palmetto, second was oak-pine.
f. % volume reported for plants, vertebrates, worms and total insects, % occurrence for different insect orders.
g. % volume of each item is based on volume of items to total volume of all items present in all stomachs.
h. % occurrence reported for plants, vertebrates, worms and total insects; insect types reported as percentage of total number of insects per stomach.
286
from those just defined (see table 26.2). Values were averaged into seasonal totals when needed.
Patterns across species
Armadillos range in dietary preferences from specialists to generalists. Redford (1985a) categorized armadillos into four groups: (1) carnivores-omnivores; (2)
myrmecophages, specializing on ants and termites;
and two groups of generalist insectivores that are either (3) terrestrial or (4) fossorial. The first three groups
reflect taxonomic subfamilies identified by molecular
data (Delsuc et al. 2003), suggesting dietary differences
were critical aspects in the evolutionary diversification
of armadillos.
Armadillos eat a wide variety of food, from plants
and insects to small vertebrates such as rodents, lizards,
and snakes (table 26.2). Nonetheless, most armadillos
ingest large numbers of insects, in both adult and larval forms. Important insect taxa include Coleoptera
(beetles), Hymenoptera (ants and wasps), and Isoptera
(termites), all of which seem to be major components
of the diet year-round (table 26.2). The importance of
other insect orders to diet seems to depend on species,
location, and season (table 26.2). As one extreme example, White (1992) reported that the number of dipterans
consumed in one D. novemcinctus population varied
by 300% depending on season and year. Other invertebrates such as centipedes, millipedes, spiders, and
worms are also eaten. Soil or sand ingestion seems to
be common; Greegor (1980b) reported 50% of stomach
volume being sand in Cha. vellerous. The most accepted
hypothesis for this is that ingesting soil is incidental to
feeding on soil invertebrates (Bushnell 1952; Beyer et al.
1994). However, soil ingestion may help in digestion or
provide needed minerals (Pacheco and Naranjo 1978;
Beyer et al. 1994), or soil may contain large numbers of
very small insects such as Collembola (springtails) that
could be used as a food source (M. Blackmore personal
communication).
Among armadillos, the Euphractinae (E. sexcinctus,
Chaetophractus spp., and Z. pichiy) seem to have the
broadest diets (Redford 1985a; table 26.2). At least two
of these species, Z. pichiy and Cha. vellerosus, are found
in stressful environments (cold and xeric, respectively).
Their broad omnivorous diets could be an adaptation
for survival in areas where food, insects in particular,
might be scarce. The evolution of such diets may have
been driven by responses to the cooling and drying patterns that were occurring in southern South America at
the time of diversification of the subfamily (Delsuc et
al. 2004). Being able to switch among food sources, for

example, seasonally, may provide further dietary flexibility necessary to withstand extreme environments.
For example, in Cha. vellerosus, plants are more important in winter and vertebrates in summer (Greegor
1980b; table 26.2). However, physiological adaptations
such as water retention in the kidneys (Greegor 1975) or
temperature regulation (Superina this volume) may be
just as important. In contrast, E. sexcinctus has a wide
distribution that overlaps that of many other armadillo
species (Eisenberg 1989; Redford and Eisenberg 1992;
Eisenberg and Redford 1999). Its carnivorous/omnivorous diet (Dalponte and Tavares-Filho 2004) may allow
it to coexist with more specialized species. Comparative
studies of armadillo diets within the same habitat are
needed to begin testing this hypothesis.
A broad diet need not be generated by indiscriminant consumption of whatever happens to be available.
For example, some evidence suggests that the omnivorous E. sexcinctus may be selective. One individual, discovered as a roadkill, had consumed four rodents. Estimates of prey availability indicated this individual may
have had a preference for a low-density rodent species
(Bezerra et al. 2001).
Species in the Tolypeutinae subfamily (P. maximus,
Tolypeutes spp. and Cabassous spp.) are considered
ant and termite specialists (Redford 1985a). Although
no new accounts have been published on Cabassous,
recent work on Tolypeutes (Bolkovic et al. 1995) and
Priodontes (Anacleto and Marinho-Filho 2001) suggest
they may forage more opportunistically. In the case
of P. maximus, analyses of fecal samples from Brazil
found plant material, dipterans, and other invertebrates
in addition to the expected ants and termites (Anacleto
and Marinho-Filho 2001). Tolypeutes species are even
more omnivorous. Previous work lists oligochaetes, insects, and carrion as food sources for T. matacus (Meritt 1973c), and a recent study documented considerable
consumption of plants (Bolkovic et al. 1995). These
findings suggest that the classification of Tolypeutinae
as myrmecophagous specialists needs to be updated.
Indeed, field studies on Cabassous may show that the
entire subfamily is more opportunistic than previously
thought.
Information on diets of wild Chlamyphorus species
is sparse. Captive individuals consumed mostly ants
(Roig 1995). In fact, one animal fed only on ants, although others were observed eating coleopteran larvae
and other invertebrates, consistent with gut content
analyses (Minoprio 1945; Roig 1995).
Armadillos in the genus Dasypus are the only ones

categorized as terrestrial generalist insectivores (Redford 1985a). D. sabanicola may be the most specialized
feeder of the group, foraging almost exclusively on insects such as hymenopterans, isopterans, and coleopterans (Pacheco and Naranjo 1978; Barreto et al. 1985).
Diets of North American populations of D. novemcinctus have been studied extensively (table 26.2). These
populations represent the northernmost distribution
of extant Xenarthra (Wetzel 1985b), extending as far
north as southern Illinois (Van Deelen et al. 2002). In
such strongly seasonal environments, one might expect
changes in diet composition attributable to seasonal
changes in prey availabilities. Although strong seasonal
differences have been reported (table 26.2), no studies
have included prey availability data along with stomach
content analyses.
Consequences of diet
The similarity in diet of D. novemcinctus over vast geographical distances (table 26.2) may provide insight
into factors influencing the distribution of this species
in North America. D. novemcinctus has one of the lowest metabolic rates of any placental mammal (McNab
1980), with consequently poor thermoregulatory abilities. Thus, Humphrey (1974) proposed that armadillos
would be unable to survive in areas where the daily
high temperature did not exceed 0C for nine or more
days. However, dispersing nine-banded armadillos
have repeatedly expanded their northern distribution
into areas that exceed this projected limit (reviewed in
Freeman and Genoways 1998). An alternative hypothesis is that rainfall, because of its link with diet, is the
key climatic variable. Humphrey (1974; see also Taulman and Robbins 1996) proposed that annual rainfall
of 38 cm/year was necessary to generate sufficient insect populations to allow for armadillo survival. However, D. novemcinctus may be able to expand its range
in spite of climatic factors by utilizing prey other than
insects. For example, consumption of vertebrate prey
may make up for an absence of insects, such as during
the winter (Gause 1980).
The diet of D. novemcinctus may have conservation
and wildlife management implications as well. Early reports of armadillos depredating nests of ground-nesting birds (Kalmbach 1943; Bushnell 1952; Breece and
Causey 1973) did not elicit much concern until recently
(Staller 2001). Round-the-clock surveillance of nests
with fiber optic video cameras revealed armadillos to
be one of the top three predators of bobwhite quail eggs
287
at one location (Staller 2001). Studies of gut contents

probably missed such predation because egg albumin
and yolk are hard to detect. Concerns about the depredation of marine turtle nests are also increasing (Bushnell 1952; Douglass and Winegarner 1977; Marshall
1987; Maffei 1997; Engeman et al. 2003). Removing armadillos from areas around endangered marine turtle
nests has increased the number of successful nesting
attempts (Engeman et al. 2003).
Diet is undoubtedly related to other aspects of armadillo behavior. For example, those species that rely
heavily on ants and termites (i.e., the Tolypeutinae subfamily) seem to have large, dispersed home ranges and
few social interactions, rarely return to sleep in a burrow, and escape from predators by immediately digging
down into the soil rather than running to a burrow.
While it is tempting to speculate that ranging widely for
termites and ants makes construction and utilization of
burrows uneconomical, we currently lack the detailed
data necessary to test this hypothesis. Unfortunately,
the same is true for other species as well, where relationships between diet and other aspects of behavior
remain unknown.
Spatial patterns
Determinations of home-range size and the extent of
overlap between sex and age groups are integral to describing the social organization of a population. Homerange size has been calculated in five armadillo species
(table 26.3); however, in many cases the data are not
that robust. For example, just one study (Carter 1985)
is responsible for all the data collected on three species
(Ca. unicinctus, E. sexcintus, and P. maximus). Also,
in many instances, home ranges were calculated from
relatively small sample sizes (table 26.3).
As with diet, the most extensive data on home ranges
are for the nine-banded armadillo, D. novemcinctus.
Home-range estimates vary widely between populations of this species (0.6320.1 ha, table 26.3). There
are at least two possible explanations for this. The first
concerns differing habitat types, with armadillos inhabiting more mesic areas having smaller home ranges
(McDonough 2000), possibly because these areas have
higher prey abundances. A second explanation concerns methodology. For example, the extremely large
home-range size reported by Fitch et al. (1952) probably resulted from using a different measure than the
minimum convex polygon. Also, some of the very small
home ranges that have been reported could be due to
288
Table 26.3. Home-Range Values and Reports of Various Social Interactions in Field Studies of Armadillos

Species
References
Home-range
size, in ha
Home-range
(sample size)
overlap?
Ca. unicinctus
21
Cha. vellerosus
22
Cha. villosus
44
D. novemcinctus
2

4

5

6

7

9

11

12

13

17

20

26

28

29

30,33,37

31

34

35

36

39

43

44
D. sabanicola
10
E. sexcinctus
21

44
P. maximus
21

44
T. matacus
8

41

44
Z. pichiy
45
64.9 (5)
3.4 (1)

3.5 (3 adults)
0.3 (1 juvenile)

19.8 (1)
7.4 (4 adults)
4.0 (4 juveniles)
2.333.36 (61)

9.2 (2)
1.9 (4)
3.5 (21 adults)
1.5 (18 juveniles)

39 (?)
0.63 (10 adults)
0.13 (1 juvenile)
4.0 (19 adults)
0.9 (10 juveniles)
0.98 (6)
0.6 (23)

6.55 (9 survivors)
11.55 (5 died)

8.7 (27)

93.3 (11)

452.2 (2)
3,000 & 15,000 (2)

2.714 (27)

Population
Social Interactions
density
Reproductive Juveniles Burrow
(per ha)
Aggression Tolerance behavior affinitive sharing

yes

0.006

no

yes

yes

yes
yes
yes
yes

yes
yes
0.15
0.35

no
yes

yes
yes
yes

yes
yes
0.73

yes

yes
yes

yes
yes
yes

yes
yes

yes

yes
yes
1.5

0.13.9

yes
no

yes
yes

yes

yes
yes
yes
1.45
yes
yes
yes
yes
yes
yes
yes

yes
3.75

yes

yes

yes
yes

yes

0.007
2.8

0.012

~6a

0.02
0.07

yes
yes

yes

yes
yes

yes
yes

yes
yes
Notes: Numbered references are the same as those in Table 26.1.

Yes = a study reported that each item occurred or was observed; No = specific mention that each item did not occur; = no information was
provided.
a. Value given is for 100 km2.
methods used in sampling, such as solely following

roads with limited visibility to the sides (Jacobs 1979;
Suttkus and Jones 1999). Finally, within-population differences could be due to unequal sampling of individuals (Breece and Dusi 1985; McDonough 2000).
No sex differences in home-range size have been
found in D. novemcinctus (Layne and Glover 1977;
Schell 1994; Suttkus and Jones 1999; McDonough 2000;

Gammons 2006). However, age differences do occur,
with home-range size increasing with age (Clark 1951;
Breece and Dusi 1985; Smith 1992; McDonough 2000).
To support a larger body size, individuals presumably
need to increase the size of their foraging area. But
body size may not be the only influence on home-range
size. In one study, male reproductive status had an effect (McDonough 2000), with breeding males having
larger ranges than nonbreeding males. Although this
finding could be an age or weight effect, anecdotal evidence suggests otherwise. During the study one male
changed from nonbreeding to breeding and more than
doubled its home range, even though its weight remained constant (McDonough 2000). A larger home
range could generate a reproductive advantage because
females tended to pair with males that overlapped their
home range the most (McDonough 2000).
Although studies on D. novemcinctus have repeatedly reported on overlapping home ranges within and
between the sexes (table 26.3), few have quantified it
(Jacobs 1979; Herbst and Redford 1991; Schell 1994; McDonough 2000). Overall patterns suggest that females
share about 2530% of their home range with each overlapping female (Herbst and Redford 1991; McDonough
2000), resulting in 6673% of the home range being
overlapped by other females (Jacobs 1979; McDonough
2000). However, a northern population in Missouri had
overlap values that were substantially less, with females
overlapping just 8% with other females (Schell 1994).
Male-male overlap is typically less than female-female
(Jacobs 1979; Schell 1994), with values in the range of
1522%. Here again, male reproductive status may be
important, as McDonough (2000) showed that breeding
males overlapped very little (12%) with one another, but
overlapped about twice as much with each nonbreeding male. Nonbreeding males overlapped each other by
about 24% and overlapped with breeding males by 41%
(McDonough 2000). The extent of overlap between the
sexes depends on the study. Schell (1994) found male
ranges overlapped those of females by 21%, but in Texas
45% of a females area was overlapped by each breeding
male (McDonough 2000).
Instead of overlap, exclusivity, that is, the amount
of home range that is not overlapped by individuals of
the same sex and reproductive status, might be a useful
way of looking at spatial patterns. McDonough (2000)
showed that both breeding and nonbreeding males had
relatively more exclusive home ranges than did females
(breeding males = 69%; nonbreeding males = 50%; females = 27%). These results were consistent with those
of Jacobs (1979), who reported exclusivity values of 34%
for females and 77.5% for males (males were not classified by reproductive status).
289
Social behavior
Armadillos are relatively solitary, asocial animals, so
discussion of social behavior may seem paradoxical.
Indeed, no species studied so far is reported to maintain sustained social groupings, with the possible exception of D. sabanicola during the rainy season (Pacheco and Naranjo 1978). Nonetheless, armadillos are
long lived, may inhabit the same areas year after year
(Loughry and McDonough 2001), and often have extensive home-range overlap with other individuals (see
above). Thus, opportunities to form long-term social
relationships may exist. Preliminary results from a D.
novemcinctus population in Texas found that ~10% of
behavioral observations could be classified as social,
that is, at least two animals within 30 m of each other
(McDonough unpublished data).
Although the probability of observing social behaviors is low, the range of social behaviors is extensive,
from aggression, such as fighting and chasing, to maintaining close proximity to another individual. While
several studies of D. novemcinctus reported no aggression between animals (Clark 1951; Layne and Glover
1977), others described fights and chases involving
adults and juveniles (Denson 1979; Jacobs 1979; Galbreath 1982; Breece and Dusi 1985; Herbst and Redford
1991; McDonough 1994; Schell 1994). Aggression can be
costly, resulting in serious and permanent damage to
the combatants (Loughry et al. 2002). The incidence of
aggression could be related to differences in population
density, with most reports of aggression coming from
high-density sites (McDonough 2000). Aggressive behavior has been associated with the breeding season in
D. novemcinctus (McDonough 1994), D. sabanicola (Pacheco and Naranjo 1978), and Z. pichiy (Superina this
volume). Chasing may have a reproductive function
in E. sexcinctus as well (Desbiez et al. 2006), although
confirmation is needed on the sex and reproductive
status of individuals involved. Competition over food
also elicits aggression in captive Z. pichiy (Superina this
volume).
In D. novemcinctus, breeding males were aggressive
toward other younger and smaller males, primarily
during the breeding season (McDonough 1994). This
supports Herbst and Redfords (1991) assertion that
male territoriality (which is maintained by aggression)
is probably related to breeding rather than to excluding
individuals from a food resource. Female D. novemcinctus also exhibit aggression but, unlike males, they
were most aggressive during late pregnancy and lacta-
290
tion, and directed their aggression mostly toward juveniles of the previous year (McDonough 1994). In a
study using captive animals, female E. sexcinctus were
also found to be most aggressive during this same time
period (Gucwinska 1971).
In D. novemcinctus an interaction observed much
more commonly than aggression involves two or more
animals feeding in proximity (i.e., within 30 m of each
other, Clark 1951; Layne and Glover 1977; Jacobs 1979;
Breece and Dusi 1985; Smith 1992). Such observations
have led some to propose a more nonterritorial view of
armadillo social systems (Clark 1951; Breece and Dusi
1985). Tolerance of the presence of others was seen
in areas where aggression also occurred (Jacobs 1979;
Herbst and Redford 1991; Schell 1994; McDonough and
Loughry 1995), suggesting armadillos vary their behavior according to context or individual identity. Jacobs
(1979) stated that tolerance interactions between adult
female neighbors and between males and females
peaked in May and June, while those between males
peaked in June. This corresponded with the onset of
the breeding season in this population and suggests
that mating may influence the formation or maintenance of such interactions. For example, tolerance of
females by breeding males may enable them to establish
social relationships with females during the year and
thus increase their likelihood of courting and successfully inseminating these females during the breeding
season. Tolerance may also be influenced by diet. As
described above, the diet of D. novemcinctus consists
mostly of insects. This type of prey base may be hard
to defend. Interactions solely involving food resources
may not elicit aggressive behavior simply because it is
not economical to do so.
Armadillos also show affinitive behavior, which differs from tolerance in that individuals actively maintain
proximity to one another. Affinitive behavior has been
observed in two social contexts: (1) breeding pairs and
(2) juvenile siblings foraging together with or without
their mother. Both types of affinitive behavior have potentially interesting evolutionary consequences, which
we describe below.
Pairing and reproduction
Pairing behavior between a male and female has been
observed in several species, both in the field (D. novemcinctus: Jacobs 1979; Gause 1980; McDonough 1997,
2000; T. tricinctus: Cardoso da Silva and Oren 1993; Z.
pichiy: Superina this volume), and in captivity (T. matacus: Meritt 1976b; Cha. villosus: Roberts et al. 1982).
In D. novemcinctus and Z. pichiy mating is seasonal. In

the United States there is a distinct peak in observations
of reproductive pairs in midsummer (JuneJuly) for D.
novemcinctus (McDonough 1997; Peppler this volume);
pairs are most commonly observed in the austral spring
(mid-SeptemberOctober) for Z. pichiy in Argentina
(Superina this volume). Normally, male armadillos do
not have visible scrotums, but the testes of male Z. pichiy increase in size during spring and summer (Superina this volume).
D. novemcinctus pairs are easily recognizable for two
reasons (Jacobs 1979; McDonough 1997). First, reproductive pairs maintain extreme proximity (typically <1
m, something never observed otherwise in adults), with
males mainly responsible for maintaining this close association (McDonough 1997). Second, males and females perform certain behaviors, such as tail-wagging
by females and dorsal touching or mounting by males,
that are never observed in any other context in the wild
(McDonough 1997; dorsal touching and mounting do
sometimes occur in interactions between nonreproductive individuals in captivity, see Christensen 1980;
Loughry, Dwyer, and McDonough 1998).
Observations of reproductive pairs led to consideration of evolutionary issues regarding armadillo mating
systems and determinants of reproductive success. In
D. novemcinctus, there is mild evidence of polygyny:
males are often seen paired with 23 females during
the same breeding season, whereas most females are
only seen paired with a single male (Jacobs 1979; McDonough 1997). In addition, males are slightly larger
than females (McDonough 2000), although this sexual
dimorphism is not found in all populations (Loughry
and McDonough 1998a). Genetic analyses of reproductive success support this view. In a sample of 68 litters
sampled over four years, Prodhl et al. (1998; see also
Prodhl et al. this volume) found six instances where
a male sired more than one litter in the same year (out
of 138 males sampled, 46 of which were reproductively
successful). Interestingly, as in so many other species,
observations of reproductive behavior did not necessarily predict reproductive success. Although observations of reproductive pairs allowed identification
of reproductively active individuals, some of which
were later identified as a parent of a particular litter,
in no case were both members of a pair identified as
the most likely parents for a particular set of juveniles
(Prodhl et al. 1998; Loughry, Prodhl, McDonough,
Nelson, and Avise 1998). Paired males were larger than
unpaired males, but this had no effect on reproductive
success, which was random with respect to all aspects of

the phenotype that were measured (Loughry, Prodhl,
McDonough, Nelson, and Avise 1998). Thus, it seems
that pairing behavior was no guarantee of reproductive
success for the armadillos in this population, although
it remains to be seen whether this conclusion holds
over longer periods of time or in other populations.
Aside from the above, reproductive behavior remains unknown for most armadillo species, despite the
fact that other lines of evidence suggest that intriguing patterns may exist. For example, Cetica and Merani
(this volume) describe how aspects of sperm morphology indicate the potential importance of sperm competition in several species. Yet at present we know nothing
of how sperm competition influences male or female
mating strategies in any of them. Similarly, the threebanded armadillo, T. matacus, has one of the longest
penises relative to body size in any mammal (Meritt
1973c, 1976b), but we know nothing of the evolutionary forces driving the development of this remarkable
trait.
Affinitive behavior and kin selection
Arguably the most intriguing aspect of reproduction in
armadillos concerns the occurrence of obligate polyembryony in the seven species belonging to the genus
Dasypus (Galbreath 1985; Craig et al. 1997). In these
species, females produce a single fertilized egg. Upon
implantation in the uterus, the egg divides one or more
times, producing multiple embryos, so that females
always give birth to litters of genetically identical offspring (Enders this volume; Prodhl et al. this volume).
Although polyembryony has been confirmed genetically only in D. novemcinctus (Prodhl et al. 1996), the
high coefficient of relatedness among siblings has led
some to consider armadillos as ideal model systems
for the study of kin selection. For example, Dawkins
(1976), speaking specifically about D. novemcinctus,
wrote: Nine-banded armadillos are born in a litter of
identical quadruplets. . . . it has been pointed out that
some strong altruism is definitely to be expected, and
it would be well worth somebodys while going out to
South America to have a look.
Could armadillos be altruistic? Observations of affinitive behavior among juvenile armadillos suggest opportunities for the operation of kin selection, perhaps
via enhanced foraging success or increased predator
detection. Similarly, kin selection could occur among
adult armadillos, perhaps via reduced aggression and
increased tolerance, increased home-range overlap,
291
or sharing of burrows (see below) among siblings. Of

course, this assumes that littermates settle near enough
to one another so that interactions of the type favored
by kin selection might occur.
To find out what, if any, consequences polyembryony
has on armadillo population structure and behavior, we
conducted a long-term study of a D. novemcinctus population (contra Dawkins, our study site was in north
Florida, not South America). To date, all our results
have been negative (Loughry, Prodhl, McDonough,
and Avise 1998), suggesting that the production of
clonal offspring has not been exploited by kin selection in this species. There are several lines of evidence
that support this conclusion: first, few adults had any
siblings present in the population. Using microsatellite
DNA markers, Prodhl et al. (1996; see also Prodhl et
al. this volume) found that only 16 of 196 adults sampled
had at least one sibling present (these 16 animals were
divided into two groups of triplets and five groups of
twins). These results were subsequently confirmed by
long-term data on juvenile recruitment, which showed
that very few littermates were recruited into the population (Loughry and McDonough 2001), possibly because
of high juvenile mortality (McDonough and Loughry
1997a). Second, while juvenile siblings were often found
in tight spatial clusters where interactions among kin
might be possible, adult siblings were widely separated
from one another (mean distance between adult sibs =
455.9 m) suggesting low rates of interaction. Thus, for
adult armadillos in this population, it appears that opportunities for kin selection were minimal. Of course,
this does not preclude the possibility of indirect fitness
benefits from interactions among more distantly related
relatives, but this seems unlikely given the absence of
any obvious kin selection where it was most expected
(i.e., among littermates). Third, although affinitive behavior among juvenile littermates certainly provides
opportunities for kin to interact, kin selection requires
that individuals discriminate between relatives and
nonrelatives and bias their behavior preferentially toward relatives (Grafen 1990; Gamboa et al. 1991; Keller
1997). The results of kin discrimination tests indicated
that, although juveniles could distinguish between the
scents of siblings and nonsiblings (Loughry and McDonough 1994), when juveniles were placed in an arena
and allowed to interact with one another, there were
no kin-biases in their behavior (Loughry, Dwyer and
McDonough 1998). In fact, all interactions among juveniles were highly amicable. Assuming these lab experiments are representative of what occurs between
292
juveniles in the wild, it thus appears unlikely that kin

selection has shaped the behavior of either juvenile or
adult armadillos in this population.
Interestingly, the lack of any obvious kin selection in
D. novemcinctus is mirrored in other vertebrates that
reproduce clonally. Polyembryony results in offspring
that are genetically identical to one another, but not to
either parent. Parthenogenetic reproduction, in which
offspring genetically identical to the mother are produced from unfertilized eggs, is found in the unisexual
fishes and several groups of reptiles and amphibians
(Cole 1975; Cuellar 1977; Vrijenhoek et al. 1989). As with
the polyembryonic armadillos, these other clonal vertebrates have long been considered ideal systems for the
study of kin selection (e.g., Alexander 1979; Bell 1997).
Yet, to our knowledge, there is no positive evidence of
kin selection occurring in any clonal vertebrate.
Assuming there are reasons that kin selection has not
been important in clonal vertebrates (see Loughry et al.
2005), why do these species produce clonal offspring?
Most reviews have concluded that clonality probably
evolved to increase parental fitness, perhaps by allowing rapid exploitation of particular environments or as a
means of bypassing certain physiological constraints on
reproduction (Cole 1975; Cuellar 1977, 1987; Vrijenhoek
1984; Lynch 1984; Galbreath 1985; Hughes and Cancino
1985; Gleeson et al. 1994; Craig et al. 1997). If true, then
clonality per se may be of little consequence. For example, Galbreath (1985) has argued that the anatomical region where female armadillos hold the fertilized
egg while they delay implantation is so small that it can
hold only a single egg (this position is controversial,
see Enders this volume). Thus, female fitness cannot be
increased by the production of additional eggs; the only
way to produce additional offspring is by having the
fertilized egg divide into multiple embryos upon implantation in the uterus. If so, then armadillo offspring
will be by necessity genetically identical, but strictly as a
consequence of increasing parental fitness, with little or
no subsequent evolutionary exploitation via kin selection.
Burrows
Because of their prevalence in the environment and
ease of measurement, burrows have been studied in
many armadillos (table 26.1). Burrows are used for
sleeping and as food shelters, refuges from predators or
inhospitable weather conditions, and birth chambers.
However, not all armadillos utilize burrows to the same
extent. For example, Cabassous spp. rarely seem to remain in or return to a previously dug burrow (Carter
and Encarnaao 1983).
Several aspects of burrows vary among species. For
example, Dasypus spp. often build nests within their
burrows. Nests are constructed by dragging vegetation
such as grass and leaves down into the burrow (Eisenberg 1961). In addition, D. novemcinctus has been observed constructing aboveground grass nests (Layne and
Waggener 1984; Bond et al. 2000; Platt and Rainwater
2003), perhaps to avoid flooding (Layne and Waggener
1984) or for thermoregulation during warmer weather
(Bond et al. 2000). However, nests have not been found
in the burrows of several other species (Ca. unicinctus,
Ca. tatouay, Cha. villosus, E. sexcinctus, P. maximus,
and Z. pichiy, Carter and Encarnaao 1983; Abba et al.
2005; Superina this volume). Similary, some species dig
burrows with only a single entrance, while others have
multiple openings (see Abba et al. 2005); some have
burrow entrances that face away from prevailing winds
(Carter and Encarnaao 1983; Gonzalez et al 2001; Abba
et al. 2005), but others do not (Zimmerman 1990; Platt
et al. 2004). In some cases, differences in burrow design
are unique enough to be species-specific, which can
be useful in assessing populations in areas or at times
of day when censusing of live animals is impractical
(Carter and Encarnaao 1983; McDonough et al. 2000;
Abba et al. 2005; Kinlaw 2006). Even so, we have yet
to fully understand the functional significance of these
interspecific differences in burrow features.
Burrow sharing
Although the evidence to date suggests kin selection
has not been an important factor in the evolution of
armadillo behavior (see above), there remain certain
aspects of behavior that could reveal kin-selected influences. Burrow sharing is one of these. Because of the
extensive overlap in their large home ranges (see above)
armadillos may often use the same burrows, but usually
not at the same time (Schell 1994). Burrow sharing is
unique in that two or more armadillos occupy the same
burrow at the same time.
Anecdotal reports of burrow sharing are widespread
(table 26.3). Some reports are impressive, with up to 12
individuals found in a burrow for T. matacus (Meritt
1973c) and D. novemcinctus (Lehmann, cited in Kalmbach 1943). In the latter case, the 12 animals were juveniles, which was taken as evidence of possible burrow
sharing by adult females. Further studies have established that adult females do share burrows, in both D.
novemcinctus (Bushnell 1952; Herbst and Redford 1991)

and D. sabanicola (females in this species may share
burrows in the absence of young, Pacheco and Naranjo 1978). Burrow sharing also may include co-occupancy by adult males and females (McDonough 1997;
Vaughan and Shoenfelder 1999), juveniles of both sexes
(these may or may not be sibling groups, Kalmbach
1943; Bushnell 1952; McDonough 2000), and adults
of unknown sex (Taber 1945; Meritt 1973c; Breece and
Dusi 1985; Schell 1994).
Why share a burrow? Certainly, harsh environmental conditions could be important. For example, armadillos have been reported sharing burrows during periods of extreme cold, probably for thermoregulatory
benefits (Meritt 1973c), and also during severe droughts
(Taber 1945). Burrow sharing between adult males and
females may be an aspect of courtship, in which males
sleep in the entrance to a burrow to maintain proximity
to the female or to prevent usurpation by other males
(McDonough 2000). In species with litter sizes >1, littermates may share burrows for some portion of their
first year (e.g., D. novemcinctus: Bushnell 1952; Galbreath 1982; Loughry and McDonough 1998b). Possible, but as yet untested, explanations for this include
enhanced thermoregulation, increased predator detection, or foraging advantages. The same benefits may apply in instances where nonlittermates share burrows as
well. However, we need to move beyond anecdotal accounts if we are ever to really understand this phenomenon. More detailed information about the seasonal
occurrence of burrow sharing and, most crucially, the
293
identity of co-occupants (age, sex, genetic relatedness,

etc.) is necessary before we can evaluate any functional
hypotheses.
Conclusions
The behavior of most armadillos remains poorly known,
with virtually no behavioral information available for
18 of 21 species. For many years the conventional wisdom in behavioral ecology has been that we need more
long-term studies where individually marked animals
are followed from birth to death (e.g., Clutton-Brock et
al. 1982; Hoogland 1995). With the possible exception
of D. novemcinctus, this has not been achieved in any
armadillo. Our understanding of this fascinating group
is severely hampered by the lack of basic ecological and
behavioral data on so many species. Given the current
conservation status of many of these taxa (Aguiar and
Fonseca this volume), it is critical that we begin accumulating such data quickly.
Acknowledgments
Thanks to Marcelo Cassini and Sergio Vizcano for
comments on an earlier draft of this chapter. Our research on armadillos has been supported over the years
by various Faculty Research Awards from Valdosta
State University, the American Philosophical Society,
the Organization of American States, the National Geographic Society, and Earthwatch.
27
Xenarthrans of the Paraguayan Chaco
Dennis A. Meritt Jr.
the north of the Paraguayan Chaco is 500 mm or less,

with increasing amounts as one moves south. Spring is
the rainy season, although the period from October to
January may see only intermittent rainfall. The Chaco
experiences prolonged periods of drought when rains
fail to come or deposit only minimal amounts of water. The period of 20022005 was the most recent such
event. An unusual and remarkable feature of the Chaco
is the near absence of stones or rock. Except for one
area in the north and west of the country, Cerro Len,
none are present (Campos et al. 2004).
Redford et al. (1990) identified the importance of
the Chaco ecosystem and described its biodiversity,
comparing and contrasting it with studies of tropical
ecosystems. Their point, a point well made, was that the
Chaco is an environment extremely rich in biodiversity, but largely ignored in terms of long-term biological
studies. Further, there were no structured conservation
efforts in place at the time of their observations. Since
then some attempt has been made to initiate conservation action programs, but at best one could describe
most of these as short term and marginally successful.
A notable exception is the work at Proyecto Tagua at
Toledo Boquern, Chaco, Paraguay. As part of a longterm conservation commitment to the endangered
Chacoan peccary, Catagonus wagneri, observations and
studies of various mammals, including xenarthrans,
have taken place in the central Chaco at Fortn Toledo,
30 km from the Mennonite community of Filadelphia.
Proyecto Tagua is the site of the captive management
and conservation action program for this peccary. Toledo is centrally located and provides an excellent location for a base of operations and research site, as well
as a logical jumping-off point for field investigations in
Resumen
Los xenartros presentes en el Chaco de Paraguay fueron
estudiados en el curso de muchas visitas a tal ecosistema.
Observaciones oportunistas proveyeron informacin
sobre la presencia o ausencia, uso del hbitat, patrones
de actividad y uso por los humanos de estos animales.
Se necesitan estudios ms acabados para completar los
huecos en el conocimiento de varias especies.
Resumo
Os xenarthras presentes no Chaco do Paraguai foram
estudados no curso de muitas visitas a este ecossistema.
Observaes oportunas produziram informaes sobre
a ocorrncia, o habitat utilizado, os padres de atividades e uso pelos humanos desses animais. Se faz necessrio estudos mais minuciosos para completar a ausncia de conhecimento sobre varias espcies.
Introduction
The Chaco of Paraguay occupies the western half of the
country and is a seasonally wet, seasonally dry plain
that drains south toward the Paraguay River basin. The
land is driest in the north near Bolivia, with vegetation
that is xeric adapted. The southern limits of the Chaco
are almost swamplike in character for a significant part
of the year as water drains from the north during and
following the rainy season. Short (1975) has described
the physical features of the Chaco from Bolivia to
northern Argentina as well as the adjacent area of Brazil. It is subtropical thorn forest and thorn scrub and is
part of the semiarid core of South America. Rainfall in
294
the states of Presidente Hayes, Boquern, Alto Chaco,

and points beyond.
Methods
The observations reported here were recorded opportunistically over the course of many visits to the Chaco.
While not particularly detailed or quantitative, my hope
is that these observations will provide additional information related to a range of xenarthran species in the
central Chaco. Specifics follow in the species accounts.
It should be noted that there are no sloths in the Chaco,
so no information on them is presented.
During various field activities in the Chaco, signs of
xenarthrans or their presence were noted whenever encountered. Beginning in 1980 and continuing through
the present, these field observations were conducted
for 24 weeks on a nearly annual basis. Observations
were carried out at various times of year including winter and summer, but the majority were obtained in the
austral spring, a time when fieldwork in the extreme
environment of the Chaco is more practical. During
spring, temperatures are more moderate and travel to
remote areas is possible. With the onset of prolonged
rains, travel is possible only on paved roads and these
are all but absent off of the Trans-Chaco Highway (Ruta
Transchaco).
Field activities consisted of walking transects, driving
dirt two-tracks, and censusing unpaved main roadways.
When encountered, living animals were pursued and
attempts made to capture them. Animals were checked
for general condition and the presence of injuries or
scars, sexed, occasionally weighed and measured, and
then released at the capture site. Some animals were
held in captivity for brief periods prior to release. Others, such as several (n = 8) individual La Plata threebanded armadillos (Tolypeutes matacus), were retained
for inclusion in captive management programs.
Foraging locations, temporary nests, burrows, tracks,
and remnants of food all provided evidence of the presence of animals. Based on the size, number, and depth
of various foraging sites or excavations it was possible
to determine the species active at a particular site. Excavations provided an indication of current or recent
use. In some cases small twigs, grass stems, or pieces
of thread were placed at burrow entrances and checked
later in an attempt to gauge activity. Sand traps were
made outside of burrows or along known animal trails
to record footprints and tail drag marks of various ani-
295
mals, including xenarthrans. When tracks or footprints

were found, attempts were made to identify these to
species. Loose or compacted soil, footprints, and spoor
outside of burrows served as additional indicators of
xenarthran presence and recency of activity.
Remnants of various xenarthrans were observed,
photographed, collected, or recorded whenever encountered. These included animals offered along roadsides as meat, roadkill along the Ruta Transchaco and
roadways within the Mennonite colonies, hunter-generated skins or carapaces at campsites, and skeletal
material resulting from subsistence hunting or other
mortality. Notations were made of date, time, location,
and circumstances. Taxonomy and order follow Redford and Eisenberg (1992). Discussions of xenarthran
taxonomy and natural history may be found in Wetzel
(1982, 1985a) and Aguiar and Fonseca (this volume).
Species accounts
Giant anteater (Myrmecophaga tridactyla)
Giant anteaters were encountered during daylight hours
in savanna and thorn forest habitat. Footprints were
observed in and around water holes or along seasonal
roadside pond edges. Live animals were observed during the day on secondary roads, crossing off-road two
tracks, and within estancias with large land holdings of
several thousand hectares or more. All living animals
observed in the central Chaco were adult-sized animals
without accompanying offspring.
Giant anteaters were infrequently seen as roadkill
along the Ruta Transchaco. As a recent example, in
June of 2005 we observed three adult individuals while
traveling from Asuncin to Fortn Toledo in the area
of km 200250. One noteworthy observation is that
the numerous black and turkey vultures do not feed on
the carcasses of giant anteater roadkills. Local folklore
explains that the skin is too tough and dense for the
vultures to tear through. Even old roadside carcasses
undergoing obvious putrefaction remained untouched
by these avian scavengers. No other mammal carcasses
in the Chaco are treated this way in my experience.
These anteaters continue to be the targets of weekend hunters, usually those traveling to the Chaco from
Asuncin for weekend or holiday recreational activities.
Giant anteaters are shot opportunistically. Their thick,
tough skins are desirable in the leather trade for use in
296
D. A. Meritt Jr.
manufacturing leather products that require durability,

for example, chaps and other riding accessories.
Aside from humans, two possible natural Chacoan
predators include the puma (Felis concolor) and jaguar
(Panthera onca). Studies in Paraguay modeled after
those of Shaw et al. (1985, 1987) in Brazil would provide
an opportunity to compare and contrast life history
traits between widely separated areas of the Chaco.
Southern tamandua (Tamandua tetradactyla)
Tamanduas were encountered more frequently in the
lower Chaco, where it is moister on a year-round basis
than in the central Chaco of Toledo. They were usually
seen in areas with more dependable sources of moisture
and associated vegetation. Tamandua presence also was
likely related to the increased presence of termite nests
in these habitats.
Tamanduas were occasionally seen as roadkill and
their skins were offered for sale by indigenous peoples,
but not with any degree of regularity. This species may
be eaten, but there were no observations of its meat
being offered for sale. The musky smell of the anteater
may be an impediment to its use as a food item.
There is a need for a population assessment of this
anteater in all Chaco habitats. The use of telemetry to
record daily activity and home range is desirable. The
food habits of this species in this habitat are unrecorded
and require study.
Chacoan naked-tail armadillo (Cabassous chacoensis)
This fossorial armadillo is rarely encountered even in
suitable Chaco habitat. Based on my observations, suitable habitat consists of open thorn forest or thorn scrub
with more porous nonclay soil. In the area around Toledo, two individuals have been seen and captured in
a 20-year period. One capture, in 2004, was a female
that produced a single offspring. Because of its nocturnal activity cycle, strong burrowing tendencies and low
population numbers compared with other armadillo
species, this armadillo is rarely encountered. However,
some encounters may go unrecognized because nakedtail armadillos are sometimes incorrectly identified as
baby Priodontes (Meritt 2006). Our understanding of
this species and related forms of Cabassous would benefit from a detailed in situ natural history study.
Summary observations of natural history and behavior for captive examples of Cabassous spp. have been
provided by Meritt (1985b). Neris et al. (2002) report
that this species is not regularly used as a staple food

item in the diets of some indigenous groups.
Small hairy armadillo (Chaetophractus vellerosus)
This diminutive armadillo, sometimes referred to as the
screaming armadillo because of its vocal capabilities, is
more common in the central Chaco today than it was 10
years ago. This species appears to have adjusted to and
benefited from some of the land conversion schemes
currently being practiced in the Chaco, particularly
in the region of the Mennonite colonies (interestingly,
this does not seem to be the case in other locations, see
Abba and Cassini this volume). While this armadillo
was only rarely encountered on roadways in the mid to
late 1990s, it is seen with increasing frequency today. It
is by far the most common armadillo encountered at
night on or adjacent to earthen roadways within the
Mennonite colonies.
Observations beginning in 2000 indicate an increased presence of this species, particularly in agricultural areas when the soil is worked and crops are
planted. These crop areas have increased over time,
usually at the expense of grassland planted with introduced grasses for cattle ranching. Generally speaking,
these fields are composed of loose soil, usually sandy
loam. The armadillo may preferentially select this soil
type and these converted agricultural areas. Further
quantitative studies are needed to verify this observation.
Studies of this species in Argentina (Greegor
1980a,b, 1985; Abba and Cassini this volume) provide
details about home range, activity cycle, and diet, but it
remains to be seen whether these findings can be generalized to other areas such as the Chaco.
Large hairy armadillo (Chaetophractus villosus)
The largest of the hairy armadillos is day active as opposed to the nocturnal life style of its smaller relative.
I have always encountered it during the day, usually
alertly foraging within open habitat adjacent to thorn
forest patches or areas of thorn scrub. Unlike the small
hairy armadillo, which has a strong flight reaction and
flees rapidly, this species in this habitat in the central
Chaco of Paraguay demonstrates a reduced reaction.
When encountered in nature, the large hairy armadillo
may pause, stand on its hind legs and sniff the air before changing direction or taking evasive action by running away. This species appears to be holding its own,
with populations neither increasing nor decreasing in

Chaco habitat as judged by long-term encounter observations.
Chacoan fairy armadillo (Chlamyphorus retusus)
This, the most fossorial of the Paraguayan armadillo
species, has been described as patchy in its distribution and more often encountered in sandy plains or
grasslands (Neris et al. 2002). However, in more than
20 years of travel in Paraguay, I have not encountered
it in any habitat. The only specimens seen were those
deposited at the former site of the Paraguayan Biological Inventory. There were two museum specimens, an
excellent taxidermy mount, and an individual in alcohol. Measurements of these individuals were not available. Readers should be aware that some authors (e.g.,
Gardner 2005) have placed this species in the genus
Calyptophractus.
Inquiries related to this species in presumed suitable
habitat are usually met with folktales that are counter to
those related in Neris et al. (2002), who say that indigenous groups consume the meat and carapace. Chaco
indigenous groups treat this armadillo as an omen of an
impending death. To see or to otherwise encounter this
mammal is bad luck and means that someone in the
immediate family of the person viewing it will soon die
(Cullar 2001; see also the chapters by Noss et al. and
Aguiar and Fonseca in this book). The animal is most
often encountered, albeit rarely, during plowing associated with agricultural operations (Meritt 1985c).
Nine-banded armadillo (Dasypus novemcinctus)
This armadillo is widespread and common in most
parts of the Chaco except for the most xeric areas. It is
encountered in almost every habitat type but appears to
be more common in the somewhat moister grasslands
and shrubby patches, in areas that have been converted
for use by cattle. The wooded boundaries required by
law between areas converted for agricultural purposes
are a prime location for D. novemcinctus. These 100
300 m wide natural areas serve as corridors between
plots and provide suitable habitat for this species as well
as a range of other life forms. Burrows, tracks, and signs
of foraging may be found here.
Chaqueos and indigenous communities rely on this
species as a food staple. Along the Trans-Chaco Highway, in areas that serve as temporary or transitory camps
for indigenous communities, it is fairly common to see
297
the freshly cleaned carcasses of this armadillo hung on

roadside vending poles. Travelers buy these and other
native animal sources of meat. In the southeast of Bolivia the nine-banded armadillo represents the single
largest armadillo resource available to and consumed
by indigenous community hunters (Cullar this volume; Noss et al. this volume). It would appear that the
situation in Paraguay is similar, both in terms of availability and consumption. There are no detailed studies
in place to quantify this and such studies are critical for
any self-sustaining management plan for this species.
The methods reported in Cullar (this volume) could
serve as a model for future Paraguayan studies.
Details of the ecology of this armadillo may be found
in the species account of McBee and Baker (1982) as
well as in McDonough and Loughry (this volume).
Six-banded armadillo (Euphractus sexcinctus)
Redford and Wetzel (1985) provide natural history
details of this species. In the Chaco this armadillo is
sometimes day active and other times active at night,
depending on time of year and whether or not it is
hunted. Six-banded armadillos are regularly encountered in open thorn scrub and thorn forest habitat in
the central Chaco, usually in areas that have not been
affected by land conversion or human activity. They are
most frequently encountered in open shrubby habitat while foraging. This armadillo is an opportunistic
feeder, taking fruits, seeds, tubers, insects, reptiles, amphibians, small mammals, eggs and carrion. Folklore
depicts these large, strong armadillos as digging into
graves to feed on corpses. This may derive from observations of this species feeding on the decomposing
carcasses of domestic livestock. Because of this habit
and also the armadillos characteristic musky odor,
some indigenous groups refuse to use it as a food item.
It has been referred to as the pig armadillo because of
its wide-ranging food selection.
When encountered the animal pauses, stands nearly
erect on its hind legs, forelegs hanging loosely, sniffing
the air while moving its nose and head from side to
side. It may then resume foraging or may suddenly bolt,
running quickly in the direction of thorny vegetation
or an available burrow. If captured it becomes agitated,
urinating and defecating on its captor while struggling
and striking out with its strong and elongated foreclaws.
Six-banded armadillos growl and will not hesitate to attack or bite during capture.
298
D. A. Meritt Jr.
Giant armadillo (Priodontes maximus)

This is the largest of the armadillos. Its distribution is
patchy and population numbers are unknown throughout its range. In the Chaco, burrows of this species are
recognizably distinct and cannot be confused with any
other armadillo (Carter 1983; Carter and Encarnaao
1983). Size of the entrance, depth of the burrow, amount
of material excavated from the burrow, and the recent
destruction of termite or other insect nests indicate the
presence of Priodontes (Meritt unpublished observations). It has been observed in the central Chaco in xeric habitat with clay soil, in habitat adjacent to seasonal
rivers and streams in more loamy soil, and in undisturbed thorn forest with sandy soil. Rainfall amounts
in these three geographically distinct areas are variable
from year to year, ranging from 375 to 500 mm.
Wherever Priodontes is found it is a highly prized
food item, representing a significant source of protein to indigenous hunters (Meritt 1973a,b). Based on
hunter comments this is as true today as it was 30 years
ago (Meritt unpublished observations, see also Aguiar
and Fonseca this volume; Noss et al. this volume).
Meritt (2006) provides a list of topic questions related to the natural history and behavior of giant armadillos. The main goal was to identify aspects of Priodontes natural history that require detailed study both
in nature and in captivity, such as activity patterns, food
requirements, behavior, and other traits that are either
poorly understood or currently unknown.
La Plata three-banded armadillo (Tolypeutes
matacus)
This is the only armadillo found in the Chaco that is
capable of rolling into a ball. Individuals in this defensive posture tuck in limbs, head, and tail, forming a
near-perfect spherical form. Tolypeutes may be found
throughout the central Chaco and can be encountered
at night along dirt roads or in open thorn bush habitat or in naturally occurring grassland patches. During
sunny warm winter days this species, especially young
adult individuals, may be active during daylight hours
as well. During periods of extreme cold, individuals
may form communal associations, sharing shallow
grass-lined nests. Up to six inactive individuals have
been found in these nest sites. These congregations are
temporary, lasting one to four days (Meritt unpublished
observations). With increased ambient temperature in-
dividuals leave these locations and resume normal activities.

Tolypeutes appears restricted in its southern distribution by temperature and its inability to store appreciable amounts of fat to use as an energy resource. Even
individuals maintained in captivity for long periods
and provided with abundant nutritious food did not
develop internal deposits of fat (Meritt unpublished
observations). This may be an adaptation related to the
ability of Tolypeutes to escape predation by rolling into
a ball, which would be difficult with large layers of fat
present.
This species in the central Chaco is a fast and elusive runner and can often escape human hunters with
speed and directional changes. Tolypeutes pursued in
roadside habitat in the central Chaco were more elusive
and difficult to capture than the small hairy armadillo
Cha. vellerosus. This armadillo is a desirable food item
throughout its range and empty shells may be found
around old campfires in the interior of the central
Chaco. It is usually cooked in its carapace by hunters.
Empty shells bearing the chew marks of puma have
been found in the Chaco. No doubt it also falls prey to
jaguars.
Tolypeutes sometimes appears in the legal animal
trade and is highly desired as a demonstration animal
in education programs, especially in zoos and wildlife
facilities. This species is easily maintained in captivity
and reproduces without difficulty, successfully raising
its single offspring (Meritt 1971b). Captive populations
in North America in recent years have been self-sustaining. This species of armadillo, if afforded habitat
protection in countries of origin, could be managed as
a self-sustaining harvestable resource.
Closing comments
As this volume amply demonstrates for many species
in a variety of locales, our knowledge of xenarthrans
in the Chaco of Paraguay would benefit from more detailed studies of their ecology and evolution, especially
for those species whose life history traits are poorly understood or unknown. Studies such as those of Abba et
al. (2005) on burrow architecture and frequency could
serve as models for studying several species of Chacoan
armadillos. Using the methodology described by Cullar (this volume), insights into xenarthran abundance
and use by indigenous communities could be gained.
Roadkill tallies conducted during travel on the TransChaco Highway and on major roads to and within
the Chaco and around various Mennonite colonies,
especially those of Filadelphia, Loma Plata, and Neuland, could provide valuable data on abundance and
distribution, seasonal presence of dispersing young,
physical and reproductive condition, nutritional state,
body size, sex, and level of maturity. Such a study has
been conducted recently in the Chaco on the maned
wolf (Chrysocyon brachyurus, Meritt and Campos unpublished observations) and could easily be done on a
number of xenarthrans (cf. Loughry and McDonough
1996; Inbar and Mayer 1999). The use of new technologies allowing mark-recapture and tracking of various
species shows promise for application to xenarthrans
and could provide much-needed and valuable home
range and activity cycle information.
Paraguayan students of natural history and ecology continue to be encouraged to conduct such studies
as time and circumstances permit. Regrettably, while
there is interest, the availability of funding and release
time from campus-based university studies has not yet
allowed many studies to be undertaken.
299
Acknowledgments
Permission to carry out scientific studies in Paraguay
was provided through the courtesy and authority of the
Ministerio, Secretaria del Medio Ambiente. Partial support for this work was provided by the Chacoan Peccary Species Survival Program (SSP) of the American
Zoo and Aquarium Association (AZA) through funds
provided by M. Scott for Proyecto Tagua. A grant
from the American Philosophical Society allowed an
initial field effort in northern Argentina and Paraguay.
In Paraguay, Jakob and Maria Unger, Juan Campos, and
Ivan Benitez, as well as others provided assistance and
shared knowledge. I am grateful to each of these organizations or individuals for their contributions. Kurt
Benirschke provided the stimulus for Chacoan studies.
I am especially grateful to him for his commitment to
the Chaco and Paraguay. This chapter has benefited
from the comments of Jim Loughry, Sergio Vizcano,
and two anonymous reviewers.
28
Ecology and conservation of three species of armadillos
in the Pampas region, Argentina
Agustn M. Abba and Marcelo H. Cassini
dancia se ven afectadas por la presencia de perros y se

observa una relacin positiva al aumentar la distancia
de los centros urbanos.
Resumen
En este captulo se realiza un resumen de la informacin
disponible sobre ecologa, uso y conservacin de tres
especies simptricas de armadillos de la regin pampeana (Chaetophractus villosus, Cha. vellerosus y Dasypus hybridus) y se plantean futuras estrategias de estudio a seguir. Cha. villosus es el armadillo ms abundante
en la regin y se lo encuentra en todos los ambientes
disponibles. Las caractersticas de esta especie son las
siguientes: (1) Desarrolla su actividad principalmente
durante la tarde/noche; (2) construye cuevas de gran
tamao con una sola entrada y realiza nidos de pasto
en su interior; (3) su dieta es omnvora, contando como
tems ms frecuentes adultos y larvas de colepteros y
lepidpteros, plantas y carroa. Al igual que las otras
dos especies de armadillos, Cha. villosus es cazada para
consumo. Sin embargo, en la regin pampeana, es la
nica especie que se ve afectada directamente por la
intensidad de la caza. La otra especie del gnero, Cha.
vellerosus, comparte con Cha. villosus el gran desarrollo
de sus cuevas, la construccin de nidos de pasto y la
dieta. Se diferencia en que las cuevas de Cha. vellerosus
tienen varias entradas, su distribucin depende de reas
de suelos arenosos o de conchilla y, aunque su patrn
de actividades no fue bien registrado en la regin pampeana, datos de otras regiones postulan que vara segn
la estacin, siendo diurno en invierno y nocturno en
verano. D. hybridus es tpica de pastizales naturales o
seminaturales; sus actividades se desarrollan durante
las horas del da; construye cuevas de desarrollo medio,
con una sola entrada y nidos de pasto en su interior;
datos asistemticos de su dieta sugieren una marcada
tendencia a la mirmecofagia. Su distribucin y abun-
Resumo
Neste captulo, sumarizamos as informaes disponveis
sobre a ecologia, uso e conservao de trs espcies
simptricas de tatus da regio pampeana (Chaetophractus villosus, Cha. vellerosus e Dasypus hybridus), como
tambm postulamos estratgias futuras para estudos
subseqentes. Cha. villosus o tatu mais abundante
nesta regio, sendo encontrado em todos os ambientes
disponveis: (1) ativo principalmente durante o crepsculo e noite; (2) constri tocas ramificadas e extensas, com apenas uma entrada, e com ninhos feitos com
grama; (3) onvoro, tendo como presas mais comuns
adultos e larvas de colepteros e lepidpteros, plantas
e tubrculos. Como as outras duas espcies, Cha. villosus caado para consumo e a nica espcie afetada diretamente pela intensa caa presente na regio
pampeana. A outra espcie do gnero, Cha. vellerosus,
compartilha com Cha. villosus o amplo desenvolvimento das tocas com ninhos de grama e sua dieta. Suas
tocas diferem por possurem diversas entradas, e pela
localizao em solos arenosos ou calcarenitos. Apesar
de seus padres de atividade diria no se encontrarem
bem descritos para a regio, dados de outras regies
sugerem uma variao sazonal, diurno no inverno e
noturno no vero. D. hybridus uma espcie tpica de
pastos naturais e semi-naturais. Ativo durante o dia,
constri tocas de tamanho intermedirio, com apenas
uma entrada e um ninho. Dados dispersos sugerem que
a sua dieta principalmente mirmecfaga. Sua distri300
Ecology and conservation of armadillos in the Pampas region, Argentina
buio e abundncia so afetadas pela presena de ces,

e uma relao positiva encontrada entre os sinais de
sua presena e a distncia de centros urbanos.
Introduction
The Pampas grassland is the most modified ecoregion
of southern South America. The original ecosystem has
almost completely disappeared and been replaced by a
matrix of pastures and crops (Bertonatti and Corcuera
2000). The Pampas plains are also one of the least protected areas in Argentina. It is the only type of ecosystem in the country without federal protection within
the framework of the National Park Administration. As
a consequence, wildlife conservation strategies must be
designed with the assumption that animals will have to
survive in human-modified habitats and in interaction
with rural activities.
Armadillos are one of the characteristic groups of
mammals in the Pampas. There are 31 species of Xenarthra, and 38% of them are threatened with extinction
(Aguiar and Fonseca this volume). This percentage is
higher than those of other more charismatic mammalian orders, such as carnivores (27%) and cetaceans
(33%, Mace and Balmford 2000). In this study, we investigated three species of armadillos with different
conservation status. Chaetophractus villosus (hairy armadillo) is a widespread and abundant taxon that is frequently persecuted as a pest. Generally speaking, Cha.
vellerosus (little hairy armadillo) is also a species with
few conservation concerns. However, a small, isolated
population of this species in east Buenos Aires province
may deserve special attention because it occupies such
a small area with intensive human activity. Dasypus hybridus (southern long-nosed armadillo) is considered
near threatened, both globally (Aguiar and Fonseca
this volume) and locally (Daz and Ojeda 2000), because it is likely to qualify as endangered in the near
future.
The objectives of this chapter are to: (1) review fundamental aspects of the ecology of Cha. villosus, Cha.
vellerosus, and D. hybridus; (2) analyze the conservation
status of the three species in the Pampas region of Argentina; and (3) propose priorities for research that will
provide key information required for an action plan for
armadillo conservation in this region.
Most information available on the ecology of South
American armadillos is based on anecdotal data or
analyses of small data sets from restricted areas (Redford 1985a; Greegor 1980a,b, 1985; see also McDonough
301
and Loughry this volume). Therefore, in order to generate this review, we utilized information from a variety
of sources including standard peer-reviewed publications, conference abstracts, and personal communications as well as local and occasional publications.
Distribution
Complete range maps for all three species can be found
in the chapter by Aguiar and Fonseca (this volume).
Cha. villosus has the widest distribution of the three
species (Braun and Daz 1999; Deferrari et al. 2002);
within Argentina, it occupies most of the country, with
the exception of the east central region (figure 28.1).
Cha. vellerosus is found in central Argentina (figure
28.1). Wetzel (1985a,b) described this species for Patagonia based on an old specimen from the Museum Bernardino Rivadavia, but the presence of the species in
this region has not been verified recently. In the Pampas, it is found in eastern Buenos Aires province and
in one small, isolated spot along the Rio de La Plata
coast.
D. hybridus has the most restricted distribution of
the three species. In Argentina, it was described for the
provinces of Mendoza, Rio Negro, Misiones (Cabrera
1957; Redford and Eisenberg 1992), Catamarca (Braun
and Daz 1999), and Jujuy (Daz 2000). However, there
is no clear evidence of its occurrence in these provinces.
In Buenos Aires province, D. hybridus occupies mainly
the eastern sector and is not present in the south (figure
28.1).
Habitat use
Cha. villosus occupies a wide variety of environments.
In Argentina, it appears to prefer grasslands and can
tolerate highly modified habitats (Cabrera and Yepes
1940; Galliari et al. 1991; Bolkovic et al. 1999; Casanave
et al. 2003; Mapelli and Grierson 2003; Abba et al. 2005,
2007; see table 28.1). The same appears true in other
parts of South America as well (e.g., Paraguay, see Meritt this volume).
The other hairy armadillo, Cha. vellerosus, occupies
mainly arid and semiarid environments where the soil
is not hard (Crespo 1944; Cabrera 1957; Wetzel 1982;
Greegor 1980a, 1985; table 28.1). On the coast of Buenos Aires, it is associated with shell ridges (Crespo
1974; Carlini and Vizcano 1987; Abba et al. 2004). Estimates of home-range sizes have been reported from
Andalgal (Catamarca province) as 3.44.6 ha (Gree-
302
A. M. Abba and M. H. Cassini
Figure 28.1. Distribution

of three species
of armadillos in the
Pampas region.
Table 28.1. Habitat Use by Three Species of Armadillos
Environment
D. hybridus
Cha. vellerosus
Cha. villosus
Grassland
Wooded
Agricultural
Semi-natural
Suburban
++
+
++
+
-
+
++
+
+
-
++
+
++
++
++
Note: ++ = used for all activities, + = used for some activities,

- = not used; see the text for sources.
gor 1980a, 1985), and 1.05.3 ha in Magdalena (Buenos

Aires province, Glaz and Carlini 1999).
In contrast to the other two species, D. hybridus is
found mostly in open areas covered by grasslands (Redford and Eisenberg 1992; Wetzel 1985a,b; Bolkovic et al.
1999; Gonzles et al. 2001; Casanave et al. 2003; Abba
et al. 2007; table 28.1). Bolkovic et al. (1999) and Abba
(unpublished data) recorded this species in agrosystems.
Activity patterns
Cha. villosus appears not to have a fixed daily cycle of
activities (table 28.2). Instead, the daily activity period
may depend on region and season. Poljak et al. (2004)
conducted a study in Pipinas (Buenos Aires province)
between August and October and found the species to
be most active between 14:00 and 18:00 h. In the northeast of Buenos Aires province, Abba (unpublished
data) observed mainly nocturnal activity. In contrast,
Casanave et al. (2003) observed activity during almost
the entire day, while in Paraguay, Meritt (this volume)

reported that all his observations of active animals occurred during daytime. Parera (2002) suggests this species is more diurnal in winter than in summer.
Similar to Pareras (2002) suggestion for Cha. villosus, Greegor (1980a, 1985) reported that Cha. vellerosus
was nocturnal in summer (22:0006:00 h) and diurnal
in winter (09:0018:00 h), relating activity with ambient temperature (table 28.2).
The few studies conducted on D. hybridus concur
that it is diurnal, with most activity during the morning (Casanave et al. 2003; Abba unpublished data; table
28.2).
Digging habits
Xenarthrans are imperfect homeotherms (Grass 1955;
Boily 2002) and construction of home structures, that
is, burrows, could be a consequence of the limited thermoregulatory capability of these animals (McNab 1980,
1985).
In a recent paper, Abba et al. (2005) conducted a detailed description of the burrows of Cha. villosus (see
also table 28.3 and figure 28.2). Burrows consisted of
a descending gallery leading to a relatively horizontal
area, and could be classified into two types: simple and
complex. Simple structures were found in humic soil
and may function for temporary protection or as foraging sites. Complex structures were home burrows,
found in calcareous soil and in the highest areas that
Table 28.2. Summary of Activity Patterns of Three Species of

Armadillos
Period
Morning
Afternoon
Evening
Night
D. hybridus
Cha. vellerosus
Cha. villosus
++
++
+
-
-
+
+
-
+
+
++
++
Note: ++ = frequent, + = intermediate, = rare; see the text for

sources.
Table 28.3. Characteristics of the Burrows of Three Species of Armadillos

Characteristic
Number of entrances
Branches?
Nesting material
Average length (cm)
Average depth (cm)
Average width (cm)
Most frequent soil used
References
D. hybridus
Cha. vellerosus
Cha. villosus
1
No
Yes
118.8
43.4
15.3
Humic
3
1
Yes
Yes
37.5138
51
1013
Calcareous and sandy
1, 2
1
Yes
Yes
66131
41.38
15.820.7
Calcareous and humic
1, 4, 5
Sources: (1) Crespo 1944, (2) Greegor 1974, (3) Gonzlez et al. 2001, (4) Abba et al. 2005,
(5) Abba unpublished data.
Figure 28.2. Diagrams of the

burrows of the three species of
armadillos.
304
A. M. Abba and M. H. Cassini
did not flood. Burrow entrances appeared oriented so

as to take advantage of prevailing winds (Gonzlez et al.
2001; Abba et al. 2005).
Greegor (1985) described the use of burrows by Cha.
vellerosus. Burrows had multiple entrances and normally several burrows were visited sequentially during
foraging trips (table 28.3 and figure 28.2). Cha. vellerosus digs burrows by scraping soil back to the midbody,
lifting the body to draw the hind leg forward, and then
thrusting the dirt backward with the hind feet. Frequently, this results in a mound of soil at the entrance to
the burrow. This style of digging also results in the burrow being sealed as it is excavated, which might help to
maintain a more constant temperature and higher humidity in the burrow. Burrows excavated in sand dunes
were frequently several meters long and 12 m deep.
In Uruguay, Gonzlez et al. (2001) conducted the
most detailed study on the burrows of D. hybridus (table 28.3 and figure 28.2). Most burrows were located in
open areas. Mounds of dry grass were commonly found
next to the entrances. Most burrows were cylindrical
with a conical end, a single entrance, and no branches.
Diet
Cha. villosus was described as a carnivore-omnivore by
Redford (1985a), based on the comments of traditional
naturalists such as Darwin (1839), Krumbiegel (1940),
and Hudson (1997; see table 28.4 and also the review
by McDonough and Loughry this volume). In a recent
paper, Casanave et al. (2003) analyzed 30 fecal samples
and found invertebrates (43%), plant parts (43%), and a
few vertebrates (1%). This species is well known for eating carrion (Cabrera and Yepes 1940; Redford 1985a).
Cha. vellerosus was also described as a carnivoreomnivore by Redford (1985a). Greegor (1980b, 1985)
reported seasonal variation in stomach contents of
specimens collected at Andalgal, Catamarca province,
Argentina, with a preponderance of plants in winter
and vertebrates in summer (table 28.4). Recently, Soilbenzon et al. (2003) analyzed diet in the Pampas and
found the opposite trend. Insects made up the vast majority of the winter diet (82.3%), but a more diverse diet
was recorded in autumn: plants (37.2%), invertebrates
(21.9%), and vertebrates (34.8%); there were no samples
from summer.
Redford (1985a) proposed D. hybridus as a generalist
insectivore, consuming a high proportion of termites
and ants. Although this hypothesis has not been verified by systematic studies, anecdotal observations sup-
Table 28.4. Diet Composition of Three Species of Armadillos

Food items
D. hybridus
Termites
Ants
Coleoptera
Coleoptera larvae
Lepidoptera
Plants
Vertebrates
Carrion
++
++
+
?
?
?
-
-
Cha. vellerosus Cha. villosus

+
++
++
++
++
++
+
+
+
+
++
++
+
++
+
++
Note: ++ = frequent, + = present, - = absent, ? = no data; see the

text for sources.
port it (Gonzlez personal communication; Abba unpublished data).
Use and conservation

Historically, armadillos from the Pampas have been
used by local people as an important source of meat
(see Noss et al. this volume). As early as the eighteenth
century, the first explorers recognized this role. For example, in 1781, Azara (1998) said that the species of tatuejos were so good as food that they should be bred in
Europe. Similarly, DOrbigny (1999), in his books published between 1835 and 1847, mentions several times
the frequent consumption of armadillos in Buenos Aires by local people and Europeans.
Because of this extensive exploitation, several authors alerted readers to the risk of extinction of these
species, beginning more than 100 years ago. For example, in 1891, Hudson (1997) named four species of
armadillos for Buenos Aires province and postulated
that, with the exception of peludos (Cha. villosus), the
other species were at risk (D. hybridus, Tolypeutes matacus, and Zaedyus pichiy). Similarly, in 1781, Azara (1998)
said that armadillos will be soon exterminated for
the inhabitants that search for their meat. Fernndez
(1915b) pointed out that it is notorious how thousands
and thousands of these animals arrive to the markets of
the big cities, destroying indifferently males and pregnant females, the species will be soon extinguished.
The present conservation status of the species that
are the subject of this chapter varies from least concern
(Cha. vellerosus and Cha. villosus) to near threatened
(D. hybridus, Aguiar and Fonseca this volume). For all
three, human impacts on populations are likely to be a
growing concern. For example, Cha. villosus is so widespread and abundant it is frequently persecuted as a
pest. Massoia (1970) included it in a list of mammalian
Ecology and conservation of armadillos in the Pampas region, Argentina
species that destroyed soil and pastures of the Pampas,

and promoted eradication campaigns. Farmers of the
Pampas frequently perceive the species as a problem for
several reasons. For one, Cha. villosus can concentrate
in relatively large numbers in places where seed or grass
is stored and can destroy the containers, for example,
plastic silo bags. Other complaints are related to burrows and the risk for cattle and horses that may break
their legs by stepping in them.
In a recent study on how land use affects the distribution of these species in the Argentine Pampas, Abba
et al. (2007) surveyed their occurrence on 34 farms of
the eastern Pampas. They used the number of burrows
as an index of habitat use and related these to four local
variables (types of vegetation and soil, plant cover, and
height), and the following farm variables: distance to
main city; total area of the farm; number of fields; hectares of pastures, grassland, exotic woodland, and native
woodland; number of husbandry heads; geomorphology; intensity of hunting; use of herbicides; burning
practices; and number of dogs.
In comparisons between farms, the distribution of
each species correlated with just a few, and different,
aspects of land use and anthropic impact. Cha. vellerosus was the most specialized, using native woodlands
composed primarily of a typical xeric tree called tala.
However, it appeared to move to adjacent grassland
while searching for food. For all activities, this species
seemed dependent on areas with calcareous soil. D. hybridus was dependent on natural grassland and avoided
cultivated pastures. Its presence correlated positively
with distance to the city of La Plata and negatively with
the number of dogs present on the farms. Cha. villosus,
the most abundant species, was a generalist, found in
all four habitat types. Its distribution between farms
depended on the intensity of hunting on these properties. Estimates of niche overlap indicated relatively
low overlap between Cha. vellerosus and the other two
species, but high overlap between Cha. villosus and D.
hybridus.
This study by Abba et al. (2007) has potentially important conservation implications. For Cha. vellerosus,
305
it seems necessary to protect native woodlands. For D.

hybridus, conservation actions should be directed toward the reduction of dog populations and preservation of natural grasslands. For Cha. villosus, the most
important consideration may be the control of hunting
activity.
Conclusions
Echoing the conclusion of McDonough and Loughrys
(this volume) broader review of armadillo behavior and
ecology, our review of the literature on the three armadillos examined here indicates the need for more research on almost all aspects of their ecology. We would
like to suggest certain priorities for providing key information for a conservation action plan for the Pampas
region. First, it is necessary to evaluate exactly how dependent Cha. vellerosus is on talar woodland, including
the functional relationships between this species and
that habitat. An additional hypothesis to be tested is
that the distribution of Cha. vellerosus is strongly limited by soil hardness because of limitations in digging
abilities. Second, for Cha. villosus, social research is
necessary that investigates hunting by local people in all
its aspects: reasons, methods, spatial and temporal distribution, etc. Finally, because so little is known about
it in the wild, D. hybridus requires the most intense investigation. One objective should be to understand why
this species avoids farms near populated cities, and to
further understand the relationship between its distribution and the occurrence of wild dogs.
Acknowledgments
Special thanks to Sergio F. Vizcano and Jim Loughry
for their invitation and for critical editing of the manuscript, and to Colleen McDonough for useful comments. Financial support was granted by the Consejo
Nacional de Investigaciones Cientficas y Tcnicas
(CONICET), Universidad Nacional de La Plata, Universidad Nacional de Lujn, and Idea Wild.
29
Biology and ecology of armadillos in the Bolivian Chaco
Erika Cullar
Resumen
Resumo
En este captulo se presenta la informacin generada

entre 1996 y 2005 en el Chaco Boliviano. Los datos
resumidos aqu representan los primeros estudios en
esta regin del pas, enfocando aspectos de la biologa
y ecologa de armadillos. El captulo seala las meto
dologas y los resultados de trabajos sobre 7 especies
de armadillos Chaetophractus villosus, Cha. vellerosus,
Dasypus novemcinctus, Euphractus sexcinctus, Toly
peutes matacus, Priodontes maximus y Chlamyphorus
retusus. Las cinco primeras especies en la lista anterior han sido relativamente ms estudiadas que las dos
ltimas, debido a que los datos generados sobre estas
5 especies fueron colectados por el auto-monitoreo de
cacera, en el Chaco Boliviano. A partir de los armadi
llos cazados y reportados por los cazadores se han ana
lizado contenidos estomacales, tractos reproductivos,
estados de reproduccin anual y patrones de actividad
(asumiendo que los armadillos son cazados mientras
estn activos y fuera de sus cuevas). Por otro lado, tambin se han realizado estudios sobre abundancia, rea
de accin y patrones de actividad, empleando mtodos
ms especficos: capturas con perros, seguimiento de
hilo en carrete pegado en el caparazn, radio telemetra
y trampas cmara. La informacin generada por el
auto-monitoreo de cacera representa una importante
fuente de datos. Sin embargo, en lo posible, es necesario
combinar la informacin de los cazadores con estudios
especficos complementarios, con el objetivo de validar
la informacin existente y desarrollar un plan de estudios enfocado de cara a la conservacin y manejo de
este grupo de especies.
Este captulo apresenta informaes obtidas no Chaco

Boliviano entre os anos de 1996 e 2005. Os dados impetrados representam os primeiros estudos conduzidos
sobre a biologia e ecologia nesta regio do Pas e tambm do Chaco. Eu descrevo os mtodos de pesquisa
e resultados dos estudos de campo conduzidos com
sete espcies de tatus: Chaetophractus villosus, Cha. vellerosus, Dasypus novemcinctus, Euphractus sexcinctus,
Tolypeutes matacus, Priodontes maximus e Chlamyphorus retusus. As primeiras cinco espcies so melhor
conhecidas que as duas ltimas em razo dos abundantes dados de um esquema de monitoramento prprio
de caa, que foi implementado na rea de estudos.
Informao a partir dos caadores permitiu anlises
de contedo estomacal, tratos reprodutivos, ciclos reprodutivos anuais e padres de atividade (assumindose que os tatus foram capturados enquanto ativos e fora
de suas tocas). Estudos adicionais foram conduzidos
sobre a abundncia dos tatus, distribuio e padres
de atividade utilizando mtodos mais especficos, por
exemplo, captura com a ajuda de ces-de-caa, spooling, telemetria e cmeras. A informao gerada a partir do monitoramento prprio de caa representa uma
fonte de dados importante, porm esse mtodo deve ser
complementado e validado por estudos especficos que
avaliem questes especficas da pesquisa. Uma agenda
completa de pesquisa que integre o monitoramento
prprio de caa e pesquisa focada necessria para promover a conservao de longo-prazo e o manejo desse
grupo de espcies.
306
Introduction
The Bolivian Gran Chaco is the northernmost area of
the Chaco biome, and is one of the best conserved parts
of the Gran Chaco in South America (Taber et al. 1997;
The Nature Conservancy et al. 2005). The Chaco of Bolivia corresponds to a vast expanse of tropical dry forest and holds the countrys highest armadillo diversity,
with a total of eight species, six of which are hunted
(Noss et al. this volume). Despite their singular diversity and important conservation concerns (Aguiar and
Fonseca this volume), armadillos were not studied in
this region before 1996 and studies remain scarce a decade later (much the same can be said for other regions
of the Chaco, see Meritt this volume).
At greater geographic scales and with larger human
populations, hunter self-monitoring methods may offer
important advantages: the local population becomes involved in wildlife research and management activities,
and effectively expands the number of data collectors
beyond the capacity of a small number of researchers
(Noss et al. 2003; Noss et al. this volume). In some areas, hunters reports can constitute almost the only data
available on certain game species.
In general, very little is known of the natural history, population density, and ecology of armadillo
species across their range (review in McDonough and
Loughry this volume). Given that such information is a
prerequisite for the effective conservation and management of armadillos, a hunter self-monitoring program
represents a potentially important source of data. Using this method, results obtained in the Isoso region
of the Bolivian Chaco indicated that more than 60%
of mammals hunted in this area were armadillos (Noss
1998; Noss et al. this volume). Five species were intensively consumed: Chaetophractus villosus, Cha. vellerosus, Dasypus novemcinctus, Euphractus sexcinctus, and
Tolypeutes matacus. Chlamyphorus retusus was also
hunted, but only for reasons associated with traditional
beliefs, in the hope of warding off curses (Cullar 2001).
Priodontes maximus occurs only in the northeast part
of the Chaco and not in the Isoso.
Utilizing data from hunter self-monitoring as well
as other techniques, in this chapter I summarize information gathered on armadillos in the Bolivian Chaco,
especially from the Isoso region (see Noss et al. this volume for further details and figure 23.1 for a map of the
study area). Data are presented concerning reproduction, diet, habitat, home ranges, abundance, and activity patterns. I conclude by highlighting the limitations
307
and advantages of the methods used in the hope this

will promote further research, which will be needed for
conservation and management of the armadillo group.
Methods
As stated above, seven armadillo species occurred in
my study area. Because of their prevalence, five species (Cha. villosus, Cha. vellerosus, D. novemcinctus, E.
sexcinctus, and T. matacus) provided most of the data.
In addition, some data on home range, abundance and
activity were available for P. maximus. However, the
final species, Chl. retusus, remains largely unknown.
The only documentation existing for this species in my
study area was a description of six individuals killed
by hunters (Cullar 2001) and a direct observation of
one live animal in November 2004. The next sections
describe how various types of data were collected.
Reproduction
Data on reproductive status were gathered from females
killed by hunters. Hunters recorded signs of lactation
and litter size for each animal (Noss et al. 2003; Noss,
Cullar, and Cullar 2004) and collected and preserved
fetuses. Using the methods of Rojas-Surez and Maffei
(2003; see also Rojas-Surez 2002), fetuses were aged so
as to estimate likely birth dates and thus evaluate potential seasonality in breeding.
Data on the proportion of females that were pregnant and lactating each month were used to further
evaluate seasonality in reproduction, and to estimate
fecundity rates (Cullar 1999; Noss et al. 2003). Two
main seasons occur in the Chaco: wet and dry. The wet
season lasts from December to April. However, the dry
season can be subdivided into a cold-dry period lasting
from May to July, and a warm-dry period lasting from
August to November (Cullar 1999).
Diet
To assess diet, stomachs from hunter-killed armadillos
were collected and preserved in formalin before being
washed and placed in 70% alcohol. Stomach contents
were then dried and weighed (Bruno 1998; Valencia
1999; Bruno and Cullar 2000; Rojas-Surez 2002). The
contents of each stomach were separated according
to their different components: animal matter (insects,
other invertebrates, vertebrates); plant matter (fiber,
stems, leaves, fruits and seeds); and other (small stones,
bone, soil, fat). When possible, insects and other invertebrates were identified to the family or genus level, and
308
E. Cullar
plant components were identified to the species level

(Bruno 1998; Valencia 1999; Bruno and Cullar 2000).
Habitat
Isoseo hunters named 83 different habitats for those
places where they caught armadillos, basing their observations on characteristics such as forest height or
dominance of a particular plant. These names were
translated to four of the habitat categories described
by Navarro and Fuentes (1999) for the Isoso region
(Cullar 1999; Noss et al. 2003). Three categories corresponded to vegetation series of the Chaco alluvial plain:
Chaco forest on well-drained or sandy soils (W); Chaco
forest on medium-drained soils (M); and Chaco forest
on poorly drained soils (P). The fourth category was a
vegetation series of the Chaco riverine landscape: phreatophytic Chaco forest and seasonally flooded Chaco
forest (R). A final category (CO) included areas converted for human use, such as farms, communities, and
corrals (Cullar 1999; Noss et al. 2003).
Activity patterns
I assumed hunters caught armadillos that were active,
that is, outside their burrows. Thus, data on time of
capture, as recorded by Isoseo hunters, were used to
describe active periods for each species (Cullar 1999;
Cullar 2002; Noss et al. 2003). These data were supplemented with camera trap records from 17,000 trap
nights (Maffei et al. 2004; Noss, Pea, and Rumiz 2004;
Pea and Cullar 2005). Camera traps were distributed
in 4 permanent study sites (chosen because they varied
in the amount of annual rainfall) within Kaa-Iya Park
and the Isoso: Tucavaca (800 mm), Ravelo (650 mm),
Cerro Cortado (400 mm), and Guanaco (<400 mm).
Times of day were grouped into six four-hour periods:
0:0003:00, 04:0007:00, 08:0011:00, 12:0015:00,
16:0019:00, 20:0023:00 (each hour representing minutes 1 to 59).
Abundance
I used several methods to estimate abundance and density of armadillos, including linear transects (Ayala and
Noss 2000), track plots (Noss and Cullar 2000), census walks with dogs (Cullar 2002), and camera traps
(Noss, Pea, and Rumiz 2004; Pea and Cullar 2005).
Home range
Three different methods were used to assess minimum
home ranges: radiotelemetry and spool-and-thread
tracking for T. matacus (Perrogn 2001; Barrientos
and Cullar 2004) and camera trapping for P. maximus (Noss, Pea, and Rumiz 2004). The latter method
consisted of a systematic camera trap survey, with the
primary objective of estimating jaguar (Panthera onca)
population density (Maffei et al. 2004).
Results
Reproduction
Table 29.1 presents the reproductive status of 926 armadillo females killed by hunters. For each species, fewer
than half the animals killed were females. The combined
proportion of pregnant or lactating females approached
40% in D. novemcinctus, whereas proportions were less
than 20% for all other species.
D. novemcinctus showed a clear reproductive peak
from July to February, during the warm-dry season.
The pregnancy rate for D. novemcinctus was 55% in
October (Cullar 1999; Noss et al. 2003; Rojas-Surez
and Maffei 2003), and there was a single peak in births
lasting from October to December (Rojas-Surez and
Maffei 2003). There was little or no reproductive activity observed between March and June.
T. matacus also exhibited a seasonal reproductive
pattern. However, pregnancy rates exhibited two peaks:
one from July to September and the other from December to February (Cullar 1999; Noss et al. 2003). RojasSurez (2002) confirmed this pattern with data from a
small sample (n = 14) of fetuses, and estimated that the
majority of births occurred in September.
Cha. villosus showed increases in reproductive activity during the warm-dry and wet seasons, and very little
activity in the cold-dry season. However, this variation
across seasons was not statistically significant (Cullar
1999; Noss et al. 2003).
E. sexcinctus and Cha. vellerosus exhibited a short,
concentrated reproductive season from October to December and November to January, respectively. Hunters reported lactating E. sexcinctus in March and July.
Maximum fecundity rates for the two species were 27%
for E. sexcinctus in December and 50% for Cha. vellerosus in January (Noss et al. 2003).
Diet
General analysis of stomach contents gave the following results: D. novemcinctus and T. matacus consumed
mostly insects; plant material made up slightly more
than half, and insects slightly less than half of the diets
of Cha. vellerosus and Cha. villosus; and plant mate-
309
Table 29.1. Reproductive Status of Female Armadillos Killed by Hunters in the

Isoso Area, 19972000

Species
Total Number
killed of females % Females % Pregnanta % Lactatinga
Cha. vellerosus
Cha. villosus
D. novemcinctus
E. sexcinctus
T. matacus
130
406
1,060
236
686
49
121
452
82
222
37.7
29.8
42.6
34.8
32.4
12.2
1.7
22.4
3.7
6.8
6.1
5.0
18.6
8.5
7.7
a. Percentages of total females of each species.
rial comprised the majority of the diet for E. sexcinctus

(Bruno 1998; Valencia 1999; Bruno and Cullar 2000;
see table 29.2). Further details on the specific insect
families consumed by each species can be found in
Bruno (1998).
Habitat
Armadillo species were captured in some habitats
much more often than in others (table 29.3). Relative
to their abundance in the study area, all species were
found disproportionately more often in habitat W, and
less often than expected in habitat M (table 29.3). D.
novemcinctus, E. sexcinctus, and Cha. vellerosus are

frequently hunted near human settlements (Noss et al.
2003), which might bias estimates of habitat use. However, hunters did not record information that would allow calculation of effort or encounter rates in particular
habitat types (Noss et al. 2003).
Activity patterns
As stated earlier, a basic assumption of this study was
that armadillos were hunted while active above ground.
Most hunters used dogs to find armadillos by detecting
their fresh scent (Cullar 2002). Hunters found arma-
Table 29.2. Principal Components of Armadillo Diets as Percentages of Total Stomach Content
Weight

Components
Insects
Other invertebrates
Vertebrates
Plant remains
Unidentified material
Total
Cha.
vellerosus
(n = 15)
Cha.
villosus
(n = 33)
D. novemcinctus
(n = 43)
41.0
2.4
0.2
56.0
0.4
100
35.5
2.2
3.4
58.9

100
84.9
5.0
1.1
8.5
0.5
100
E. sexcinctus
(n = 14)
15.8
1.2
12.9
68.3
1.8
100
T. matacus
(n = 37)
86.1
12.3
1.6
100
Source: Bruno and Cullar 2000.

Note: Sample sizes are numbers of stomachs examined.
Table 29.3. Percentage of Animals Collected by Hunters in Five Habitats

Species (sample size)
Cha. vellerosus (150)
Cha. villosus (420)
D. novemcinctus (1,182)
E. sexcinctus (250)
T. matacus (784)
P (5%)
9
2
1
4
3
Habitat (% of study area)

M (70%)
R (5%)
W (16%)
35
46
59
50
59
19
5
12
8
5
29
45
22
28
32
CO (4%)
9
2
6
9
2
Source: Noss et al. 2003.

Notes: Habitat categories are based on Navarro and Fuentes (1999). Abbreviations: P = low xeric Chaco forest on
poorly drained soils, M = low xeric Chaco forest on medium-drained soils, R = Chaco riverine landscape, W =
Chaco forest on well-drained sandy soils, CO = converted to human use.
310
E. Cullar
dillos at all times of day, but their effort was not equally
distributed across the entire day, nor across seasons.
For example, hunters changed from nocturnal searching to more diurnal activity during the cold-dry season,
and therefore their reports reflected a high percentage
of diurnally active armadillos. Their results contrast
with those obtained from camera traps, which showed
slightly different activity patterns, for example, a higher
proportion of nocturnal records for T. matacus and
Cha. vellerosus (figure 29.1). However, this difference
may be related to the fact that camera traps were placed
primarily during the warm-dry and rainy seasons and
very few during the cold-dry season.
Despite these differences, records from both hunters and camera traps showed that armadillo activity
periods differed between species (figure 29.1). Both
methods documented a nocturnal activity peak for D.
novemcinctus and marked diurnal activity during the

middle of the day for E. sexcinctus. However, there were
disparities. For example, hunters reports for T. matacus and Cha. vellerosus indicated two activity peaks
(diurnal and nocturnal), whereas camera trap records
indicated only a nocturnal peak. In other cases, data
were not available to compare results between methods.
Cha. villosus showed diurnal activity in hunters data
but it was never recorded by camera traps. Similarly,
data from camera traps suggested a highly nocturnal
activity pattern in P. maximus, but there were no observations by hunters to corroborate this.
Abundance
Under the assumption that hunters do not select for
particular species, implying that capture rates should
reflect the true relative abundance of different arma-
Figure 29.1. Activity patterns for Dasypus novemcinctus, Euphractus sexcinctus, Tolypeutes matacus, and Chaetophractus vellerosus.
311
Figure 29.2. Armadillos

captured by camera traps in
different areas of the Bolivian Chaco.
dillo species, D. novemcinctus was most abundant (n =

1352), followed by T. matacus (n = 978), Cha. villosus
(n = 519), E. sexcinctus (n = 281), and Cha. vellerosus
(n = 176, Noss et al. 2004; note that these values differ
from those in table 29.1 because the values here include
animals where more detailed information, e.g., sex and
reproductive condition, was not recorded).
To confirm estimates derived from animals killed by
hunters, several attempts were made to estimate population densities of live animals. Initial attempts were
based on traditional line transect methods. However,
such methods are not ideal in zones of thick and thorny
vegetation because of difficulty of access without causing disturbance. In addition, poor visibility and poor
detectability of small, nocturnal, and fossorial species
may make transects unreliable. One solution is to adapt
the traditional hunting method of using dogs to survey for armadillos in nonhunted areas; however, this
method has to be carefully planned with regard to
health issues associated with taking domestic dogs into
protected areas.
Density analysis from 6,560 km of line transects
sampled in the Isoso area (Ayala and Noss 2000)
showed the highest value for E. sexcinctus (1.18 animals/
km2, 95% confidence interval [c.i.] = 0.821.68, n = 58)
followed by D. novemcinctus (0.65 animals/km2, 95%
c.i. = 0.391.09, n = 40) and Cha. villosus (0.58 animals/
km2, 95% c.i. = 0.380.90, n = 40). Other information,
including hunting records and track observations, sug-
gests these estimates are misleading, particularly for the

nocturnal D. novemcinctus (only daytime transects were
conducted). The data are presented here only because
no other information exists for the region. Using hunting dogs, Cullar (2002) estimated a T. matacus density
of 1.9 individuals/km2 for a nonhunted area. Barrientos
and Cullar (2004) used radiotelemetry to estimate a
density of 7 individuals/km2 for the same species in a
more open area with a different soil type (sandy instead
of poorly drained), only 7 km from the previous study
area. Finally, two separate surveys at Tucavaca, utilizing
camera trap capture and recapture of identified individuals, resulted in estimates of P. maximus population
density of 5.77 (SE = 1.95) and 6.28 2.76 individuals/100 km2 (Noss, Pea, and Rumiz 2004).
Not surprisingly, species abundances varied with
habitat types. Using camera traps, Pea and Cullar
(2005) found a relationship between capture frequencies of armadillos and rainfall in the Chaco (figure
29.2). Though present at all sites surveyed, T. matacus
appeared most abundant at the driest sites. In contrast,
P. maximus was absent from the drier sites and most
abundant where rainfall was highest.
A final set of data came from indirect indexes of
density derived from track plots. Track plots (Noss and
Cullar 2000) were established in the study area to record the incidence of medium and large-sized mammals; however, because armadillo tracks are very easy
to recognize, this method was used to make basic com-
312
E. Cullar
parisons between species. Using track plots, T. matacus

was the most abundant species in both hunted (45.8
observations/1,000 track plots) and nonhunted areas
(27.6 observations/1,000 track plots), followed by D.
novemcinctus (31.7 and 2.0 respectively), E. sexcinctus
(21.9 and 0.6), Cha. villosus (4.5 and 0), and Cha. vellerosus (1.7 and 0.2).
Home range
Data are available for two species. Studies using radiotelemetry (n = 27 individuals) and spool-and-thread (n
= 7) showed the home ranges of individual T. matacus
varied from 2.7 to 14 ha (Barrientos and Cullar 2004).
Multiple camera-trap records for two male P. maximus
provided home-range estimates, based on the minimum convex polygon, of 3 km2 for one animal and 15
km2 for the other (Noss, Pea, and Rumiz 2004).
Conclusions
The hunter monitoring program in the Bolivian Chaco
has proven to be an important source of information
(through written records and specimens) on the biology and ecology of armadillos. However, data collection in this program was not designed specifically for
such objectives. Nonetheless, the enormous quantity of
animals hunted as well as involvement of the community are important factors that enabled the collection
of sound information on some aspects of armadillo biology. Given that the Isoso hunting area extends over
4,000 km2 (see figure 23.2 of Noss et al. this volume),
the use of information contributed by hunters allowed
the compilation of a much more comprehensive and
substantial dataset, representative of the entire study
area, than could ever have been generated solely by researchers, who are often bound by considerable financial and time constraints.
Other methods of data collection all have advantages and disadvantages. Camera traps are expensive
and cannot be used indiscriminately. However, camera
traps showed promise in estimating population densities and home ranges for species in which individuals
are recognizable, for example, P. maximus. For all species, camera traps can provide useful data on habitat
usage and activity patterns. Radiotelemetry and spooland-thread studies provided information on home
ranges and movement patterns, but the former method
requires a better way of fixing radios on armadillos in
order to maximize data collection. For example, transmitters glued to the carapace of T. matacus sloughed off
within three months (Barrientos and Cullar 2004).
After almost a decade of studies yielding baseline
results, the next step is to develop more appropriate
research methods, as suggested by others in this book
(e.g., Abba and Cassini, Aguiar and Fonseca, Merrit,
Noss et al.). Such methods will allow us to gather more
accurate data and thus conduct more detailed studies of
behavioral ecology, taxonomy, genetics, and population
density. At this time, additional creativity and dedication are required to advance the study of armadillos.
Acknowledgments
I especially thank the hunters and their families who
took part in the hunter monitoring program, as well as
the parabiologists of the Isoso, for their participation
and support. Thanks to Andy Noss, Sergio Vizcano, Jim
Loughry, Alan Hesse, and two anonymous reviewers
who made important comments on this chapter. I also
thank the Capitana de Alto y Bajo Isoso, the Kaa-Iya
del Gran Chaco National Park, the Sistema Nacional de
reas Protegidas, and the Direccin General de Biodiversidad for institutional support and permissions. This
publication was made possible in part by financial support from the United States Agency for International
Development (USAID, Cooperative Agreement No.
511-A-00-01-00005). The opinions expressed are those
of the author and do not necessarily reflect the views
of USAID.
30
The natural history of the pichi, Zaedyus pichiy,
in western Argentina
Mariella Superina
cionalmente mais esbelta que exemplares do norte da

Provncia. Os piches so os nicos xenartros dos que se
sabe que podem entrar em torpor. Sua temperatura corporal altamente varivel, com um promedio de 35,1C
em exemplares que no esto em torpor e pode baixar
a 10C durante o torpor. Em Mendoza se alimentam
preferentemente de insetos, larvas, lagartas e material
vegetal, mas tambm ingerem carnia e pequenos vertebrados. Estes animais de hbitos predominantemente
diurnos e solitrios se reproduzem na primavera e
vero. Tem-se observado que os machos defendem seu
territrio agressivamente contra outros machos. A forte
presso cinegtica para usar esta espcie como fonte de
protena tem levado a uma reduo de avistamentos de
piches silvestres e sugere que o impacto humano esta
rapidamente reduzindo as populaes silvestres.
Resumen
Los piches Zaedyus pichiy son pequeos armadillos que
viven en zonas ridas con suelos arenosos y firmes. Se
estudiaron piches en estado silvestre y en cautiverio en
la Provincia de Mendoza, Argentina, para obtener datos
sobre distintos aspectos de su historia natural. Las hembras tienden a tener un peso corporal ms bajo y un
caparazn ms angosto que los machos, mientras que
los piches del sur tienen una cabeza proporcionalmente
ms esbelta que ejemplares del norte de la Provincia.
Los piches son los nicos xenartros de los que se sabe
que pueden entrar en sopor. Su temperatura corporal
es altamente variable, con un promedio de 35,1C en
ejemplares que no estn en sopor y puede bajar a 10C
durante el sopor. En Mendoza se alimentan preferentemente de insectos, larvas, gusanos y material vegetal,
pero tambin ingieren carroa y pequeos vertebrados.
Estos animales de hbitos predominantemente diurnos
y solitarios se reproducen en primavera y verano. Se
ha observado que los machos defienden su territorio
agresivamente contra otros machos. La fuerte presin
cinegtica para usar esta especie como fuente de protena ha llevado a una reduccin de avistamientos de
piches silvestres y sugiere que el impacto humano est
reduciendo rpidamente las poblaciones silvestres.
Introduction
Locally known as piche, quirquincho, or blanquito,
Zaedyus pichiy occurs farther south than any other armadillo species. Its range extends from the provinces
of Mendoza, San Luis, and Buenos Aires, south to the
Santa Cruz river in Argentina, and south to the Strait
of Magellan in Chile, at altitudes up to 2,500 m (Meritt
and Benirschke 1973; Wetzel 1985b; Aguiar and Fonseca
this volume). Z. pichiy rarely has been kept and studied
in captivity and seldom studied in the wild (Superina
2000), which explains the scarcity of scientific information about its life history; most knowledge of this species is based on anecdotal reports from locals.
Resumo
Os piches Zaedyus pichiy so pequenos tatus que vivem
em zonas ridas com solos arenosos e firmes. Estudaram-se piches Zaedyus pichiy em estado silvestre e em
cativeiro na Provncia de Mendoza, Argentina, para
obter dados sobre distintos aspectos de sua histria
natural. As fmeas tendem a ter um peso corporal mais
baixo e uma carapaa mais estreita que os machos, enquanto que os piches do sul tm uma cabea propor-
Methods
The present report is based on published literature, field
studies, and observations of captive pichis carried out
between 2001 and 2005. To obtain samples from wild
313
314
M. Superina
pichis, vehicle, walking, and horseback transects of

varying length were run in randomly chosen areas in
the entire provincial territory of Mendoza, Argentina
(3200'S to 3730'S and 6630'W to 7000'W; figure
30.1), except in the high Andes in the west, where pichis
are absent. Transects were sampled in all seasons and
at all times of day. Pichis were run down, captured by
hand, and restrained manually to take morphometric
measurements, measure rectal temperature with a thermocouple, and collect blood and fecal samples. Rectal
temperature was measured no later than 5 minutes after
capture and blood samples extracted within 15 minutes.
All pichis were released at their capture site within 45
minutes.
Observations of captive pichis were obtained from
five males and six females kept individually in pens
made of wire mesh (2 1.5 2.5 m each); soil to a depth
of 2 m provided a natural substrate for digging. The
complex of enclosures was located in Lujn de Cuyo,
Mendoza province, within the pichis native range, to
ensure their exposure to natural light cycles and variations in ambient temperature. Males and females were
brought together during the mating season.
A temperature data logger (Thermochron iButton,
Maxim Dallas Semiconductors, Sunnyvale, CA) was
implanted into the subcutaneous tissue of a captive
adult male to record its body temperature at intervals
of 90 minutes during the estimated torpor season. During the experiment, the animal was kept in a metal box
measuring 80 60 50 cm filled with wood shavings,
and had free access to food and fresh water.
Independent-samples t-tests were used to test for
significant differences between morphometric data of
male and female pichis, and of northern and southern
populations, respectively. Quantitative data are reported as means SD.
Results
Morphology
Pichis are small, hairy armadillos that can be distinguished from other species by their sharply pointed
marginal scutes, short ears, and a light longitudinal
dorsal line extending from the posterior end of the
scapular shield (or the first band) to the posterior end
of the pelvic shield. Their scapular and pelvic shields
are separated by seven bands (range 69), and one
complete movable band is present on the anterior edge
of the scapular shield (Wetzel 1985a; Superina unpublished data). Carapace color varies from light yellow to
Figure 30.1. Sampling areas in Mendoza province, Argentina.
almost black. The tan ventral hair increases in density

and length during winter. In contrast to Euphractus
sexcinctus, Chaetophractus villosus and Chaetophractus
vellerosus, pichis lack pelvic glands (Wetzel 1985b). Also
unlike other Euphractinae, pichis lack teeth on the premaxillary and have eight maxillary and nine mandibular pairs of teeth (Wetzel 1985b).
Morphometric data for adult, wild-caught specimens
from Mendoza province are shown in table 30.1. There
is some evidence of sexual dimorphism. Female pichis
weigh less than males (mean for 28 females = 906 127
g, mean for 53 males = 983 177 g, p < 0.05), and this
difference is reflected in a significantly smaller carapace
width for females (21.2 1.1 cm, n = 41) than males (22.0
1.1 cm, n = 71, p < 0.001). Geographic variation also
occurs. Individuals from southern Mendoza are smaller
and have a proportionally more slender head than their
conspecifics from northern Mendoza. The head length:
width ratio of northern individuals was 0.09 0.02 (n
= 14), but 0.08 0.01 (n = 92) in southern pichis. The
The natural history of the pichi, Zaedyus pichiy, in western Argentina
Table 30.1. Morphometric Measurements of Adult Wild-Caught

Pichis from All Localities in Mendoza Province, Argentina

Head length
Head width
Ear length
Carapace length
Carapace width
Tail length
Hind foot length
Mean (cm)
SD
6.9
5.3
1.5
18.8
21.7
10.6
4.8
0.3
0.3
0.2
1.4
1.2
0.9
0.4
117
117
80
118
115
116
81
Note: Head length was measured from the tip of the snout to the
upper border of the head shield; head width at the broadest part
of the head shield; carapace length from the anterior edge of the
scapular shield to the posterior edge of the pelvic shield; and
carapace width on the fourth band.
small sample of northern animals precludes statistical

testing at this time; nonetheless, the difference is visually distinctive (figure 30.2) and makes it quite easy to
tell whether a pichi is from the north or south.
Habitat
Pichis live in arid habitats with firm sandy soils, such
as the Patagonian steppe in southern Mendoza (figure
30.3) and the Chacoan monte desert in its northern
reaches (figure 30.4). In Patagonia, the arid climate
and poor volcanic soil, covered with sands of variable depth, together determine the scarce vegetation
315
(Candia et al. 1993), which is composed of grasslands

with isolated shrubs, areas with open bush communities, and basaltic steps (Candia and Dalmasso 1995). In
the Chacoan monte, the dry climate and sandy ground
favor the growth of characteristic monte vegetation,
which consists of small trees, shrubs, and scarce pasture
(Peralta de Galmarini and Martinez Carretero 1995). In
both habitats, pichis are more commonly found near
shrubs than on open lands.
Physiology
A major difference between armadillos and most other
eutherian mammals is their relatively low body temperature and a reduced ability to maintain it at a constant
level. The body temperature of wild pichis averages
35.1C, but may fluctuate considerably over the course
of a day as well as with changes in ambient temperature. Compared with other armadillo species, Z. pichiy
shows the largest daily variation in body temperature,
in the range of 2236C under laboratory conditions
(Roig 1971). I recorded body temperatures in 58 wild,
nontorpid individuals from Mendoza province and
found ranges of 32.238.3C.
Z. pichiy seems to be very susceptible to sudden
changes in environmental temperature. Laboratory
studies suggest that pichis display clear signs of inactivity and respiratory arrhythmia during hypothermia.
Figure 30.2. Head

morphology of
Zaedyus pichiy from
southern (left) and
northern (right)
Mendoza province,
Argentina.
Figure 30.3. Patagonian steppe in southern Mendoza province, Argentina (36.2S, 68.9W).
Figure 30.4. Chacoan monte in northern Mendoza province, Argentina (32.3S, 67.9W).
The natural history of the pichi, Zaedyus pichiy, in western Argentina
Their behavior returns to normal after prolonged cold

exposure. A sudden rise in environmental temperature
from 0C to 15C overexerts the thermoregulatory ability of cold-acclimated animals, leading to an increase in
rectal temperature of 16C within an hour and eventually to death (Roig 1971).
The pichi is the only known xenarthran that can
enter torpor (Roig 1971). Torpor usually lasts from
mid-May to mid-August. Examinations of carcasses of
wild pichis confiscated by law enforcement agencies in
Mendoza indicate the animals build a subcutaneous fat
layer (up to 1.5 cm thick) just prior to the onset of torpor. The rectal temperature of a wild, torpid pichi was
18.8C, whereas an adult male in captivity lowered its
body temperature to 10C during several torpor bouts.
According to several reports from locals, pichis do not
leave their burrows at all during winter, which suggests
they might hibernate to cope with the harsh climate. It
is still not clear, however, whether they undergo several
short torpor bouts interrupted by periods of activity, or
if their metabolism and body temperature are lowered
for longer periods. In the wild, only subadult individuals, who may lack the necessary energy reserves to enter
torpor, have been observed outside their burrows during winter; it remains unknown whether adult pichis
remain in their burrows during the entire cold period
or emerge during periodic arousals. Preliminary observations on captive pichis have provided ambiguous results: while some captive pichis remained in their burrows for up to seven days, other individuals, in similar
body condition, emerged every day. This occurred even
though all animals were maintained in similar conditions and provided with equal amounts of food.
Food habits
Pichis are considered carnivores-omnivores with a
preference for carrion. According to Redford (1985a),
their natural diet also includes insects, especially ants
and their larvae, other invertebrates, and plant material. Preliminary studies of stomach contents from confiscated carcasses suggest that, in Mendoza, pichis feed
predominantly on insects, grubs, and plant material
(especially Prosopis pods in northern Mendoza). Carrion and small vertebrates seem to constitute a minor
part of their diet.
317
Reproduction
Field studies indicate pichis begin their mating season
in mid to late September. Mating has been observed in
September and might extend through October. Male
armadillos are ascrotal (Anderson and Benirschke
1966); the testes of adult pichis cannot be seen or palpated in nonreproductive individuals, but are easily
recognized during the reproductive season. In seasonal
breeders, testis size increases during the reproductive
season because of structural modifications of the gonads and enhanced production of spermatozoa (Goeritz et al. 2003). Pichis with prominent testes have
been observed only between September and February,
which suggests sperm production is limited to spring
and summer. Pregnant females have been observed in
October and November, and lactating females between
November and February. Pichis with small carapaces
and a body weight below 600 g, suggesting they were
juveniles, have been recorded in January, February, and
March, but no other months. Similarly, most subadult
pichis, that is, individuals weighing between 600 and
840 g and showing no signs of reproductive activity,
have been caught between January and March.
Captive pichis reached sexual maturity at 910
months of age. Mounting was observed between September and December and births of one or two offspring
occurred after a gestation period of approximately 60
days. The offspring did not emerge until they were fully
weaned at 40 days of age (figure 30.5). These observations, as well as the field data described above, are consistent with those of Redford and Eisenberg (1992) and
suggest pichis are seasonal breeders that reproduce in
spring and have only one litter per year.
Behavior
Wild pichis seem to have predominantly diurnal and
solitary habits. No wild pichis were caught before 10:00
or after 20:30. Several individuals kept in captivity, however, were more active at night (Altmann and Scheel
1980; Superina unpublished data). Social interactions
between wild individuals were observed only during
the breeding season. During this same time, captive
adult males aggressively defended their territory against
other males. Additionally, fights between males and females occurred. Aside from aggression connected with
reproduction, fights over food resources have been observed between individuals sharing an enclosure.
318
M. Superina
Figure 30.5. Captive-born pichis, Zaedyus pichiy, aged approximately 40 days, at first emergence.
Several types of vocalizations have been recorded

in captive pichis. A purring sound is emitted when an
animal is threatened, which can change into a gruntlike
vocalization or a scream (similar to that of Cha. vellerosus) when it is picked up.
In the field, pichis either lie motionless on the
ground when threatened, run to hide in an existing
burrow, or run under a shrub where they start digging a
new shelter. Pichis tend to hide their burrow entrances
under vegetation. The roots not only provide support
and protection (Greegor 1985); pichis might also feed
on the roots or on associated insects. Similar to Cha.
vellerosus (Greegor 1985), pichis thrust sand backward
while digging and seal the burrow as they excavate. Burrows can be several meters long and of variable depth;
maximum depth of burrows excavated during winter
was 1.5 m. Only one of >70 excavated burrows of wild
pichis contained nesting material, which suggests that,
unlike Dasypus novemcinctus (Taber 1945; Clark 1951),
this species does not build nests inside its burrows. In
captivity, pichis may return to the same burrow for several days but still dig small, temporary burrows almost
every day.
Conservation status
Virtually nothing is known about the size or density
of pichi populations in any part of their current range.
Although hunting pichis is illegal in Mendoza province,

this species is currently one of the main prey of poachers, who use it as a protein source. Intense hunting
pressure seems to have depleted populations over the
years, as reflected in lower encounter rates and reduced
catch rates by poachers. The latter has been confirmed
during interviews with locals and hunters, who have
reported reduced sightings of pichis in the past years,
and from law enforcement agencies, who have noted a
considerable decrease in numbers of confiscated pichis.
Increasing human population size and expansion of
human activities, such as agriculture, into pichi habitat
will likely cause further losses (see Aguiar and Fonseca
this volume).
Acknowledgments
I thank the many people who helped with fieldwork,
especially the rangers at the Department of Natural
Renewable Resources of Mendoza province. Special
thanks are due to Guillermo Ferraris for his invaluable
help with fieldwork and care of captive animals, and
to my advisor Patrice Boily. Fieldwork was supported
by the University of New Orleans and grants from the
IUCN Edentate Conservation Action Fund and Pittsburgh Zoo.
Bibliography
antigen but dispensable for induction of specific T-cell responses. Journal of Immunology 171: 114855.
Aiello, A. 1985. Sloth hair: unanswered questions. In G. G.
Montgomery, ed., The evolution and ecology of armadillos,
sloths and vermilinguas, 21318. Smithsonian Institution
Press, Washington, D.C.
Alberdi, M.-T., J. L. Prado, and R. Salas. 2004. The Pleistocene
Gomphotheriidae (Proboscidea) from Peru. Neues Jahrbuch fr Geologie und PalontologieAbhandlungen und
Monatshefte 231: 42352.
Alexander, R. D. 1979. Darwinism and human affairs. University of Washington Press, Seattle.
Alexander, R. McN. 1983. Animal mechanics. Blackwell, Oxford.
. 1985. Mechanics of posture and gait of some large dinosaurs. Zoological Journal of the Linnean Society 83: 125.
Alexander, R. McN., R. A. Faria, and S. F. Vizcano. 1999. Tail
blow energy and carapace fractures in a large glyptodont
(Mammalia, Edentata). Zoological Journal of the Linnean
Society 125: 4149.
Altman, P. L., and D. S. Dittmer. 1962. Growth, including reproduction and development. In Federation of American Societies for Experimental Biology Handbook, 163.
Altmann, D. and H. Scheel. 1980. Zur Aktivitt beim Zwerggrteltier, Euphractus pichiy, im Tierpark Berlin. Zoologischer
Garten N.F., Jena 50: 34552.
Altschul, S. F., et al. 1990. Basic local alignment search tool.
Journal of Molecular Biology 215: 40310.
Alvard, M. S., et al. 1997. The sustainability of subsistence hunting in the Neotropics. Conservation Biology 11: 97782.
Alvarez, S. 2004. Densidad y preferencia de hbitat del perezoso de dos dedos, Choloepus hoffmanni, en un Bosque
andino. Tesis de grado, Universidad Nacional de Colombia,
Bogot.
Ameghino, C. 1919. Sobre mamferos fsiles del piso araucanense de Catamarca y Tucumn. Primera Reunin Nacional
de la Sociedad Argentina de Ciencias Naturales 1: 15053.
Ameghino, F. 1884. Filogenia: principios de clasificacin transformista basados sobre leyes naturales y proporciones
matemticas. Coni, Buenos Aires.
. 1889. Contribucin al conocimiento de los mamferos
fsiles de la Repblica Argentina. Actas de la Academia Nacional de Ciencias en Crdoba 6: 11027. Atlas: plates 198.
. 1891. Sobre algunos restos de mamferos fsiles recogidos por el seor Manuel B. Zavaleta en la formacin miocena de Tucumn y Catamarca. Revista Argentina de Historia
Natural 1: 88101.
. 1893. Nouvelles dcouvertes palontologiques das la
Patagonie australe. Revue Scientifique 51: 731.
. 1894. Enumration synoptique des espces de mam-
Abba, A. M., D. E. Udrizar Sauthier, and S. F. Vizcano. 2005.

Distribution and use of burrows and tunnels of Chaetophractus villosus (Mammalia, Xenarthra) in the eastern Argentine pampas. Acta Theriologica 50: 11524.
Abba, A. M., S. F. Vizcano, and M. H. Cassini. 2007. Effects of
land use on the distribution of three species of armadillos in
the Argentinean Pampas. Journal of Mammalogy 88: 5027.
Abba, A. M., et al. 2004. Determinantes de la distribucin de
armadillos (Xenarthra, Dasypodidae) en La Plata y Magdalena, Buenos Aires, Argentina. Abstracts XIX Jornadas
Argentinas de Mastozoologa 19: 36.
Acha, P. N., and B. Szyfres, eds. 2003. Zoonoses and communicable diseases common to man and animals. 3rd ed. Pan
American Health Organization, Washington, D.C.
Adamoli, V. C., et al. 2001. Comparative morphologic placental
types in Dasypodidae (Chaetophractus villosus, Cabassous
chacoensis, Tolypeutes matacus and Dasypus hybridus). Biocell 25: 1722.
Affanni, J. M., et al. 1987. Absence of apnea in armadillos covered with soil. Respiration Physiology 67: 23945.
Afzelius, B. A. 1979. Sperm structure in relation to phylogeny
in lower Metazoa. In D. W. Fawcett and J. M. Bedford, eds.,
The spermatozoon, 24351. Urban and Schwarzenberg, Baltimore and Munich.
Agar, N. S. 1996. Erythrocyte metabolism in the Australian
marsupials. Experimental Hematolology 24: 725.
Aggio, M. C. 1987. Filogenia, embriologa y patologa de la
mielopoyesis esplnica. Acta Bioqumica Clnica Latinoamericana 21: 24954.
Aguiar, F. B., et al. 1988a. Estudos dos volumes e capacidades
pulmonares em preguias (B. variegatus) anestesiadas. Resumos III Reunio Anual da Federao das Sociedades de
Biologia Experimental 2.54: 81.
Aguiar, F. B., et al. 1988b. Propriedades mecanicas ativas do
sistema respiratorio de preguias (B. variegatus) anestesiadas. Resumos III Reunio Anual da Federao das Sociedades
de Biologia Experimental 2.53: 81.
Aguiar, J. M. 2004. Species summaries and species discussions.
Edentata 6: 326.
Aguilar, R. F., F. Dunker, and M. Garner. 2002. Dilated cardiomyopathy in two giant anteaters (Myrmecophaga tridactyla).
Abstract, Proceedings of the American Association of Zoo
Veterinarians, 510 October: 16972.
Aguirre, L. F., et al. 1999. El monitoreo de aprovechamiento
de fauna en la EBB: un anlisis crtico. In T. G. Fang, O. L.
Montenegro, and R. E. Bodmer, eds., Manejo y conservacin
de fauna silvestre en Amrica Latina, 97108. Instituto de
Ecologa, La Paz, Bolivia.
Aichele, P., et al. 2003. Macrophages of the splenic marginal
zone are essential for trapping of blood-borne particulate
319
320
Bibliography
mifres fossiles des formations ocnes de Patagonie. Boletn de la Academia Nacional de Ciencias en Crdoba 13:
259455.
. 1895. Sur les dents fossiles de lArgentine: examen critique, revision et correction de louvrage de R. Lydekker: The
extint edentates of Argentina. Revista del Jardn Zoolgico de
Buenos Aires 3: 97192.
. 1902. Notices prliminaires sur des mammifres nouveaux des terrains crtacs de Patagonie. Boletn de la Academia Nacional de Ciencias de Crdoba 17: 168.
. 1904. Nuevas especies de Mamferos cretceos y terciarios de la Repblica Argentina. Anales de la Sociedad
Cientfica Argentina 57: 16275, 32741.
. 1905. La faceta articular inferior nica del astrgalo de
algunos mamferos no es un carcter primitivo. Anales del
Museo Nacional de Buenos Aires 12: 164.
Amezcua, M. E., et al. 1984. Wild Mexican armadillo with leprosy-like infection [letter]. International Journal of Leprosy
52: 25455.
Amrine-Madsen, H., et al. 2003. A new phylogenetic marker,
apolipoprotein B, provides compelling evidence for eutherian relationships. Molecular Phylogenetics and Evolution 28:
22540.
Anacleto, T. C. S., and J. Marinho-Filho. 2001. Habito alimentar
do tatu-canastra (Xenarthra, Dasypodidae) em uma area de
cerrado do Brasil Central. Revista Brasileira de Zoologia 18:
68188.
Anderson, J. M., and K. Benirschke. 1963. Fetal circulations in
the placenta of Dasypus novemcinctus Linn. and their significance in tissue transplantation. Transplantation 1: 30610.
. 1966. The armadillo, Dasypus novemcinctus, in experimental biology. Laboratory Animal Care 16: 20216.
Anderson, M. 1978. Leprosy in an armadillo from Texas. Leprosy Scientific Memorandum L-967: 5.
Anderson, M. J., and A. F. Dixson. 2002. Sperm competition:
motility and the midpiece in primates. Nature 416: 496.
Anderson, R. P., and C. O. Handley, Jr. 2001. A new species
of three-toed sloth (Mammalia: Xenarthra) from Panama,
with a review of the genus Bradypus. Proceedings of the Biological Society of Washington 114: 133.
. 2002. Dwarfism in insular sloths: biogeography, selection, and evolutionary rate. Evolution 56: 104558.
Andrade da Costa, B. L. S. 1987. Organizao retiniana das clulas ganglionares nas preguias Bradypus variegatus Schinz
e Choloepus didactylus L. Masters thesis, Universidade Federal de Pernambuco, Pernambuco, Brazil.
Andrade da Costa, B. L. S., et al. 1987. Unusual distribution of
ganglion cells in the retina of the three-toed sloth (Bradypus
variegatus). Brazilian Journal of Medical and Biological Research 20: 74148.
Andrade da Costa, B. L. S., et al. 1989. Ganglion cell size and
distribution in the retina of the two-toed sloth (Choloepus
didactylus L.). Brazilian Journal of Medical and Biological
Research 22: 23336.
Anthony, H. B. 1926. Mammals of Porto Rico, living and extinctRodentia and Edentata. New York Academy of Sciences, Scientific Survey of Porto Rico and the Virgin Islands
9: 97241.
Aquino, R. M., R. E. Bodmer, and J. G. Gil. 2001. Mamferos de
la cuenca del ro Samiria: ecologa poblacional y sustentabili-
dad de la caza. Universidad Nacional Mayor de San Marcos,

Lima, Peru.
Aramayo, S. A., and T. Manera de Bianco. 1996. Edad y nuevos
hallazgos de icnitas de mamferos y aves en el yacimiento
paleo iconolgico de Pehuen-Co (Pleistoceno Tardo), Provincia de Buenos Aires, Argentina. Asociacin Paleontolgica Argentina, Publicacin Especial 4: 4757.
Arenas, P. 2003. Etnografa y alimentacin entre los Toba-achilamolek y Wich-Lhukutas del Chaco Central (Argentina).
Pastor Arenas, Buenos Aires.
Arias, J. R., et al. 1982. Isolation of Histoplasma capsulatum
from an armadillo (Dasypus novemcinctus) in the eastern
Amazon of Brazil. Transactions of the Royal Society of Tropical Medicine and Hygiene 76: 7056.
Arispe, R., and D. Rumiz. 2002. Una estimacin del uso de los
recursos silvestres en la zona del Bosque Chiquitano, Cerrado y Pantanal de Santa Cruz. Revista Boliviana de Ecologa
y Conservacin Ambiental 11: 1736.
Aristizabal, B. H., et al. 2002. Experimental Paracoccidioides
brasiliensis infection in mice: influence of the hormonal status of the host on tissue responses. Medical Mycology 40:
16978.
Asher, R. J., M. J. Novacek, and J. H. Geisler. 2003. Relationships
of endemic African mammals and their fossil relatives based
on morphological and molecular evidence. Journal of Mammalian Evolution 10: 13194.
Austin, C. R., and M. W. H. Bishop. 1958. Some features of the
acrosome and perforatorium in mammalian spermatozoa.
Proceedings of the Royal Society of London, Series B 149:
23440.
Avise, J. C. 2000. Phylogeography. Harvard University Press,
Cambridge.
. 2004. Molecular markers, natural history, and evolution.
2nd ed. Sinauer Associates, Sunderland, Mass.
Ayala, C. J. M. 1997. Utilizacin de la fauna silvestre del grupo
tnico Ayorode en la comunidad Tobit, Santa Cruz, Bolivia. Undergraduate thesis, Universidad Autonoma Gabriel
Ren Moreno, Santa Cruz, Bolivia.
Ayala, J., and A. Noss. 2000. Censo por transectas en el Chaco:
limitaciones biolgicas y sociales de la metodologa. In E.
Cabrera, C. Mercolli, and R. Resquin, eds., Manejo de fauna silvestre en Amazona y Latinoamrica, 2936. CITES
Paraguay, Fundacin Moiss Bertoni, University of Florida,
Asuncin, Paraguay.
Azara, F. 1998. Viajes por la Amrica Meridional. Vol. I. Ediciones el Elefante Blanco, Buenos Aires. (reprint of work originally published in 1781)
Baccetti, B. 1979. The evolution of the acrosomal complex. In D.
W. Fawcett and J. M. Bedford, eds., The spermatozoon, 305
29. Urban and Schwarzenberg, Baltimore and Munich.
. 1985. Evolution of the sperm cell. In C. B. Metz and A.
Monroy, eds., Biology of fertilization, 358. Academic Press,
Orlando.
. 1986. Evolutionary trends in sperm structure. Comparative Biochemistry and Physiology 85: 2936.
. 1987. News on phylogenetical and taxonomical spermatology. In H. Mohr, ed., New horizons in sperm cell research,
33348. Japan Scientific Societies Press, Tokyo.
Bagagli, E., et al. 1998. Isolation of Paracoccidioides brasiliensis
from armadillos (Dasypus novemcinctus) captured in an en-
Bibliography
demic area of paracoccidioidomycosis. American Journal of

Tropical Medicine and Hygiene 58: 50512.
Bagagli, E., et al. 2003. High frequency of Paracoccidioides
brasiliensis infection in armadillos (Dasypus novemcinctus):
an ecological study. Medical Mycology 41: 21723.
Bagagli, E., et al. 2006. Phylogenetic and evolutionary aspects of
Paracoccidioides brasiliensis reveal a long coexistence with
animal hosts that explain several biological features of the
pathogen. Infection, Genetics and Evolution 6: 34451.
Bagatto, B., D. A. Crossley II, and W. W. Burggren. 2000. Physiological variability in neonatal armadillo quadruplets: within- and between-litter differences. Journal of Experimental
Biology 203: 173340.
Baker, R. H. 1943. May food habits of armadillos in eastern
Texas. American Midland Naturalist 29: 37980.
Balia, L. M., et al. 1985. Experimental reproduction of leprosy in Dasypus hybridus. International Journal of Leprosy 53:
59599.
Ballard, W. B., et al. 1998. Comparison of VHF and satellite telemetry for estimating sizes of wolf territories in northwest
Alaska. Wildlife Society Bulletin 26: 82329.
Barny, M., et al. 1967. Relation of properties of isolated myosin to those of intact muscles of the cat and sloth. European
Journal of Biochemistry 2: 15664.
Barbarn, F. R. 2000. Recursos alimenticios derivados de la
caza, pesca y recoleccin de los Wichi del Ro Pilcomayo
(Provincia de Salta, Argentina). In E. Cabrera, C. Mercolli, and R. Resquin, eds., Manejo de fauna silvestre en
Amazona y Latinoamrica, 50727. CITES Paraguay, Fundacin Moises Bertoni, University of Florida, Asuncin,
Paraguay.
Bargo, M. S. 2001a. El aparato masticatorio de los perezosos terrestres (Xenarthra, Tardigrada) del Pleistoceno de la Argentina. Morfometra y biomecnica. Ph.D. diss., Universidad
Nacional de La Plata, La Plata, Argentina.
. 2001b. The ground sloth Megatherium americanum:
skull shape, bite forces, and diet. Acta Paleontologica Polonica 46: 17392.
. 2003. Biomechanics and palaeobiology of the Xenarthra: state of the art. Senckenbergiana biologica 83: 4150.
. 2004. Cenozoic giants of South America. Journal of
Morphology 260: 276.
Bargo, M. S., G. De Iuliis, and S. F. Vizcano. 2006. Hypsodonty
in Pleistocene ground sloths (Xenarthra, Tardigrada). Acta
Paleontologica Polonica 51: 5361.
Bargo, M. S., N. Toledo, and S. F. Vizcano. 2006. Muzzle reconstruction of the Pleistocene ground sloths. Journal of
Morphology 267: 24863.
Bargo, M. S., et al. 2000. Limb bone proportions, strength and
digging in some Lujanian (Late Pleistocene-Early Holocene)
mylodontid ground sloths (Mammalia, Xenarthra). Journal
of Vertebrate Paleontology 20: 60110.
Barreto, M., P. Barreto, and A. DAlessandro. 1985. Colombian
armadillos: stomach contents and infection with Trypanosoma cruzi. Journal of Mammalogy 66: 18893.
Barrientos, J. 2004. Monitoreo de corechis (Tolypeutes matacus): el rea de accin mediante la telemetra en Cerro Colorado del Chaco boliviano. In R. L. Cullar, ed., II Congreso
de investigadores en recursos naturales del Parque Nacional
Kaa-Iya y zonas de influencia: Memoria (San Jos de Chiq-
321
uitos, 78 noviembre de 2003), 1618. WCS, CABI, Santa

Cruz, Bolivia.
Barrientos, J., and R. L. Cullar. 2004. rea de accin de Tolypeutes matacus por telemetra y seguimiento por hilos en
Cerro Cortado del Parque Kaa-Iya. In Memorias: Manejo
de fauna silvestre en la Amazona y Latinoamrica. VI Congreso, Iquitos, Per, 510 de septiembre de 2004, 11115.
www.revistafauna.com.pe/memo.htm.
Barros, M. C., I. Sampaio, and H. Schneider. 2003. Phylogenetic
analysis of 16S mitochondrial DNA data in sloths and anteaters. Genetics and Molecular Biology 26: 512.
Barroso, C. M. L., and H. Seunez. 1991. Chromosome studies
on Dasypus, Euphractus and Cabassous genera (Edentata:
Dasypodidae). Cytobios 68: 17996.
Bartmann, C. P., C. Beyer, and H. Widorf. 1991. Topographie
der Beckenhhlenorgane sowie Befunde zur Makroskopie
und Histologie der Geschlechtsorgane eines mnnlichen
Groen Ameisenbren (Myrmecophaga tridactyla) im Hinblick auf seine Fertilitt. Berliner und Mnchener Tierrztliche Wochenschrift 104: 4146.
Bartmann, W. 1983. Haltung und Zucht von grossen Ameisenbren, Myrmecophaga tridactyla, im Dortmunder Tierpark.
Zoologischer Garten N.F. 53: 131.
. 1985. Geburt und Handaufzucht eines Groen Ameisenbren (Myrmecophaga tridactyla) im Tierpark Dortmund. Zeitschrift des Klner Zoo 28: 5160.
. 1994. Breeding anteaters in captivity. In Report of the
Scientific Session, 49th Annual Conference IUDZG/The
World Zoo Organization, 16. So Paulo, Brazil.
Baur, E. H. 1998. Study of subsistence hunting in the forestry
concession of Carmelita, San Andrs, Petn. Final Report
19971998. ProPetn, Conservation International, Flres,
Petn, Guatemala.
Beath, M. M., K. Benirschke, and L. E. Brownhill. 1962. The
chromosomes of the nine-banded armadillo, Dasypus
novemcinctus. Chromosoma 13: 2738.
Becher, H. 1921. Zur Kenntnis der Placenta von Bradypus tridactylus. Zeitschrift fr Anatomie und Entwicklungsgeschichte 61: 11436.
. 1931. Placenta und uterusschleimhaut von Tamandua
tetradactyla (Myrmecophaga). Gegenbauers Morphologisches
Jahrbuch 67: 381458.
Bedford, J. M. 1999. The implications of unusual sperm/female
relationships in mammals. In C. Gagnon, ed., The male gamete: from basic science to clinical applications, 8192. Cache
River Press, St. Louis.
Bedford, J. M., and D. D. Hoskins. 1990. The mammalian spermatozoon: morphology, biochemistry and physiology. In
G. E. Lamming, ed., Marshalls physiology of reproduction,
379568. Churchill Livingstone, London.
Beebe, W. 1926. The three-toed sloth Bradypus cuculliger cuculliger Wagler. Zoologica 7: 167.
Bell, G. 1978. The evolution of anisogamy. Journal of Theoretical
Biology 73: 24770.
. 1997. The basics of selection. Chapman and Hall, New
York.
Benirschke, K., ed. 1969. Comparative mammalian cytogenetics.
Springer-Verlag, New York.
. 2005. Comparative placentation. <http://medicine.
ucsd.edu/cpa>.
322
Bibliography
Benirschke, K., R. J. Low, and V. H. Ferm. 1969. Cytogenetic

studies of some armadillos. In K. Benirschke, ed., Comparative mammalian cytogenetics, 33045. Springer-Verlag, New
York.
Benirschke, K., and H. C. Powell. 1985. The placentation of
sloths. In G. G. Montgomery, ed., The evolution and ecology
of armadillos, sloths, and vermilinguas, 23741. Smithsonian
Institution Press, Washington, D.C.
Benirschke, K., and D. H. Wurster. 1969. The chromosomes
of the giant armadillo, Priodontes giganteus Geoffroy. Acta
Zoologica et Pathologica Antverpiensia 49: 12530.
Bennett, A. F. 2003. Experimental evolution and the Krogh Principle: generating biological novelty for functional and genetic
analyses. Physiological and Biochemical Zoology 76: 111.
Benson, R. H., R. E. Chapman, and A. F. Siegel. 1982. On the
measurement of morphology and its change. Paleobiology
8: 32839.
Bergqvist L. P., . A. L. Abrantes, and L. d. S. Avilla. 2004. The
Xenarthra (Mammalia) of So Jos de Itabora Basin (upper
Paleocene, Itaboraian), Rio de Janeiro, Brazil. Geodiversitas
26: 32337.
Berlin, B., and E. A. Berlin. 1983. Adaptation and ethnozoological classification: theoretical implications of animal resources and diet of the Aguaruna and Huambisa. In R. B. Hames
and W. T. Vickers, eds., Adaptive responses of native Amazonians, 30125. Academic Press, New York.
Bermdez, P., N. N. Polini, and E. B. Casanave. 2004. A study
of platelets in the armadillo Chaetophractus villosus (Xenarthra, Dasypodidae). Platelets 15: 27985.
Bertonatti, C., and G. Aprile. 1999. Pichiciego menor. Vida Silvestre 63: 2122.
Bertonatti, C., and J. Corcuera. 2000. Situacin ambiental Argentina 2000. Fundacin Vida Silvestre Argentina, Buenos
Aires.
Best, R. C., and A. Y. Harada. 1985. Food habits of the silky anteater (Cyclopes didactylus) in the Central Amazon. Journal
of Mammalogy 66: 78081.
Beyer, W. N., E. S. Connor, and S. Gerould. 1994. Estimates of
soil ingestion by wildlife. Journal of Wildlife Management
58: 37582.
Bezerra, A. M. R., F. H. G. Rodrigues, and A. P. Carmignotto.
2001. Predation of rodents by the yellow armadillo (Euphractus sexcinctus) in the cerrado of central Brazil. Mammalia 65: 8688.
Bialek, R., et al. 2000. Small subunit ribosomal DNA sequence
shows Paracoccidioides brasiliensis closely related to Blastomyces dermatitidis. Journal of Clinical Microbiology 38:
319093.
Binford, C. H., E. E. Storrs, and G. P. Walsh. 1976. Disseminated
infection in the nine-banded armadillo (Dasypus novemcinctus) resulting from inoculation with M. leprae. Observations made on 15 animals studied at autopsy. International
Journal of Leprosy 44: 8083.
Binford, C. H., et al. 1977. Naturally acquired leprosy-like disease in the nine-banded armadillo Dasypus novemcinctus:
histopathologic and microbiologic studies of tissues. Journal
of the Reticuloendothelial Society 22: 37788.
Birkhead, T. R. 1995. Sperm competition: evolutionary causes
and consequences. Reproduction, Fertility and Development
7: 75575.
Blanco, R. E., and A. Czerwonogora. 2003. The gait of Megatherium Cuvier 1796 (Mammalia, Xenarthra, Megatheriidae).
Senckenbergiana biologica 83: 6168.
Blouin, M. S. 2003. DNA-based methods for pedigree reconstruction and kinship analysis in natural populations. Trends
in Ecology and Evolution 18: 50311.
Blue, J., and L. Weiss. 1981. Species variation in the structure
and function of the marginal zone. An electron microscope
study of cat spleen. American Journal of Anatomy 161: 169
87.
Bogado, M. M., et al. 1983. Aislamiento de Histoplasma capsulatum en Dasypus hybridus (tatu mulita de 7 bandas). Revista
Argentina de Micologa 6: 1417.
Boggs, D. F., P. B. Frappell, and D. L. Kilgore, Jr. 1998. Ventilatory, cardiovascular and metabolic responses to hypoxia
and hypercapnia in the armadillo. Respiration Physiology
113: 1019.
Boily, P. 2002. Individual variation in metabolic traits of wild
nine-banded armadillos (Dasypus novemcinctus) and the
aerobic capacity model for the evolution of endothermy.
Journal of Experimental Biology 205: 320714.
Boily, P., and F. M. Knight. 2004. Cold-induced fever and peak
metabolic rate in the nine-banded armadillo (Dasypus
novemcinctus). Physiological and Biochemical Zoology 77:
65157.
Boily, P., P. G. ONeil, and P. A. Prodhl. 2000. Testing the
predictions of the aerobic capacity model using wild, freeranging nine-banded armadillos (Dasypus novemcinctus)
as models. Comparative Biochemistry and Physiology B126:
S13.
Bolkovic, M. L. 1999. Usos de fauna silvestre de pobladores de
las cercanas de la Reserva Provincial Copo, Santiago del
Estero, Argentina. In T. G. Fang, O. L. Montenegro, and R.
E. Bodmer, eds., Manejo y conservacin de fauna silvestre
en Amrica Latina, 11724. Instituto de Ecologa, La Paz,
Bolivia.
Bolkovic, M. L., J. M. Affanni, and C. M. Ghersa. 1999. El peludo C. villosus y la mulita D. hybridus en agroecosistemas
de la pampa interior. Abstracts XIV Jornadas Argentinas de
Mastozoologa 14: 15.
Bolkovic, M. L., S. M. Caziani, and J. J. Potomastro. 1995. Food
habits of the three-banded armadillo (Xenarthra: Dasypodidae) in the dry Chaco, Argentina. Journal of Mammalogy
76: 1199204.
Bond, B. T., M. I. Nelson, and R. J. Warren. 2000. Home range
dynamics and den use of nine-banded armadillos on Cumberland Island, Georgia. Proceedings of the Annual Conference of Southeastern Fish and Wildlife Agencies 54: 41424.
Bond, B. T., R. J. Warren, and M. I. Nelson. 2002. Winter mortality of adult nine-banded armadillos (Dasypus novemcinctus) on Cumberland Island, Georgia. Georgia Journal of
Science 60: 20915.
Borrero, L., J. Lanata, and F. Borrela. 1988. Reestudiando huesos: nuevas consideraciones sobre sitios de ltima Esperanza. Anales del Instituto de Patagonia 18: 13356.
Bosco, S. M. G., et al. 1999. Isolation of Sporothrix schenckii in
soil collected at a burrow of a Paracoccidioides brasiliensis
positive armadillo (Dasypus novemcinctus). Foz do Igua/PR, XXI Congresso Brasileiro de Microbiologia, MS 053:
264.
Bibliography
Bosco, S. M. G., et al. 2005. Molecular detection of Paracoccidioides brasiliensis DNA in stool and tissue samples of naturally infected armadillos. Revista do Instituto de Medicina
Tropical de So Paulo 47(Supplement 14): 63.
Bouchard, C., R. M. Malina, and L. Perusse. 1997. Genetics of
fitness and physical performance. Human Kinetics, Champaign, Ill.
Bowman, B. H., T. J. White, and J. W. Taylor. 1996. Human
pathogenic fungi and their close nonpathogenic relatives.
Molecular Phylogenetics and Evolution 6: 8996.
Bowman, B. H., et al. 1992. Molecular evolution of the fungi:
relationship of the Basidiomycetes, Ascomycetes, and Chytridiomycetes. Molecular Biology and Evolution 9: 28596.
Brando, D. 1996. Estudo sobre o padro de revoadas de uma
comunidade de trmites no Parque Nacional das Emas,
Gois. Ph.D. diss., Universidade Estadual de Campinas,
Brazil.
Brandoni, D., et al. 2004. The pes of Pyramiodontherium bergi
(Xenarthra, Tardigrada): the most complete pes of a Tertiary
Megatheriinae. Geodiversitas 26: 64359.
Braun, J. K., and M. M. Daz. 1999. Key to the native mammals
of Catamarca province, Argentina. Occasional Papers, Oklahoma Museum of Natural History 4: 116.
Breece, G. A. 1974. The ecology of the nine-banded armadillo
in southwest Alabama. Masters thesis, Auburn University,
Auburn, Alabama.
Breece, G. A., and M. K. Causey. 1973. Armadillo depredation
of dummy bobwhite quail nests in southwest Alabama.
Proceedings of the Annual Conference of Southeastern Fish
and Wildlife Agencies 27: 1822.
Breece, G. A., and J. L. Dusi. 1985. Food habits and home ranges
of the common long-nosed armadillo Dasypus novemcinctus in Alabama. In G. G. Montgomery, ed., The evolution
and ecology of armadillos, sloths, and vermilinguas, 41927.
Smithsonian Institution Press, Washington, D.C.
Breed, W. G., et al. 1988. Sperm head structure of a murid rodent
from southern Africa: the red veld rat Aethomys chrysophilus. Gamete Research 19: 191202.
Brinkmann, H., and H. Philippe. 1999. Archaea sister group
of Bacteria? Indications from tree reconstruction artifacts
in ancient phylogenies. Molecular Biology and Evolution 16:
81725.
Briskie, J. V., and R. Montgomerie. 1992. Sperm size and sperm
competition in birds. Proceedings of the Royal Society of London, Series B 247: 8995.
Britton, S. W. 1941. Form and function in the sloth. Quarterly
Review of Biology 16: 1334, 190207.
Britton, S. W., and B. F. Kline. 1939. On deslothing the sloth.
Science 90: 1617.
Brown, S. A., K. Langford, and S. Tarver. 1997. Effects of certain
vasoactive agents on the long-term pattern of blood pressure, heart rate, and motor activity in cats. American Journal
of Veterinary Research 58: 64752.
Bruce, S., et al. 2000. Armadillo exposure and Hansens disease:
an epidemiologic survey in southern Texas. Journal of the
American Academy of Dermatology 43: 22328.
Bruno, N. 1998. Hbitos alimenticios de cinco dasypodidos en
el Chaco Boliviano. Technical Report Number 7, Kaa-Iya
Project (WCS, CABI), Santa Cruz, Bolivia.
Bruno, N., and E. Cullar. 2000. Hbitos alimenticios de cinco
323
dasypodidos en el Chaco Boliviano. In E. Cabrera, C. Mercolli, and R. Resquin, eds., Manejo de fauna silvestre en Amazona y Latinoamrica, 40112. CITES Paraguay, Fundacin
Moiss Bertoni, University of Florida, Asuncin, Paraguay.
Buchanan, G. D. 1967. The presence of two conceptuses in the
uterus of a nine-banded armadillo. Journal of Reproduction
and Fertility 13: 32931.
Buchanan, G. D., A. C. Enders, and R. V. Talmage. 1956. Implantation in armadillos ovariectomized during the period of
delayed implantation. Journal of Endocrinology 14: 12128.
Buckland, S. T., et al. 1993. Distance sampling: estimating abundance of biological populations. Chapman and Hall, London.
Buffetaut, E., and R. Hoffstetter. 1977. Dcouverte dun crocodilien Sebecus dans le Miocne du Prou oriental. Compte
Rendu de lAcadmie des Sciences 284: 166366.
Bugge, J. 1979. Cephalic arterial pattern in New World edentates and Old World pangolins with special reference to
their phylogenetic relationships and taxonomy. Acta Anatomica 105: 3746.
Bulyk, M. L. 2003. Computational prediction of transcriptionfactor binding site locations. Genome Biology 5: 201.
Burchfield, H. P., et al. 1976. Improved caging for nine-banded
armadillos. Laboratory Animal Science 26: 23436.
Burggren, W. W. 1999. Developmental physiology, animal models, and the August Krogh Principle. Zoology 102: 14856.
Burke, T., and M. W. Bruford. 1987. DNA fingerprinting in
birds. Nature 327: 14952.
Burleigh, J. G., and S. Mathews. 2004. Phylogenetic signal in
nucleotide data from seed plants: implications for resolving
the seed plant tree of life. American Journal of Botany 91:
1599613.
Burmeister, G. 18701874. Monografa de los glyptodontes en el
Museo Pblico de Buenos Aires. Anales del Museo Pblico
de Buenos Aires 2: 1412.
Burridge, M. J., et al. 1979. Serologic survey for Toxoplasma
gondii in wild animals in Florida. Journal of the American
Veterinary Medical Association 175: 96467.
Bush, M., and B. A. Gilroy. 1979. A bleeding technique from
nonpalpable vessels in anesthetized two-toed sloths (Choloepus didactylus)plus hematologic data. Journal of Zoo
Animal Medicine 10: 2627.
Bushnell, R. 1952. The place of the armadillo in Florida wildlife
communities. Masters thesis, Stetson University, DeLand,
Fla.
Cabanac, A. J., et al. 1999. The structure and blood storing function of the spleen of the hooded seal Cystophora cristata.
Journal of Zoology 248: 7581.
Cabral, A. M. S., C. Peres da Costa, and S. E. Huggins. 1980.
Cardiac output in the three-toed sloth, Bradypus tridactylus.
Comparative Biochemistry and Physiology 67A: 52730.
Cabrera, A. 1957. Catlogo de los mamferos de Amrica del
Sur. Revista del Museo Argentino de Ciencias Naturales Bernardino Rivadavia, Ciencias Zoolgicas 41: 1307.
Cabrera, A., and J. Yepes. 1940. Mamferos Sudamericanos. Historia Natural Ediar, Buenos Aires.
Cabrera, E., C. Mercolli, and R. Resquin, eds. 2000. Manejo
de fauna silvestre en Amazona y Latinoamrica. CITES
Paraguay, Fundacin Moises Bertoni, University of Florida,
Asuncin, Paraguay.
324
Bibliography
Cceres, F. 1995. Actual status of quirquincho Chaetopractus

[sic] nationi, endangered Andean species. Unpublished report, consultancy to DNCB, La Paz, Bolivia.
Cadavid, D., and A. Restrepo. 1993. Factors associated with
Paracoccidiodes brasiliensis infection among permanent
residents of three endemic areas in Colombia. Epidemiology
and Infection 111: 12133.
Caldwell, H. D., W. F. Kirchheimer, and T. M. Buchanan. 1979.
Identification of a Mycobacterium leprae specific protein
antigen(s) and its possible application for the serodiagnosis
of leprosy. International Journal of Leprosy 47: 47783.
Camilo-Alves, C. de S. P. 2003. Adaptaes dos tamandusbandeira (Myrmecophaga tridactyla Linnaeus, 1758) variao da temperatura ambiente no Pantanal da Nhecolndia,
MS. Masters thesis, Universidade Federal de Mato Grosso
do Sul, Brazil.
Camilo-Alves, C. de S. P., and G. Mouro. 2006. Responses of
a specialized insectivorous mammal (Myrmecophaga tridactyla) to variation in ambient temperature. Biotropica 38:
5256.
Campos, J. M., I. Benitez, and D. A. Meritt. 2004. On the occurrence of the owl monkey (Aotus azarae) in Cerro Leon,
Paraguay. Neotropical Primates 12: 5556.
Candia, R., and A. D. Dalmasso. 1995. Dieta del guanaco (Lama
guanicoe) y productividad del pastizal en la Reserva La Payunia, Mendoza (Argentina). Multequina (Mendoza, Argentina) 4: 515.
Candia, R., et al. 1993. Diseo del plan de manejo para la Reserva Provincial La Payunia (Malarge, Mendoza). Multequina
(Mendoza, Argentina) 2: 587.
Cannon, M. D., and D. J. Meltzer. 2004. Early Paleoindian foraging: examining the faunal evidence for large mammal
specialization and regional variability in prey choice. Quaternary Science Reviews 23: 195587.
Cao, Y., N. Okada, and M. Hasegawa. 1997. Phylogenetic position of guinea pigs revisited. Molecular Biology and Evolution 14: 46164.
Cardoso da Silva, J. M., and D. C. Oren. 1993. Observations on
the habitat and distribution of the Brazilian three-banded
armadillo Tolypeutes tricinctus, a threatened Caatinga endemic. Mammalia 57: 14952.
Carleton, J. P. 1996. Cells, molecules and adaptative radiation
in mammals. In H. H. Genoways and R. J. Baker, eds., Contributions in mammalogy: a memorial volume honoring Dr.
J. Knox Jones Jr., 124. Museum of Texas Tech University,
Lubbock.
Carlini, A. A., and G. J. Scillato-Yan. 1996. Chorobates recens
(Xenarthra, Dasypodidae) y un anlisis de la filogenia de
los Euphractini. Revista del Museo de La Plata (NS), Paleontologa IX 59: 22538.
. 2004. The oldest Megalonychidae (Xenarthra: Tardigrada): phylogenetic relationships and an emended diagnosis
of the family. Neues Jahrbuch fr Geologie und Palontologie
Abhandlungen 233: 42343.
Carlini, A. A., G. J. Scillato-Yan, and S. F. Vizcano. 1993. Un
Myrmecophagidae (Vermilingua) del Mioceno de Cerro
Boleadoras (Santa Cruz, Argentina). Ameghiniana 30: 102.
Carlini, A. A., and S. F. Vizcano. 1987. A new record of the
armadillo Chaetophractus vellerosus (Gray, 1865) in the
Buenos Aires province of Argentina: possible causes for the

disjunct distribution. Studies on Neotropical Fauna and Environment 22: 5357.
Carlini, A. A., S. F. Vizcano, and G. J. Scillato-Yan. 1997. Armored xenarthrans: a unique taxonomic and ecologic assemblage. In R. F. Kay, R. H. Madden, R. L. Cifelli, and J.
J. Flynn, eds., Vertebrate paleontology in the Neotropics: the
Miocene fauna of La Venta, Colombia, 21326. Smithsonian
Carlini, A. A., et al. 1989. Un singular Paratheria del Eoceno
temprano de Patagonia, Argentina. Ameghiniana 26: 241.
Carlini, A. A., et al. 1992. Un singular Myrmecophagidae (Xenarthra, Vermilingua) de Edad Colhuehuapense (Oligoceno tardo, Mioceno temprano) de Patagonia, Argentina.
Ameghiniana 29: 176.
Carlini, A. A., et al. 2002. Una nueva especie de megaterio
(Xenarthra, Megatheriidae) de Mioceno Tardo-Plioceno
de Catamarca, Argentina. Ameghiniana 39: 36777.
Carmo, N. A. D. 2002. Distribuio, densidade, padro de atividades, dieta e parasitas de Bradypus tridactylus (Mammalia, Xenarthra) em fragmento florestal na Amaznia Central. Masters thesis, Universidade Federal do Amazonas,
Manaus, Amazonas.
Carretero, E. M., A. Garca, and M. A. Dacar. 2004. First data
on differential use of the environment by Pleistocene megafauna species (San Juan, Argentina). Current Research in the
Pleistocene 21: 9192.
Carson, D. D., et al. 2000. Review: embryo implantation. Developmental Biology 223: 21737.
Cartelle, C., and G. De Iuliis. 1995. Eremotherium laurillardi:
the Panamerican late Pleistocene megatheriid sloth. Journal
. 2006. Eremotherium laurillardi (Lund) (Xenarthra,
Megatheriidae), the Panamerican giant ground sloth: taxonomic aspects of the ontogenetic development of skull and
dentition. Journal of Systematic Paleontology 4: 199209.
Cartelle, C., and M. T. V. Leite. 1989. Myrmecophagidae (Edentata, Vermilingua) do Pleistoceno final no Centro-Oeste da
Bahia. XI Congresso Brasileiro de Paleontologa, Curitiba
Anais 11: 57181.
Carter, A. M., et al. 2006. The fetal membranes of the otter
shrews and a synapomorphy for Afrotheria. Placenta 27:
25868.
Carter, T. S. 1983. The burrows of the giant armadillo, Priodontes maximus (Edentata: Dasypodidae). Sugetierkundliche
Mitteilungen 31: 4753.
. 1985. Armadillos of Brazil. National Geographic Society
Research Reports 20: 1017.
Carter, T. S., and C. D. Encarnao. 1983. Characteristics and
use of burrows by four species of armadillo in Brazil. Journal
Carvalho, C. T. d. 1960. Notes on the three-toed sloth, Bradypus
tridactylus. Mammalia 24: 15556.
Carvalho, R. A., Z. C. Lins-Lainson, and R. Lainson. 1997.
Breeding nine-banded armadillos (Dasypus novemcinctus)
in captivity. Laboratory Animal Science 36: 6668.
Casadevall, A. 2005. Fungal virulence, vertebrate endothermy,
and dinosaur extinction: is there a connection? Fungal Genetics and Biology 42: 98106.
Bibliography
Casadevall, A., J. N. Steenberg, and J. D. Nosanchuk. 2003.

Ready made virulence and dual use virulence factors
in pathogenic environmental fungithe Cryptococcus
neoformans paradigm. Current Opinion in Microbiology 6:
33237.
Casamiquela, R. M., and F. Sepulveda. 1974. Catalogacin critica
de algunos vertebrados fsiles chilenos. III. Los Megatherioideos. Sobre Megatherium medinae Philippi. Ameghiniana
11: 97123.
Casanave, E. B., and J. M. Affanni. 1994. Body temperature of
the armadillo Chaetophractus villosus (Mammalia, Dasypodidae). Archives Internationales de Physiologie, Biochimie
et Biophysique 102: 24346.
. 1995. Decrease of body temperature in armadillos experimentally covered by soil. Archives of Physiology and Biochemistry 103: 2932.
Casanave, E. B., P. M. Bermdez, and N. N. Polini. 2005. Haemostatic mechanisms of the armadillo Chaetophractus villosus (Xenarthra, Dasypodidae). Comparative Clinical Pathology 13: 17175.
. 2006. Principal coagulation factors and natural anticoagulants in the armadillo Chaetophractus villosus (Mammalia, Xenarthra, Dasypodidae). Comparative Clinical Pathology 14: 21016.
Casanave, E. B., L. Garca Samartino, and J. M. Affanni. 1995.
Bradycardia in armadillos experimentally covered with soil.
Archives of Physiology and Biochemistry 103: 5153.
Casanave, E. B., M. C. Manfredi, and E. M. Luengos Vidal. 2003.
Ecologa comportamental de los armadillos en un pastizal
serrano. II Jornadas Interdisciplinarias del Sudoeste Bonaerense, EdiUNS, Baha Blanca 3: 11525.
Casanave, E. B., and N. N. Polini. 1999. Comparative study of
some haematological parameters of two wild Chaetophractus villosus (Mammalia, Dasypodidae) populations. Comparative Haematology International 9: 1316.
Casinos, A. 1996. Bipedalism and quadrupedalism in Megatherium: an attempt at biomechanical reconstruction. Lethaia
29: 8796.
Cassinello, J., et al. 1998. Characteristics of the semen of three
endangered species of gazelles (Gazella damamhorr, G. neglecta and G. cuvieri). Journal of Reproduction and Fertility
113: 3545.
Cassini, M. H. 1993. Searching strategies within food patches in
the armadillo Chaetophractus vellerosus. Animal Behaviour
46: 400402.
Cassini, M. H., A. Kacelnik, and E. T. Segura. 1990. The tale of
the screaming hairy armadillo, the guinea pig and the marginal value theorem. Animal Behaviour 39: 103050.
Castellanos, A. 1931. La librera del Glyptodon de Ameghino.
Cultura, rgano de la biblioteca popular Bernardino Rivadavia 3: 49.
. 1932. Nuevos gneros de gliptodontes en relacin con su
filogenia. Physis 11: 92100.
. 1937. Anotaciones sobre la lnea filogentica de los
clamiterios. Publicaciones del Instituto de Fisiografa y Geologa de la Facultad de Ciencias Matemticas Fsico-Qumicas y Naturales aplicadas a la Industria de la Universidad
Nacional del Litoral, Serie tcnico-cientfica 1(1): 135.
. 1940. A propsito de los gneros Plohophorus, Nopach-
325
thus y Panochthus. Publicaciones del Instituto de Fisiografa

y Geologa de la Facultad de Ciencias Matemticas FsicoQumicas y Naturales aplicadas a la Industria de la Universidad Nacional del Litoral, Serie tcnico-cientfica 1(6): 1279.
Cattoi, N. 1962. Un nuevo Xenarthra del Terciario de Patagonia,
Chubutherium ferelloii gen. et. sp. nov. (Megalonychoidea,
Mylodontidae). Revista del Museo Argentinia Ciencias Nacional Bernardino Rivadavia Zoologa 8: 12333.
Caviglia, S. E., H. D. Yacobaccio, and L. A. Borrero. 1986. Las
Buitreras: convivencia del hombre con fauna extinta en Patagonia Meridional. In A. L. Bryan, ed., New evidence for the
Pleistocene peopling of the Americas, 295317. Center for the
Study of Early Man, Orono, Maine.
Cetica, P., and M. S. Merani. 2004. El espermatozoide tambin
tiene su historia. Ciencia Hoy 14: 1015.
Cetica, P., H. J. Aldana Marcos, and M. S. Merani. 2005. Morphology of female genital tracts in Dasypodidae (Xenarthra, Mammalia): a comparative survey. Zoomorphology
124: 5765.
Cetica, P., et al. 1993. Comparative spermatology in Dasypodidae (Priodontes maximus, Chaetophractus villosus & Dasypus hybridus). Biocell 18: 89103.
Cetica, P., et al. 1997. Comparative spermatology in Dasypodidae II (Chaetophractus vellerosus, Zaedyus pichiy, Euphractus sexcinctus, Tolypeutes matacus, Dasypus septemcinctus &
Dasypus novemcinctus). Biocell 21: 195204.
Cetica, P., et al. 1998. Evolutionary sperm morphology and
morphometry in armadillos. Journal of Submicroscopic Cytology and Pathology 30: 30914.
Chapman, R. E. 1990. Conventional Procrustes approaches. In
F. J. Rohlf and F. L. Bookstein, eds., Proceedings of the Michigan Morphometrics Workshop, 25167. Museum of Zoology,
Special Publication no. 2, University of Michigan.
Chatters, J. C., S. Hackenberger, and H. G. McDonald. 2004.
A Jeffersons ground sloth from the terminal Pleistocene of
Central Washington. Current Research in the Pleistocene 21:
9394.
Cheadle, M. A., et al. 2001. The nine banded armadillo Dasypus
novemcinctus is an intermediate host for Sarcocystis neurona. International Journal of Parasitology 31: 33035.
Chebez, J. C. 1994. Tatu carreta, pichiciego menor, pichiciego
mayor. In J. C. Chebez, ed., Los que se van, 191204. Albatros, Buenos Aires.
Chiarello, A. G. 1998a. Diet of the Atlantic forest maned sloth
Bradypus torquatus (Xenarthra: Bradypodidae). Journal of
Zoology 246: 1119.
. 1998b. Activity budgets and ranging patterns of the Atlantic forest maned sloth Bradypus torquatus (Xenarthra:
Bradypodidae). Journal of Zoology 246: 110.
Chiarello, A. G., et al. 2004. A translocation experiment for the
conservation of maned sloths, Bradypus torquatus (Xenarthra, Bradypodidae). Biological Conservation 118: 42130.
Chimpanzee Sequencing and Analysis Consortium. 2005. Initial sequence of the chimpanzee genome and comparison
with the human genome. Nature 437: 6987.
Christensen, C. G. 1980. Behavior patterns of the nine-banded
armadillo (Dasypus novemcinctus). Masters thesis, Southern Illinois University, Carbondale.
Christiansen, P., and R. A. Faria. 2003. Mass estimation of two
326
Bibliography
fossil ground sloths (Mammalia, Xenarthra, Mylodontidae).

Churakov, G., et al. 2005. A novel abundant family of retroposed elements (DAS-SINEs) in the nine-banded armadillo
(Dasypus novemcinctus). Molecular Biology and Evolution
22: 88693.
Cingolani, H. E., M. M. Mita, and C. P. da Costa. 1979. Car
actersticas mecnicas del miocardio de perezoso (Bradypus
tridactylus). Acta Physiologica Latino Americana 29: 6770.
Cisternas, P. A., and P. J. Armati. 2000. Immune system cell
markers in the Northern brown bandicoot, Isoodon macrourus. Developmental and Comparative Immunology 24:
77182.
Clark, K. A., et al. 1987. Leprosy in armadillos (Dasypus novemcinctus) from Texas. Journal of Wildlife Disease 23: 22024.
Clark, W. K. 1951. Ecological life history of the armadillo in the
eastern Edwards Plateau region. American Midland Naturalist 46: 33758.
Clift, W. 1835. Some account of remains of the Megatherium
sent to England from Buenos Aires by Woodbine Parish.
Transactions of the Geological Society 2: 43750.
Clutton-Brock, T. H. 1989. Mammalian mating systems. Proceedings of the Royal Society of London, Series B 236: 339
72.
Clutton-Brock, T. H., F. E. Guinness, and S. D. Albon. 1982. Red
deer: the ecology of two sexes. University of Chicago Press,
Chicago.
Cole, C. J. 1975. Evolution of parthenogenetic reptiles. In R. Reinboth, ed., Intersexuality in the animal kingdom, 34055.
Springer-Verlag, Berlin.
Cole, F. R., D. M. Reeder, and D. E. Wilson. 1994. A synopsis of
distribution patterns and the conservation of mammal species. Journal of Mammalogy 75: 26676.
Coltrain, J. B., et al. 2004. Rancho La Brea stable isotope biogeochemistry and its implications for the palaeoecology of
late Pleistocene, coastal California. Palaeogeography, Palaeoclimatology, Palaeoecology 205: 199219.
Consentino, L. N. 2004. Aspectos do comportamento da
Preguia Comum, Bradypus variegatus (Xenarthra, Bradypodidae) em uma rea semi-cativeiro no municpio de ValenaRio de Janeiro. Masters Thesis, Universidade Federal
de Rio de Janeiro, Rio de Janeiro, Brazil.
Cooper, D. W., et al. 1993. X inactivation in marsupials and
monotremes. Seminars in Developmental Biology 4: 11728.
Cooper, T. G. 1984. The onset and maintenance of hyperactivated motility of spermatozoa from the mouse. Gamete Research 9: 5594.
Copa, M. E., and W. R. Townsend. 2004. Aprovechamiento de
la fauna por dos comunidades Tsimane: un subsidio del
bosque a la economa familiar. Revista Boliviana de Ecologa
y Conservacin Ambiental 16: 4148.
Coppo, J. A., et al. 1979. Valores hemticos del armadillo Dasypus spp. Gaceta Veterinaria (Buenos Aires) 41: 493501.
Corin-Frederic, J. 1969. Les formulas gonosomiques dites aberrants chez les mammifres Euteries. Chromosoma 27:
26887.
Corredor, G. G., et al. 1999. Isolation of Paracoccidioides brasiliensis from the nine-banded armadillo Dasypus novemcinctus in an endemic area for paracoccidioidomycosis in Colombia. Revista Iberoamericana de Micologia 16: 21620.
Corredor, G. G., et al. 2005. The naked-tailed armadillo Cabas-
sous centralis (Miller 1899): a new host to Paracoccidioides

brasiliensis. Molecular identification of the isolate. Medical
Mycology 43: 27580.
Costa, E. O., L. S. M. Diniz, and C. Fava-Netto. 1995. The prevalence of positive intradermal reactions to paracoccidioidin
in domestic and wild animals in So Paulo, Brazil. Veterinary Research Communications 19: 12730.
Costa, E. O., et al. 1995. Delayed hypersensitivity test with paracoccidioidin in captive Latin Americam wild mammals.
Journal of Medical and Veterinary Mycology 33: 3942.
Costa, F. A. M., et al. 2001. Coccidioidomicose pulmonar em
caador de tatus. Jornal de Pneumologia 27: 27578.
Costa, L. P., et al. 2000. Biogeography of South American forest mammals: endemism and diversity in the Atlantic forest.
Biotropica 32: 87281.
Coutinho, H. B., et al. 1995. Immunocytochemical study on the
ontogeny of the marsupial Didelphis albiventris immune
system. Journal of Anatomy 187: 3746.
Coutinho, M., et al. 1997. Aspectos ecolgicos dos vertebrados
terrestres e semi-aquticos no Pantanal. In Ministrio do
Meio Ambiente, dos Recursos Hdricos e da Amaznia Legal, ed., Brasil: plano de conservao da bacia do alto ParaguaiPCBAP. Diagnsticos do meios fsicos e bitico: meio
bitico, 183322. Braslia, Brazil.
Coutinho, Z. F., et al. 2002. Paracoccidioidomycosis mortality
in Brazil (19801985). Cadernos de Sade Pblica, Rio de Janeiro 18: 144154.
Craig, S. F., et al. 1997. The paradox of polyembryony: a review of cases and a hypothesis for its evolution. Evolutionary
Ecology 11: 12743.
Crawford, C. L. 1992. Leprosy vaccine. Nature 359: 100.
Crawshaw, G. J., and S. E. Oyarzun. 1996. Vertebral hyperostosis in anteaters (Tamandua tetradactyla and Tamandua
mexicana): probable hypervitaminosis A and/or D. Journal
of Zoo and Wildlife Medicine 27: 15869.
Crespo, J. A. 1944. Contribucin al conocimiento de la ecologa
de algunos dasipodidos (Edentata) argentinos. Revista Argentina de Zoogeografia 4: 839.
. 1974. Comentarios sobre nuevas localidades para
mamferos de Argentina y de Bolivia. Revista del Museo
Argentino de Ciencias Naturales Bernardino Rivadavia,
Zoologa 11: 1013.
Croft, D. A. 2001. Cenozoic environmental change in South
America as indicated by mammalian body size distributions
(cenograms). Diversity and Distributions 7: 27187.
Cuba Caparo, A. 1979. Atlas de histologa del armadillo de 7
bandas Dasypus hybridus. 1st ed. Centro Panamericano de
Zoonosis, OEA, Buenos Aires.
Cullar, E. 1999. Hunting of armadillos by Izoceo Indians of
the Gran Chaco, Bolivia. Masters thesis, University of Kent,
Canterbury, England.
. 2001. The tatujeikurajoyava (Chlamyphorus retusus) in
the Izozog communities of the Bolivian Gran Chaco. Edentata 4: 1415.
. 2002. Census of the three-banded armadillo Tolypeutes
matacus using dogs, southern Chaco, Bolivia. Mammalia
66: 44851.
Cullar, E., and A. Noss. 2003. Mamferos del Chaco y de la
Chiquitania de Santa Cruz, Bolivia. Editorial FAN, Santa
Cruz de la Sierra, Bolivia.
Cuellar, O. 1977. Animal parthenogenesis. Science 197: 83743.
Bibliography
. 1987. The evolution of parthenogenesis: a historical

perspective. In P. B. Moens, ed., Meiosis, 43104. Academic
Press, New York.
Cullar, R. L. 1997. Aprovechamiento de la fauna silvestre en
una comunidad de agricultores: los guaranes de Akae, Santa Cruz, Bolivia. Masters thesis, Universidad Mayor de San
Andrs, La Paz, Bolivia.
. 1999. Aprovechamiento de la fauna silvestre en una comunidad de agricultores: los guaranes de Akae, Santa Cruz,
Bolivia. In T. G. Fang, O. L. Montenegro, and R. E. Bodmer,
eds., Manejo y conservacin de fauna silvestre en Amrica
Latina, 14748. Instituto de Ecologa, La Paz, Bolivia.
. 2000. Uso de los animales silvestres por pobladores
Izoceos. In E. Cabrera, C. Mercolli, and R. Resquin, eds.,
Manejo de fauna silvestre en Amazona y Latinoamrica,
47184. CITES Paraguay, Fundacin Moises Bertoni, University of Florida, Asuncin, Paraguay.
Cully, W. 1939. Day and night in the life of an armadillo. Bulletin
of the New York Zoological Society 42: 18082.
Cummins, J. M., and P. F. Woodall. 1985. On mammalian
sperm dimensions. Journal of Reproduction and Fertility
75: 15375.
Currah, R. S. 1985. Taxonomy of the Onygenales: Arthrodermataceae, Gymnoascaceae, Myxotrichaceae and Onygenaceae.
Mycotaxon 24: 1216.
Cuvier, G. 1796. Notice sur le squelette dune trs-grande espce de quadrupde inconnue jusqu prsent, trouv au
Paraquay, et dpos au cabinet dhistoire naturelle de Madrid. Magasin Encyclopdique: ou Journal des Sciences, des
Lettres et des Arts 1: 30310.
. 1823. Rchrches sur les ossemens fossiles. 2nd ed. Dufour
& DOcagne, Paris.
Czaplewski, N. J., and C. Cartelle. 1998. Pleistocene bats from
cave deposits in Bahia, Brazil. Journal of Mammalogy 79:
784803.
Czekala, N. M., et al. 1980. Annual circulating testosterone
levels in captive and free-ranging male armadillos (Dasypus novemcinctus). Journal of Reproduction and Fertility 59:
199204.
DAddamio, G. H., J. D. Roussel, and E. E. Storrs. 1977. Response
of the nine-banded armadillo (Dasypus novemcinctus) to
gonadotropins and steroids. Laboratory Animal Science 27:
48289.
Dalcq, A. M. 1973. Cyto-morphologie normale du testicule
et spermatogense chez les mammifres. Mmoires de
lAcadmie Royale de Mdecine de Belgique, Palais des Acadmies, Bruxelles 46: 531648.
Dalponte, J. C., and J. A. Tavares-Filho. 2004. Diet of the yellow
armadillo, Euphractus sexcinctus, in south-central Brazil.
Edentata 6: 3741.
Damuth, J. 1990. Problems in estimating body masses of archaic
ungulates using dental measurements. In J. Damuth and B.
J. MacFadden, eds., Body size in mammalian paleobiology:
estimation and biological implications, 22953. Cambridge
University Press, Cambridge.
Darwin, C. 1839. Voyage of the Beagle. John Colburn, London.
Davies, A. G., E. L. Bennett, and P. G. Waterman. 1988. Food
selection by two south-east Asian colobine monkeys (Presbytes rubicunda and Presbytes melalophos) in relation to
plant chemistry. Biological Journal of the Linnean Society 34:
3356.
327
Davies, D. V. 1947. The cardiovascular system of the slow loris

(Nycticebus tardigradus malaianus). Proceedings of the Zoological Society of London 117: 377410.
Dawkins, R. 1976. The selfish gene. Oxford University Press,
Oxford.
De Iuliis, G. 1994. Relationships of the Megatheriinae, Nothrotheriinae, and Planopsinae: some skeletal characteristics
and their importance for phylogeny. Journal of Vertebrate
Paleontology 14: 57791.
. 1996. A systematic review of the Megatheriinae (Mammalia: Xenarthra: Megatheriidae). Ph.D. diss., University of
Toronto, Ontario.
. 2006. On the status of Megatherium sundti Philippi 1893
(Mammalia: Xenarthra: Megatheriidae). Ameghiniana 43:
16169.
De Iuliis, G., M. S. Bargo, and S. F. Vizcano. 2000. Skull morphology variation and mastication in the fossil giant armadillos Pampatherium spp.; with remarks on related genera
(Mammalia: Xenarthra: Pampatheriidae). Journal of Vertebrate Paleontology 20: 743754.
De Iuliis, G., D. Brandoni, and G. J. Scillato-Yan. Forthcoming. New remains of Megathericulus patagonicus Ameghino,
1904 (Xenarthra, Megatheriidae): information on primitive
features of megatheriines. Journal of Vertebrate Paleontology.
De Iuliis, G., and C. Cartelle. 1993. The medial carpal and metacarpal elements of Eremotherium and Megatherium (Xenarthra, Mammalia). Journal of Vertebrate Paleontology 13:
52533.
. 1999. A new giant megatheriin ground sloth (Mammalia: Xenarthra: Megatheriidae) from the late Blancan to early
Irvingtonian of Florida. Zoological Journal of the Linnean
Society 127: 495515.
De Iuliis, G., G. H. R, and S. F. Vizcano. 2004. The Toro Negro megatheriine (Mammalia, Xenarthra): a new species
of Pyramiodontherium and a review of Plesiomegatherium.
Journal of Vertebrate Paleontology 24: 21427.
De Iuliis, G., and P.-A. St-Andr. 1997. Eremotherium sefvei nov.
sp. (Mammalia, Xenarthra, Megatheriidae) from the Pleistocene of Ulloma, Bolivia. Geobios 30: 45361.
De Iuliis, G., et al. 2005. El legado del megaterio. Revista Museo
3: 3136.
de Jong, W. W., J. A. M. Leunissen, and G. J. Wistow. 1993. Eye
lens crystallins and the phylogeny of placental orders: evidence for a Macrocelid-Paenungulate clade? In F. S. Szalay,
M. J. Novacek, and M. C. McKenna, eds., Mammal phylogeny (Placentals), 512. Springer-Verlag, New York.
de Jong, W. W., et al. 1985. Protein sequence analysis applied to
Xenarthran and Pholidote phylogeny. In G. G. Montgomery, ed., The evolution and ecology of armadillos, sloths and
vermilinguas, 6576. Smithsonian Institution Press, Washington, D.C.
de Jong, W. W., et al. 2003. Indels in protein-coding sequences
of Euarchontoglires constrain the rooting of the eutherian
tree. Molecular Phylogenetics and Evolution 28: 32840.
de Lange, D. 1926. Quelques remarques sur la placentation de
Bradypus. Comptes Rendues Association Anatomie Lige 21:
32133.
. 1933. Placentarbildung. Handbuch der vergleichen Anatomie 6: 155234.
De Souza-Mazurek, R. R., et al. 2000. Subsistence hunting
328
Bibliography
among the Waimiri Atroari Indians in central Amazonia,

Brazil. Biodiversity and Conservation 9: 579596.
Deferrari, G., et al. 2002. Presencia de Chaetophractus villosus
en Tierra del Fuego: nueva especie introducida, Nuevo
problema? Abstracts XVII Jornadas Argentinas de Mastozoologa 17: 24.
Delsuc, F., H. Brinkmann, and H. Philippe. 2005. Phylogenomics and the reconstruction of the tree of life. Nature Reviews
Genetics 6: 361375.
Delsuc, F., M. J. Stanhope, and E. J. P. Douzery. 2003. Molecular
systematics of armadillos (Xenarthra, Dasypodidae): contribution of maximum likelihood and Bayesian analyses of
mitochondrial and nuclear genes. Molecular Phylogenetics
and Evolution 28: 26175.
Delsuc, F., S. F. Vizcano, and E. J. P. Douzery. 2004. Influence of
Tertiary paleoenvironmental changes on the diversification
of South American mammals: a relaxed molecular clock
study within xenarthrans. BMC Evolutionary Biology 4: 11.
Delsuc, F., et al. 2001. The evolution of armadillos, anteaters
and sloths depicted by nuclear and mitochondrial phylogenies: implications for the status of the enigmatic fossil
Eurotamandua. Proceedings of the Royal Society of London,
Series B 268: 160515.
Delsuc, F., et al. 2002. Molecular phylogeny of living xenarthrans and the impact of character and taxon sampling on
the placental tree rooting. Molecular Biology and Evolution
19: 165671.
Denson, R. D. 1979. Aggression and tumbling among armadillos. Southwestern Naturalist 24: 69798.
Deps, P. D., N. S. Michalany, and J. Tomimori-Yamashita. 2004.
False positive reaction of the immunohistochemistry technique using anti-BCG polyclonal antibodies to identify Mycobacterium leprae in wild nine-banded armadillos. International Journal of Leprosy 72: 32730.
Deps, P. D., A. R. Santos, and J. Yamashita-Tomimori. 2002.
Detection of Mycobacterium leprae DNA by PCR in blood
sample from nine-banded armadillo: preliminary results.
International Journal of Leprosy 70: 3435.
Desbiez, A. L. J., P. A. L. Borges, and I. M. Medri. 2006. Chasing
behavior in yellow armadillos, Euphractus sexcinctus, in the
Brazilian Pantanal. Edentata 7: 5153.
Dewey, C., et al. 2004. Accurate identification of novel human
genes through simultaneous gene prediction in human,
mouse, and rat. Genome Research 14: 66164.
Daz, G. B., and R. Ojeda. 2000. Libro rojo de los mamferos
amenazados de la Argentina. Sociedad Argentina para el
Estudio de los Mamferos, Mendoza.
Daz, M. M. 2000. Key to the native mammals of Jujuy province, Argentina. Occasional Papers, Oklahoma Museum of
Natural History 7: 120.
Dillehay, T. 2000. The settlement of the Americas, a new prehistory. Basic Books, New York.
Diniz, L. S. M., E. O. Costa, and P. M. A. Oliveira. 1997. Clinical
disorders in armadillos (Dasypodidae, Edentata) in captivity. Journal of Veterinary Medicine 44: 577582.
. 1995. Clinical disorders observed in anteaters (Myrmecophagidae, Edentata) in captivity. Veterinary Research
Communications 19: 409415.
Diniz, L. S. M., and P. M. Oliveira. 1999. Clinical problems of
sloths (Bradypus sp. and Choloepus sp.) in captivity. Journal
of Zoo and Wildlife Medicine 30: 7680.
Dobigny, G., et al. 2005. Low rate of genomic repatterning in

Xenarthra inferred from chromosome painting data. Chromosome Research 13: 65163.
DOrbigny, A. D. 1999. Viaje por Amrica Meridional II. Emec,
Buenos Aires, Argentina. (reprint of work originally published in 1835)
Dosapey, T. 2003. Importancia medicinal y uso tradicional de
cinco mamferos y un reptil de la comunidad Ayorea Santa
Teresita. In R. L. Cullar, ed., I Congreso de investigadores en
el Parque Nacional Kaa-Iya del Gran Chaco y zonas de influencia: Memoria (La Brecha, 12 de Octubre de 2002), 6062.
Kaa-Iya Project (WCS, CABI), Santa Cruz, Bolivia.
Douglass, J. F., and C. E. Winegarner. 1977. Predators of eggs
and young of the gopher tortoise, Gopherus polyphemus
(Reptilia, Testidines, Testudinidae) in southern Florida.
Journal of Herpetology 11: 23638.
Downing, K. F., and R. S. White. 1995. The cingulates (Xenarthra) of the Leisey Shell Pit local fauna (Irvingtonian), Hillsborough County, Florida. Bulletin of the Florida Museum of
Dresdner, R. D., and D. F. Katz. 1981. Relationship of mammalian sperm motility and morphology to hydrodynamic
aspects of the cell function. Biology of Reproduction 25:
92030.
Drumond, M. A. 1992. Padres de forrageamento do tamandubandeira (Myrmecophaga tridactyla) no Parque Nacional da
Serra da Canastra. Masters thesis, Universidade Federal de
Minas Gerais, Brazil.
Duarte, D. P. F., C. P. da Costa, and S. E. Huggins. 1982. The
effects of posture on blood pressure and heart rate in the
three-toed sloth, Bradypus tridactylus. Comparative Biochemistry and Physiology 73A: 697702.
. 1983. The autonomic component in blood pressure and
heart rate: effects of tilting in the three-toed sloth. Comparative Biochemistry and Physiology 74C: 48592.
Duarte, D. P. F., et al. 1987. Responses of the three-toed sloth,
Bradypus tridactylus, to some commonly used pharmacologic agents. I. Autonomic drugs. Comparative Biochemistry
and Physiology 88C: 27585.
Duarte, D. P. F., et al. 2003. Circadian rhythms in blood pressure in free-ranging three-toed sloths (Bradypus variegatus). Brazilian Journal of Medical and Biological Research
34: 27378.
Duarte, D. P. F., et al. 2004. Cardiovascular responses to locomotor activity and feeding in unrestrained three-toed
sloths, Bradypus variegatus. Brazilian Journal of Medical and
Biological Research 37: 155761.
Dustan, R. W., et al. 1986. Feline sporotrichosis: a report of five
cases with transmission to humans. Journal of the American
Academy of Dermatology 15: 3745.
Dybas, L. K., and H. S. Dybas. 1981. Coadaptation and taxonomic differentiation of sperm and spermathecae in featherwing beetles. Evolution 35: 16874.
Eddy, S. R. 2005. A model of the statistical power of comparative genome sequence analysis. PLoS Biology 3: e10.
Edmonds, R. H., and F. Nagy. 1974. Fine structure of an edentate (Dasypus novemcinctus) male reproductive system: the
epididymis. Journal of Anatomy 118: 33751.
Eisenberg, J. F. 1961. Observations on the nest-building behavior of armadillos. Proceedings of the Zoological Society of
London 137: 32224.
Bibliography
. 1989. Mammals of the Neotropics. Vol. 1, The northern

Neotropics. University of Chicago Press, Chicago.
Eisenberg, J. F., and E. Maliniak. 1978. Reproduction of the
two-toed sloth Choloepus hoffmanni in captivity. Abstract,
58th Annual Meeting of the American Society of Mammalogists.
. 1985. Maintenance and reproduction of the two-toed
sloth Choloepus didactylus in captivity. In G. G. Montgomery, ed., The evolution and ecology of armadillos, sloths, and
Eisenberg, J. F., and K. H. Redford. 1999. Mammals of the Neotropics. Vol. 3, The central Neotropics. University of Chicago
Press, Chicago.
Eisenberg, J. F., and R. W. Thorington, Jr. 1973. A preliminary
analysis of a Neotropical mammal fauna. Biotropica 5:
15061.
Eisenmann, V. 1984. Sur quelques caractres adaptatifs du
squelette dEquus et leurs implications palocologiques.
Bulletin du Musum National dHistoire Naturelle de Paris,
4me srie, 6, section C, 2: 18595.
Eisenmann, V., and P. Sondaar. 1998. Pliocene vertebrates locality of alta, Ankara, Turkey. 7. Hipparion. Geodiversitas 20:
40939.
Eisentraut, M. 19321933. Biologische Studien im bolivianischen Chaco. IV. Die Wrmeregulation beim Kugelgrteltier
(Tolypeutes conurus Js. Geoff.). Zeitschrift fr Vergleichende
Physiologie 18: 17485.
Elrod, D. A. 2000. Mitochondrial DNA sequence variation in
populations of the nine-banded armadillo, Dasypus novemcinctus. Ph.D. diss., University of North Texas, Denton.
Emmons, L. H., and F. Feer. 1990. Neotropical rainforest mammals: a field guide. University of Chicago Press, Chicago.
Encke, B. 1978. Sieben Jahre Tamanduas (Tamandua tetradactyla) im Krefelder Zoo. Zoologischer Garten N.F. 48: 1930.
Enders, A. C. 1960a. Development and structure of the villous
haemochorial placenta of the nine-banded armadillo (Dasypus novemcinctus). Journal of Anatomy 94: 3445.
. 1960b. Electron microscopic observations on the villous
haemochorial placenta of the nine-banded armadillo (Dasypus novemcinctus). Journal of Anatomy 94: 20515.
. 1960c. A histological study of the cortex of the ovary of
the adult armadillo, with special reference to the question of
neoformation of oocytes. Anatomical Record 136: 491499.
. 1962. The structure of the armadillo blastocyst. Journal
of Anatomy 96: 3948.
. 1963. Fine structural studies of implantation in the armadillo. In A. C. Enders, ed., Delayed implantation, 28190.
University of Chicago Press, Chicago.
. 1966. The reproductive cycle of the nine-banded armadillo (Dasypus novemcinctus). In I. W. Rowlands, ed.,
Comparative biology of reproduction in mammals, 295310.
Academic Press, New York.
. 2002. Implantation in the nine-banded armadillo: how
does a single blastocyst form four embryos? Placenta 23:
7186.
Enders, A. C., and G. D. Buchanan. 1959. The reproductive tract
of the female nine-banded armadillo. Texas Reports in Biology and Medicine 17: 32340.
Enders, A. C., G. D. Buchanan, and R. V. Talmage. 1958. Histological and histochemical observations on the armadillo
329
uterus during the delayed and post-implantation periods.

Anatomical Record 130: 63957.
Enders, A. C., and A. O. Welsh. 1993. Structural interactions of
trophoblast and uterus during hemochorial placenta formation. Journal of Experimental Zoology 266: 57887.
Engelmann, G. 1978. The logic of phylogenetic analysis and the
phylogeny of the Xenarthra. Ph.D. diss., Columbia University, New York.
. 1985. The phylogeny of the Xenarthra. In G. G. Montgomery, ed., The evolution and ecology of armadillos, sloths,
and vermilinguas, 5164. Smithsonian Institution Press,
Washington, D.C.
. 1987. A new Deseadan sloth (Mammalia: Xenarthra)
from Salla, Bolivia, and its implications for the primitive
condition of the dentition in edentates. Journal of Vertebrate
Paleontology 7: 21723.
Engeman, R. M., et al. 2003. Monitoring predators to optimize
their management for marine turtle nest protection. Biological Conservation 13: 17178.
Enna, C. D., et al. 1979. Leprosy in the United States. Public
Health Reports 93: 46873.
Ergueta S., P., and C. de Morales. 1996. Libro rojo de los vertebrados de Bolivia. Centro de Datos para la Conservacin, La
Paz, Bolivia.
Esnault, C., and J. C. Nicolle. 1976. Evolution de lADN et des
protines nuclaires basiques au cours de la maturation
des cellules germinales du blier. Annales dhistochimie 21:
18997.
Eullio, K. D., et al. 2001. Coccidioides immitis isolated from
armadillos (Dasypus novemcinctus) in the state of Piaui,
northeast Brazil. Mycopathologia 149: 5761.
European Association of Zoos and Aquaria (EAZA). 2005. European breeding programmes. <http://www.eaza.net>. Accessed 30 June 2005.
Falkner, T. 1774. Descripcin de la Patagonia y de las parte contiguas de la america del sur: traduccin y notas de Samuel A.
Lafone Quevedo, corregida por Salvador Canals Frau. Hachette, Buenos Aires (reprinted in 1957).
Fang, T. G., O. L. Montenegro, and R. E. Bodmer, eds. 1999.
Manejo y conservacin de fauna silvestre en Amrica Latina.
Instituto de Ecologa, La Paz, Bolivia.
Faria, R. A. 1985. Some functional aspects of mastication in
Glyptodontidae (Mammalia). Fortschritte der Zoologie 30:
27780.
. 1988. Observaciones adicionales sobre la biomecnica
masticatoria en Glyptodontidae (Mammalia, Edentata). Boletn de la Sociedad Zoolgica 4: 59.
. 1995. Limb bone strength and habits in large glyptodonts. Lethaia 28: 18996.
. 1996. Trophic relationships among Lujanian mammals.
Evolutionary Theory 11: 12534.
. 2002. Megatherium, el pelado: sobre la apariencia de los
grandes perezosos (Mammalia; Xenarthra) cuaternarios.
Faria, R. A., and R. E. Blanco. 1996. Megatherium, the stabber. Proceedings of the Royal Society, London, Series B, 263:
172529.
Faria, R. A., and S. F. Vizcano. 1997. Allometry of the leg
bones in armadillos (Mammalia, Dasypodidae): a comparison with other mammals. Zeitschrift fr Sugetierkunde 62:
6570.
330
Bibliography
. 2001. Carved teeth and strange jaws: how glyptodonts

masticated. Acta Paleontologica Polonica 46: 87102.
. 2003. Slow moving or browsers? A note on nomenclature. Senckenbergiana biologica 83: 34.
Faria, R. A., S. F. Vizcano, and M. S. Bargo. 1998. Body mass
estimations in Lujanian (late Pleistocene-early Holocene of
South America) mammal megafauna. Mastozoologa Neotropical 5: 87108.
Faria, R. A., S. F. Vizcano, and G. Storch, eds. 2003. Morphological studies in fossil and extant Xenarthra (Mammalia).
Fava, S. C., and C. Fava-Netto. 1998. Epidemiologic surveys
of histoplasmin and paracoccidioidin sensitivity in Brazil.
Revista do Instituto de Medicina Tropical de So Paulo 40:
15564.
Fawcett, D. W. 1970. A comparative view of sperm ultrastructure. Biology of Reproduction, Supplement 2: 90127.
Feito, R., and M. Gallardo. 1982a. Sperm morphology of the
Chilean species of Ctenomys (Octontidae). Journal of Mammalogy 63: 65861.
. 1982b. Ultrastructure of Ctenomys maulinus spermatozoon with special reference to the nucleus. Gamete Research
5: 31721.
Flix, C. H., C. P. da Costa, and W. A. Zin. 1988. Propriedades mecnicas passivas do pulmo e parede torcica em
preguias (B.variegatus) anestesiadas. Resumos da III Reunio Anual da Federao das Sociedades de Biologia Experimental 2.56: 82.
Flix, C. H., et al. 1987. Morfologia do pulmo da preguia
(Bradypus variegatus). Resumos da II Reunio Anual da Federao das Sociedades de Biologia Experimental 2.50: 182.
Ferigolo, J. 1985. Evolutionary trends of the histological pattern
in the teeth of Edentata (Xenarthra). Archives of Oral Biology 30: 7182.
Fernandes, G. F., et al. 2004. IgM and IgG antibody response
to Paracoccidioides brasilensis in naturally infected wild
armadillos (Dasypus novemcinctus). Medical Mycology 42:
36368.
Fernandes-Pinto, E., and A. C. Krger. 2000. Utilizao da fauna silvestre na regio sudoeste do Estado do Paran, Brasil.
In E. Cabrera, C. Mercolli, and R. Resquin, eds., Manejo
de fauna silvestre en Amazona y Latinoamrica, 48594.
CITES Paraguay, Fundacin Moises Bertoni, University of
Florida, Asuncin, Paraguay.
Fernandez, M. 1909. Beitrge zur Embryologie der grteltiere.
I. Zur Keimblatter-inversion und spezifischen polyembryonie der Mulita (Tatusia hybrida Desm.). Morphologische
Jahrbuch 39: 30233.
. 1914. Zur Anordnung der Embryonen und Form der
Placenta bei Tatusia novemcincta. Anatomischer Anzeiger
46: 253.
. 1915a. Ueber einige Entwickelungsstadien des peludo
(Dasypus villosus) und ihre Beziehung zum Problem der
spezifischen Polyembryonie des Genus Tatusia. Anatomische Anzeiger 48: 30527.
. 1915b. Die Entwick der Mulita. Revista del Museo de La
Plata 21: 481.
Fernicola, J. C. 2005. Anlisis filogentico de la familia Glyptodontidae Gray 1869. Ph.D. diss., Universidad de la Repblica, Montevideo, Uruguay.
Ferrari, C. C., et al. 1997. Breeding in captivity of the southern

lesser long-nosed armadillo Dasypus hybridus (Edentata =
Xenarthra: Dasypodidae). Zoocriaderos 2: 15.
Filice, G. A., R. N. Greenberg, and D. W. Fraser. 1977. Lack of
observed association between armadillo contact and leprosy
in humans. American Journal of Tropical Medicine and Hygiene 26: 13739.
Fisher, M. C., et al. 2001. Biogeographic range expansion into
South America by Coccidioides immitis mirrors New World
patterns of human migration. Proceedings of the National
Academy of Sciences, U.S.A. 98: 455862.
Fitch, H. S., P. Goodrum, and C. Newman. 1952. The armadillo
in the southeastern United States. Journal of Mammalogy
33: 2137.
Fleury, R. N., et al. 2001. Zoonotic sporotrichosis. Transmission to humans by infected domestic cat scratching: report
of four cases in So Paulo, Brazil. International Journal of
Dermatology 40: 31822.
Flint, M. P. 1997. Zoo standards for keeping xenarthrans (edentates) in captivity. In A. Shoemaker, ed., Minimum husbandry guidelines for mammals, 1524. American Zoo and
Aquarium Association, Bethesda, Maryland.
Flower, W. H. 1882. On the mutual affinities of the animals
composing the order Edentata. Proceedings of the Zoological
Society of London 25: 35867.
Flynn, J. J., and C. C. Swisher III. 1995. Cenozoic South American land mammal ages: correlation to global chronologies.
Society for Sedimentary Geology (SEPM), Special Publication
54: 31733.
Flynn, J. J., et al. 2003. The Tinguiririca fauna, Chile: biochronology, paleoecology, biogeography, and a new earliest Oligocene South American Land Mammal Age. Palaeogeography, Palaeoclimatology, Palaeoecology 195: 22959.
Flynn, J. J., et al. 2005. Geochronology of Hemphillian-Blancan
age strata, Guanajuato, Mexico, and implications for timing
of the Great American Biotic Interchange. Journal of Geology 113: 287307.
Foley, W. J., W. V. Engelhardt, and P. Charles-Dominique. 1995.
The passage of digesta, particle size and in vitro fermentation rate in the three-toed sloth Bradypus tridactylus (Edentata: Bradypodidae). Journal of Zoology 236: 68196.
Fonseca, G. A. B., and J. M. Aguiar. 2004. The 2004 Edentate
Species Assessment Workshop: introduction. Edentata 6:
12.
Fonseca, G. A. B., and M. T. Fonseca. 1997. Editorial. Edentata
3: 2.
Fonseca, G. A. B., et al. 1994. Livro vermelho dos mamferos
Brasileiros ameaados de extino. Fundao Biodiversitas,
Belo Horizonte, Brazil.
Fonseca, G. A. B., et al. 1999. Brazilian Cerrado. In R. A. Mittermeier, N. Meyers, P. R. Gil, and C. G. Mittermeier, eds.,
Hotspots, 14855. CEMEX/Conservation International,
Mexico City.
Fontana, C. S., G. A. Bencke, and R. E. Reis, eds. 2003. Livro
vermelho da fauna ameaada de extino no Rio Grande do
Sul. Edipucrs, Porto Alegra, Brazil.
Foster, R. B., and S. P. Hubbell. 1990. The floristic composition
of the Barro Colorado Island. In A. H. Gentry, ed., Four Neotropical rainforests, 8598. Yale University Press, New Haven
and London.
Bibliography
Foster, S. A., and J. A. Endler, eds. 1999. Geographic variation in

behavior: perspectives on evolutionary mechanisms. Oxford
University Press, Oxford.
Fox, M. D., D. C. Anderson, and A. F. Kaufman. 1977. Leprosylike disease in a wild armadilloMississippi. Leprosy Scientific Memoranda L865: 5.
Franco, M., et al. 1989. Chlamydospore formation by Paracoccidioides brasiliensis mycelial form. Revista do Instituto de
Medicina Tropical de So Paulo 31: 15157.
Franco, M., et al., eds. 1994. Paracoccidioidomycosis. 2nd ed.
CRC Press, Boca Raton, Fla.
Franco, M., et al. 2000. A critical analysis of isolation of Paracoccidioides brasiliensis from soil. Medical Mycology 38:
18591.
Franzen, A. 1977. Sperm structure with regard to fertilization
biology and phylogenetics. Verhandlungen der Deutschen
Zoologischen Gessellschaft 1977: 12338.
Franzen, M. 2001. Las pautas de cacera de los Huaorani: la reaccin a la economa-mercado en la Amazona del Ecuador.
In J. P. Jorgenson and M. C. Rodrguez, eds., Conservacin
y desarrollo sostenible del Parque Nacional Yasun y su rea
de influencia: Memorias del seminario-taller 2001, 12836.
WCS, Ministerio del Ambiente, UNESCO, Quito, Ecuador.
Frappell, P. B., D. F. Boggs, and D. L. Kilgore, Jr. 1998. How
stiff is the armadillo? A comparison with the allometrics of
mammalian respiratory mechanics. Respiration Physiology
113: 11122.
Freedman, M. H., and E. F. Saunders. 1985. Hematopoiesis
in the human spleen. American Journal of Hematology 11:
27175.
Freeman, P. W., and H. H. Genoways. 1998. Recent northern
records of the nine-banded armadillo (Dasypodidae) in Nebraska. Southwestern Naturalist 43: 491504.
Freiberger, H. G., and H. Fudenberg. 1981. An appetite for armadillo. Hospital Practice 16: 137.
Frutos, S. D., and R. A. Van Den Bussche. 2002. Genetic diversity and gene flow in nine-banded armadillos in Paraguay.
Journal of Mammalogy 83: 81523.
Gage, M. J. 1998. Mammalian sperm morphometry. Proceedings
of the Royal Society, London, Series B 265: 97103.
Gaitn, M. B. 1999. Patrones de cacera en una comunidad indgena Ticuna de la Amazona Colombiana. In T. G. Fang,
O. L. Montenegro, and R. E. Bodmer, eds., Manejo y conservacin de fauna silvestre en Amrica Latina, 7175. Instituto
de Ecologa, La Paz, Bolivia.
Galbreath, G. J. 1980. Aspects of natural selection in Dasypus
novemcinctus. Ph.D. diss., University of Chicago, Chicago,
Ill.
. 1982. Armadillo. In J. A. Chapman and G. A. Feldhamer,
eds., Wild mammals of North America, 7179. Johns Hopkins University Press, Baltimore.
. 1985. The evolution of monozygotic polyembryony in
Dasypus. In G. G. Montgomery, ed., The evolution and ecology of armadillos, sloths, and vermilinguas, 24346. Smithsonian Institution Press, Washington, D.C.
Galndez, E. J., and M. C. Aggio. 1998. The spleen of Mustelus
schmitti (Chondrichthyes, Triakidae). A light and electron
microscopic study. Ichthyological Research 45: 17986.
Galndez, E. J., S. M. Codn, and E. B. Casanave. 2000. Spleen
of Dasypus hybridus (Mammalia, Dasypodidae): a light
331
and electron microscopic study. Anatomical Record 258:

28691.
Galndez, E. J., S. Estecondo, and E. B. Casanave. 2003. The
spleen of Zaedyus pichiy (Mammalia, Dasypodidae): a light
and electron microscopic study. Anatomia, Histologia, Embryologia 34: 19499.
. 2006. The spleen of a specially adapted mammal, the
little hairy armadillo Chaetophractus vellerosus, (Xenarthra,
Dasypodidae). A light and electron microscopic study. International Journal of Morphology 24: 33948.
Galndez, E. J., et al. 1997. Estructura esplnica de Chaetophractus villosus: adulto y neonato. Revista Brasileira de Biologia
57: 393401.
Gallego, M., et al. 1995. Ultrastructural identification of the
splenic follicular dendritic cells in the chicken. Anatomical
Record 242: 22024.
Galliari, C. A., W. D. Berman, and F. J. Goin. 1991. Situacin
ambiental de la provincia de Buenos Aires. A. Recursos y
rasgos naturales en la evaluacin ambiental: mamferos.
Special Publication of the Comisin de Investigaciones Cientficas (CIC) de la Provincia de Buenos Aires, Ao I 5: 335.
Galton, F. 1875. The history of twins, as a criterion of the relative
powers of nature and nurture. Frasers Magazine November:
566576.
Galvo de Moura Filho, A., S. E. Huggins, and S. G. Lines. 1983.
Sleep and waking in the three-toed sloth, Bradypus tridactylus. Comparative Biochemistry and Physiology 76A: 34555.
Gamboa, G. J., H. K. Reeve, and W. G. Holmes. 1991. Conceptual issues and methodology in kin-recognition research: a
critical discussion. Ethology 88: 10927.
Gammons, D. 2006. Radiotelemetry studies of armadillos in
southwestern Georgia. Masters thesis, University of Georgia, Athens.
Garcia, J. E. 2003. Genetic studies in living Xenarthra (a review) (Mammalia, Xenarthra). Senckenbergiana biologica
83: 1926.
Garcia, J. E., et al. 2005. Identification of microsatellite DNA
markers for the giant anteater Myrmecophaga tridactyla.
Journal of Heredity 96: 600602.
Garca Samartino, L., et al. 1987. On the presence of a peculiar
alpha rhythm in the olfactory tubercle of waking armadillos. Electroencephalography and Clinical Neurophysiology
66: 18590.
Gardner, A. L. 2005. Order Cingulata and order Pilosa. In D. E.
Wilson and D. M. Reeder, eds., Mammal species of the world.
Vol. 1, 94103. Johns Hopkins University Press, Baltimore.
Garland, T. D., Jr., and P. A. Carter. 1994. Evolutionary physiology. Annual Reviews of Physiology 56: 579621.
Gaudin, T. J. 1995. The ear region of edentates and the phylogeny of the Tardigrada (Mammalia, Xenarthra). Journal of
Vertebrate Paleontology 15: 672705.
. 1999a. Xenarthra. In R. S. Singer, ed., Encyclopedia of
paleontology. Vol. 2, 134753. Fitzroy Dearborn, Chicago.
. 1999b. The morphology of xenarthrous vertebrae (Mammalia, Xenarthra). Fieldiana (Geology), n.s. 41: 138.
. 2003. Phylogeny of the Xenarthra (Mammalia). Senckenbergiana biologica 83: 2740.
. 2004a. Phylogenetic relationships among sloths (Mammalia, Xenarthra, Tardigrada): the craniodental evidence.
Zoological Journal of the Linnean Society 140: 255305.
332
Bibliography
. 2004b. Phylogenetic analysis of extinct and extant pangolins (Mammalia, Pholidota) and related taxa using postcranial data. Journal of Vertebrate Paleontology 24(3, Supplement): 63A.
. 2005. Is Eurotamandua a pangolin? A reconsideration
of its affinities to Xenarthra, Pholidota and Palaeanodonta.
Journal of Vertebrate Paleontology 25(3, Supplement): 62
63A.
Gaudin, T. J., and D. G. Branham. 1998. The phylogeny of the
Myrmecophagidae (Mammalia, Xenarthra, Vermilingua)
and the relationship of Eurotamandua to the Vermilingua.
Journal of Mammalian Evolution 5: 23765.
Gaudin, T. J., and J. R. Wible. 1999. The entotympanic of pangolins and the phylogeny of the Pholidota (Mammalia). Journal of Mammalian Evolution 6: 3965.
. 2006. The phylogeny of living and extinct armadillos
(Mammalia, Xenarthra, Cingulata): a craniodental analysis.
In M. T. Carrano, T. J. Gaudin, R. W. Blob, and J. R. Wible,
eds., Amniote paleobiology: perspectives on the evolution of
mammals, birds and reptiles, 15398. University of Chicago
Press, Chicago.
Gaudin, T. J., et al. 1996. Reexamination of the morphological
evidence for the cohort Epitheria (Mammalia, Eutheria).
Gause, G. E. 1980. Physiological and morphometric responses
of the nine-banded armadillo (Dasypus novemcinctus) to
environmental factors. Ph.D. diss., University of Florida,
Gainsville.
Genoways, H. H., and R. M. Timm. 2003. The xenarthrans of
Nicaragua. Mastozoologa Neotropical 10: 23153.
Gentry, A. H. 1990. Floristic similarities and differences between southern Central America and upper and central
Amazonia. In A. H. Gentry, ed., Four Neotropical rainforests,
14157. Yale University Press, New Haven and London.
Gentry, A. H., and J. Terborgh. 1990. Composition and dynamics of the Cocha Cashu mature floodplain forest. In A. H.
Gentry, ed., Four Neotropical rainforests, 54264. Yale University Press, New Haven and London.
Gezuele, E. 1972. Fatal infection by Nocardia brasiliensis in an
armadillo. Sabouraudia 10: 6365.
Gibbs, R. A., et al. 2004. Genome sequence of the brown Norway rat yields insights into mammalian evolution. Nature
428: 493521.
Gilbert, B. M., and L. Giacometti. 1972. An environmental caging system and maintenance program for the nine-banded
armadillo Dasypus novemcinctus. Laboratory Animal Science 22: 73940.
Gillette, D. D., and C. E. Ray. 1981. Glyptodonts of North America. Smithsonian Contributions to Paleobiology 40: 1255.
Gilmore, D. P., and C. P. da Costa. 1995. The three-toed sloth in
biomedical research: an update on the endocrine and reproductive systems. Medical Science Research 23: 57981.
Gilmore, D. P., C. P. da Costa, and D. P. F. Duarte. 2000. An update on the physiology of two- and three-toed sloths. Brazilian Journal of Medical and Biological Research 33: 12946.
. 2001. Sloth biology: an update on their physiological
ecology, behavior and role as vectors of arthropods and arboviruses. Brazilian Journal of Medical and Biological Research 34: 925.
Gilmore, D. P., et al. 1991. Effects of exogenous LHRH on plasma

LH and sex steroid levels in the three-toed sloth Bradypus
tridactylus. Medical Science Research 19: 33335.
Gilmore, D. P., et al. 1994. Further studies on reproductive function in the three-toed sloth, Bradypus tridactylus. Medical
Science Research 22: 25556.
Ginther O. J. 1989. Twin embryos in mares. II: post fixation embryo reduction. Equine Veterinary Journal 21: 17174.
Glass, B. 1985. History of classification and nomenclature in
Xenarthra (= Edentata). In G. G. Montgomery, ed., The
ecology and evolution of armadillos, sloths, and vermilinguas,
13. Smithsonian Institution Press, Washington, D.C.
Glaz, D., and A. A. Carlini. 1999. Estudio preliminar del home
range y rea mxima de actividad en Chaetophractus vellerosus (Mammalia, Dasypodidae). Abstracts XIV Jornadas
Argentinas de Mastozoologa 14: 19.
Gleeson, S. K., A. B. Clark, and L. A. Dugatkin. 1994. Monozygotic twinning: an evolutionary hypothesis. Proceedings of
the National Academy of Sciences, U.S.A. 91: 1136367.
Goeritz, F., et al. 2003. Seasonal timing of sperm production in
roe deer: interrelationship among changes in ejaculate parameters, morphology and function of testis and accessory
glands. Theriogenology 59: 1487502.
Goffart, M. 1971. Function and form in the sloth. Pergamon
Press, Oxford.
Goldschmidt, B. 1992. Estudo citogentico em preguias (Bradypus variegatus). Masters thesis, Universidade Federal de
Rio de Janeiro, Rio de Janeiro, Brazil.
Goldschmidt, B., and J. C. C. Almeida. 1993. Cytogenetic studies in sloths (Bradypus variegatus). Revista Brasileira de
Gentica 16: 93948.
Goldschmidt, B., J. C. C. Almeida, and V. L. Oliveira. 1995. Visualization of nucleolar organizer regions in sloths (Bradypus variegatus and Scaeopus torquatus). Revista Brasileira
de Gentica 18: 11113.
Gomendio M., and E. R. S. Roldan. 1991. Sperm competition
influences sperm size in mammals. Proceedings of the Royal
Society of London, Series B 243: 18185.
. 1994. The evolution of gametes. In E. E. Bittar and N.
Bittar, eds., Principles of medical biology, 11551. JAI Press,
Greenwich, Conn.
Gonalves, A. M. P. M. 2003. As preguias urbanas de Tefilo
Otoniaspectos citogenticos e ecolgicos. Monograph, Universidade Federal de Minais Gerais, Belo Horizonte, Brazil.
Gonzlez, E. M., A. Soutullo, and C. A. Altuna. 2001. The burrow of Dasypus hybridus (Cingulata: Dasypodidae). Acta
Theriologica 46: 5359.
Gonzlez, J. A. 2003. Patrones generales de caza y pesca en
comunidades nativas y asentamientos de colonos aledaos
a la Reserva Comunal Yanesha, Paso, Per. In R. PolancoOchoa, ed., Manejo de fauna silvestre en Amazona y Latinoamrica: seleccin de trabajos V Congreso Internacional,
89102. Fundacin Natura, Bogot, Colombia.
Gordon, D. H., and C. R. B. Watson. 1986. Identification of
cryptic species of rodents (Mastomys, Saccostomys) in the
Kruger National Park. South African Journal of Zoology 21:
9599.
Gould, S. J. 1996. Can we truly know sloth and rapacity? Natural
History 105: 1822.
Bibliography
Gowaty, P. A. 1996. Field studies of parental care in birds: new

data focus questions on variation among females. Advances
in the Study of Behavior 27: 476531.
Grabowski, N. T., et al. 2002. Comparison of major and minor
milk constituents of a milk sample from the two-toed sloth
(Choloepus didactylus) in early lactation to common rearing
formulae. Der Zoologische Garten 3: 17788.
Grafen, A. 1990. Do animals really recognize kin? Animal Behaviour 39: 4254.
Grass, P. P. 1955. Ordre des Edentes (Edentata Cuvier, 1798;
Edentati Vicq DAzyr, 1792; Bruta Linn, 1758 p.p), formes
actuelles. In P. P. Grass, ed., Trait de zoologie, anatomie,
systematique, biologie, Mammiferes, Tome XVII, Fascicule II,
1182246. Masson et Cie. Editeurs, Libraires de Lacademie
de Medicine, Paris.
Greegor, D. H. 1974. Comparative ecology and distribution of
two species of armadillos, Chaetophractus vellerosus and
Dasypus novemcinctus. Ph.D. diss., University of Arizona,
Tucson.
. 1975. Renal capabilities of an Argentine desert armadillo. Journal of Mammalogy 56: 62632.
. 1980a. Preliminary study of movements and home range
of the armadillo, Chaetophractus vellerosus. Journal of Mammalogy 61: 33435.
. 1980b. Diet of the little hairy armadillo Chaetophractus
vellerosus of northwestern Argentina. Journal of Mammalogy 61: 33134.
. 1985. Ecology of the little hairy armadillo, Chaetophractus vellersosus. In G. G. Montgomery, ed., The evolution and
ecology of armadillos, sloths, and vermilinguas, 397405.
Greenwood, A. D., et al. 2001. A molecular phylogeny of two
extinct sloths. Molecular Phylogenetics and Evolution 18:
94103.
Gregory, W. K. 1910. The orders of mammals. Bulletin of the
American Museum of Natural History 27: 1524.
Griffith, S. C., I. P. F. Owens, and K. A. Thuman. 2002. Extra
pair paternity in birds: a review of interspecific variation
and adaptive function. Molecular Ecology 11: 2195212.
Griffiths-Jones, S., et al. 2005. Rfam: annotating non-coding
RNAs in complete genomes. Nucleic Acids Research 33:
D12124.
Grimaldi, D., and D. Agosti. 2000. A formicine in New Jersey
Cretaceous amber (Hymenoptera: Formicidae) and early
evolution of the ants. Proceedings of the National Academy of
Sciences, U.S.A. 97: 1367883.
Gucwinska, H. 1971. Development of six-banded armadillos at
Wroclaw Zoo. International Zoo Yearbook 11: 8889.
Guerra, M. M., and E. J. Naranjo. 2003. Cacera de subsistencia
en dos localidades de la selva Lacandona, Chiapas, Mxico. In R. Polanco-Ochoa, ed., Manejo de fauna silvestre en
Amazona y Latinoamrica: seleccin de trabajos V Congreso
Internacional, 33944. Fundacin Natura, Bogot, Colombia.
Guimaraes, M. M. 1997. Area de vida, territorialidade, e dieta
do tatu-bola, Tolypeutes tricinctus (Xenarthra, Dasypodidae) num Cerrado do Brasil Central. Masters thesis, Universidade de Brasilia, Brazil.
Guinart S., D. 1997. Los mamferos del bosque semideciduo
333
neotropical de Lomero (Bolivia): interaccin indgena.

Ph.D. diss., Universidad de Barcelona, Spain.
Gupte, M. D., et al. 1998. Comparative leprosy vaccine trial in
South India. Indian Journal of Leprosy 70: 36988.
Guth, C. 1961. La rgion temporale des Edents. Ph.D. diss.,
Universit de Paris, France.
Hamerton, J. L. 1968. Robertsonian translocation in man: evidence for prezygotic selection. Cytogenetics 7: 26076.
Hames, R. B. 1991. Wildlife conservation in tribal societies. In
M. L. Oldfield and J. B. Alcorn, eds., Biodiversity: culture,
conservation and ecodevelopment, 17299. Westview Press,
Boulder, Colo.
Hames, R. B., and W. T. Vickers. 1982. Optimal diet breadth as a
model to explain variability in Amazonian hunting. American Ethnologist 9: 35878.
Hamlett, G. W. D. 1932. The reproductive cycle of the armadillo.
Zeitschrift fr Wissenschaftliche, Zoologie 141: 14357.
. 1933. Polyembryony in the armadillo: genetic or physiological? Quarterly Review of Biology 8: 34858.
. 1935. Delayed implantation and discontinuous de
velopment in mammals. Quarterly Review of Biology 10:
43247.
Handley, C. O. Jr. 1976. Mammals of the Smithsonian Venezuelan project. Brigham Young University Science Bulletin, Biological Series 20: 189.
Hansen, R. M. 1978. Shasta ground sloth food habits, Rampart
Cave, Arizona. Paleobiology 4: 30219.
Harding, H. R., K. Aplin, and C. D. Shorey. 1987. Parsimony
analysis of marsupial sperm structure: a preliminary report.
In H. Mohri, ed., New horizons in sperm cell research, 375
85. Japan Scientific Societies Press, Tokyo.
Hartwig, H., and H. G. Hartwig. 1985. Structural characteristics
of the mammalian spleen indicating storage and release of
red blood cells. Aspects of evolutionary and environmental
demands. Experientia 41: 15963.
Hasiotis, S. T., and R. F. Dubiel. 1995. Termite (Insecta: Isoptera)
nest ichnofossils from the Upper Triassic Chinle Formation,
Petrified Forest National Park, Arizona. Ichnos 4: 11930.
Hausman, L. A. 1936. Further studies of the hair of fossil ground
sloth (Nothrotherium shastense) and of its problematical
ovate bodies. American Journal of Science 31: 22328.
Hayes, T. G. 1968a. Studies of a primitive mammalian spleen,
the opossum (Didelphis virginiana). Journal of Morphology
124: 445450.
. 1968b. Electron and light microscopic observations
of the armadillo Dasypus novemcinctus mexicanus spleen.
Anatomical Record 160: 473.
. 1970. Structure of the ellipsoid sheath in the spleen of
the armadillo Dasypus novemcinctus. A light and electron
microscopic study. Journal of Morphology 132: 20724.
Hayman, D. L. 1990. Marsupial cytogenetics. Australian Journal
of Zoology 37: 33149.
Hayman, D. L., and P. G. Martin, eds. 1974. Animal cytogenetics. Vol. 4, Chordata, Number 4, Mammalia, I. Monotremata
and Marsupiala. Gebrder Borntraeger, Berlin.
Hayssen, V., A. van Tienhoven, and A. van Tienhoven. 1993.
Asdells patterns of mammalian reproduction. Cornell University Press, Ithaca, N.Y.
Heath, E., et al. 1987. Rouleaux formation by spermatozoa in
334
Bibliography
the naked-tail armadillo, Cabassous unicinctus. Journal of

Reproduction and Fertility 79: 15358.
Hebeler-Barbosa, F., M. R. Montenegro, and E. Bagagli. 2003.
Virulence profiles of ten Paracoccidioides brasiliensis isolates
obtained from armadillos (Dasypus novemcinctus). Medical
Mycology 41: 8996.
Hebeler-Barbosa, F., et al. 2003. Comparison of the sequences
of the internal transcribed spacer regions and pbgp43 genes
of Paracoccidioides brasiliensis from patients and armadillos
(Dasypus novemcinctus). Journal of Clinical Microbiology 41:
573537.
Herbig-Sandreuter, A. 1964. Neue Beobachtungen am venezolanischen Dreizehenfaultier. Acta Tropica 21: 97113.
Herbst, L. and K. Redford. 1991. Home-range size and social
spacing among female common long-nosed armadillos
(Dasypus novemcinctus). National Geographic Research and
Exploration 7: 23637.
Hernndez, M. B., et al. 2001. Presencia de polmeros de casenas en mamferos. Mastozoologa Neotropical 8: 1520.
Hernndez Betancourt, S., A. Segovia C., and F. Martnez C.
1998. Datos preliminares de la cacera de subsistencia en el
sur de Yucatn, Mxico. VI Simposio sobre venados de Mxico. Universidad Nacional Autnoma de Mxico, Mexico
City.
Herr, R. A., et al. 2001. Phylogenetic analysis of Lacazia loboi
places this previously uncharacterized pathogen within the
dimorphic Onygenales. Journal of Clinical Microbiology 39:
30914.
Heuser, C. H., and G. B. Wislocki. 1935. Early development of
the sloth (Bradypus griseus) and its similarity to that of man.
Contributions to Embryology, Carnegie Institution of Washington 25: 113.
Hickman, C. 1988. Analysis of form and function in fossils.
American Zoologist 28: 77583.
Hill, K., and J. Padwe. 2000. Sustainability of Ach hunting in
the Mbaracay Reserve, Paraguay. In J. G. Robinson and E.
L. Bennett, eds., Hunting for sustainability in tropical forests, 79105. Columbia University Press, New York.
Hill, K., et al. 1997. Impact of hunting on large vertebrates in
the Mbaracayu Reserve, Paraguay. Conservation Biology 11:
133953.
Hill, R. 2004. Morphological and histological diversity in the
integument of extant and fossil cingulates (Mammalia,
Xenarthra). Journal of Vertebrate Paleontology 24(3, Supplement): 70A.
Hillis, D. M. 1998. Taxonomic sampling, phylogenetic accuracy
and investigator bias. Systematic Biology 47: 38.
Hirschfeld, S. E. 1976. A new fossil anteater (Edentata, Mammalia) from Colombia, S. A., and evolution of the Vermilingua.
Journal of Paleontology 50: 41932.
. 1981. Pliometanastes protistus (Edentata, Megalonychidae) from Knights Ferry, California with discussion of early
Hemphillian megalonychids. PaleoBios 36: 116.
. 1985. Ground sloths from the Friasian La Venta fauna,
with additions to the Pre-Friasian Coyaima fauna of Colombia, South America. University of California Publications in
Geological Sciences 128: 191.
Ho, T.-Y. 1967. Relationship between amino acid contents of
mammalian bone collagen and body temperature as a basis
for estimation of body temperature of prehistoric animals.

Comparative Biochemistry and Physiology 22: 11319.
Hoffman, E. A., and E. L. Ritman. 1985. Effect of body orientation on regional lung expansion in dog and sloth. Journal of
Applied Physiology 59: 48191.
Hoffstetter, R. 1952. Les mammifres plistocnes de la Rpublique de lEquateur. Mmoires de la Socit Gologique de
France 66.
. 1954. Les gravigrades cuirasss du dsadien de Patagonie (note prliminaire). Mammalia 18: 15969.
. 1956. Contribution ltude des Orophodontoidea
gravigrades cuirasss de la Patagonie. Annales de Palontologie 42: 2764.
. 1958. Xenarthra. In J. Piveteau, ed., Trait de palontologie mammifres volution, 535636. Masson et Cie, Paris.
. 1982. Les dents xnarthres, un groupe singulier de la
faune Neotropical. In E. M. Gallitelli, ed., Paleontology, essential of historical geology, 385443. STEM Mucchi, Modena, Italy.
. 1986. High Andean mammalian faunas during the PlioPleistocene. In F. Vuilleumier and M. Monasterio, eds., High
altitude tropical biogeography, 21945. Oxford University
Press, New York.
Hofreiter, M., et al. 2000. A molecular analysis of ground sloth
diet through the last glaciation. Molecular Ecology 9: 1975
84.
Hofreiter, M., et al. 2003. Phylogeny, diet, and habitat of an extinct ground sloth from Cuchillo Cur, Neuqun Province,
southwest Argentina. Quaternary Research 59: 36478.
Holland, B. R., F. Delsuc, and V. Moulton. 2005. Visualizing
conflicting evolutionary hypotheses in large collections of
trees: using consensus networks to study the origins of placentals and hexapods. Systematic Biology 54: 6676.
Holland, B. R., and V. Moulton. 2003. Consensus networks: a
method for visualising incompatibilities in collections of
trees. Lecture Notes in Bioinformatics 2812: 16576.
Holmquist, G., and B. M. Dancis. 1980. A general model of
karyotype evolution. Genetica 52/53: 15163.
Hoogland, J. L. 1995. The black-tailed prairie dog: social life of a
burrowing mammal. University of Chicago Press, Chicago.
Hss, M., et al. 1996. Molecular phylogeny of the extinct ground
sloth Mylodon darwinii. Proceedings of the National Academy of Sciences, U.S.A. 93: 18185.
Howerth, E. W., et al. 1990. Survey for leprosy in nine-banded
armadillos (Dasypus novemcinctus) from the southeastern
United States. Journal of Wildlife Disease 26: 11215.
Hsu T. C. 1965. Chromosomes of two species of anteaters. Mammalian Chromosomes Newsletter 15: 1089.
Hsu T. C., and K. Benirschke. 1969. Tamandua tetradactyla. An
Atlas of Mammalian Chromosomes 3: 1089.
Huang, X., and A. Madan. 1999. CAP3: a DNA sequence assembly program. Genome Research 9: 86877.
Hubbell, S. P., and R. B. Foster. 1990. Structure, dynamics, and
equilibrium status of old-growth forest on Barro Colorado
island. In A. H. Gentry, ed., Four Neotropical rainforests,
522541. Yale University Press, New Haven and London.
Huchon, D., et al. 1999. Armadillos exhibit less genetic polymorphism in North America than in South America: nuclear and mitochondrial data confirm a founder effect in
Bibliography
Dasypus novemcinctus (Xenarthra). Molecular Ecology 8:

174348.
Hudelot, C., et al. 2003. RNA-based phylogenetic methods:
application to mammalian mitochondrial RNA sequences.
Hudson, W. H. 1997. El naturalista en el plata. Ediciones el
Elefante Blanco, Buenos Aires. (reprint of work originally
published in 1890)
Hughes, A. 1977. The topography of vision in mammals of contrasting life style: comparative optics and retinal organisation. In F. Crescitelli, ed., Handbook of sensory physiology,
613756. Springer-Verlag, Berlin.
Hughes, R. N., and J. M. Cancino. 1985. An ecological overview
of cloning in Metazoa. In J. B. C. Jackson, L. W. Buss, and
R. E. Cook, eds., Population biology and evolution of clonal
organisms, 15386. Yale University Press, New Haven.
Humphrey, S. R. 1974. Zoogeography of the nine-banded armadillo (Dasypus novemcinctus) in the United States. BioScience 24: 457462.
Hunter, S. W., T. Fujiwara, and P. J. Brennan. 1982. Structure
and antigenicity of the major specific glycolipid antigen of
Mycobacterium leprae. Journal of Biological Chemistry 257:
1507278.
Hurtado, A. M., et al. 1985. Female subsistence strategies among
Ache hunter-gatherers of eastern Paraguay. Human Ecology
13: 128.
Huson, D. H., and D. Bryant. 2006. Application of phylogenetic networks in evolutionary studies. Molecular Biology and
Evolution 23: 25467.
Inbar, M., and R. T. Mayer. 1999. Spatio-temporal trends in
armadillo diurnal activity and road-kills in central Florida.
Wildlife Society Bulletin 27: 86572.
Innami, S. 1978. The study of the armadillo in Paraguay. In The
armadillo as an experimental model in biomedical research,
8995. Scientific publication 366, Pan American Health Organization, Washington, D.C.
Instituto Brasileiro do Meio Ambiente e dos Recursos Naturais
Renovveis (IBAMA). 2005. <www.ibama.gov.br>. Accessed
30 June, 2005.
International Human Genome Sequencing Consortium. 2001.
Initial sequencing and analysis of the human genome. Nature 409: 860921.
. 2004. Finishing the euchromatic sequence of the human
genome. Nature 431: 93145.
International Species Information System (ISIS). 2005. ISIS
species holdings <www.isis.org>. Accessed 29 June, 2005.
International Union for the Conservation of Nature (IUCN).
1996. IUCN red list of threatened animals. IUCN, Gland,
Switzerland.
. 2001. IUCN red list categories and criteria: version 3.1.
IUCN Species Survival Commission. IUCN, Gland, Switzerland and Cambridge, U.K.
. 2003. Guidelines for application of IUCN red list criteria at regional levels: version 3.0. IUCN Species Survival
Commission. IUCN, Gland, Switzerland.
. 2006. IUCN red list of threatened species <http://www.
iucnredlist.org>. Accessed 5 April 2006.
Izor, R. J. 1985. Sloths and other mammalian prey of the harpy
eagle. In G. G. Montgomery, ed., The evolution and ecology
335
of armadillos, sloths and vermilinguas, 34346. Smithsonian

Jacobs, J. F. 1979. Behavior and space usage patterns of the ninebanded armadillo (Dasypus novemcinctus) in southwestern
Mississippi. Masters thesis, Cornell University, Ithaca, New
York.
Jackson, D. 2003. Evaluating evidence of cultural associations
of Mylodon in the semiarid region of Chile. In L. Miotti, M.
Salemme, and N. Flegenheimer, eds., Where the south winds
blow: ancient evidence of Paleo South Americans, 7781.
Center for the Study of the First Americans, Texas A&M
University, College Station.
Jamieson, B. G. M. 1987. A biological classification of sperm
types, with special reference to annelids and mollusks, and
an example of spermiocladistics. In H. Mohri, ed., New horizons in sperm cell research, 31132. Japan Scientific Societies
Press, Tokyo.
Jamieson, B. G. M., J. Ausi, and J. L. Justine. 1995. Advances in
spermatozoal phylogeny and taxonomy. Mmoires du Musum National dHistoire Naturelle, Tome 166, Editions du
Musum, Paris.
Janis, C. M. 1988. An estimation of tooth volume and hypsodonty indices in Ungulate mammals, and the correlation of
these factors with dietary preference. Mmoires du Musum
National dHistoire Naturelle (Srie C), 53: 36787.
. 1990. Correlation of cranial and dental variables with
body size in ungulates and macropodids. In J. Damuth and
B. J. MacFadden, eds., Body size in mammalian paleobiology:
estimation and biological implications, 22599. Cambridge
. 1995. Correlations between craniodental morphology
and feeding behavior in ungulates: reciprocal illumination
between living and fossil taxa. In J. Thomason, ed., Functional morphology in vertebrate palaeontology, 7698. Cambridge University Press, Cambridge.
Janis, C. M., and D. Ehrhardt. 1988. Correlation of muzzle
width and relative incisor width with dietary preference
in ungulates. Zoological Journal of the Linnean Society 92:
26784.
Janis, C. M., and M. Fortelius. 1988. On the means whereby
mammals achieve increased functional durability of their
dentitions, with special reference to limiting factors. Biological Reviews 63: 197230.
Jeffreys, A. J., V. Wilson, and S. L. Thein. 1985. Individual-specific fingerprints of human DNA. Nature 316: 7679.
Jenkins, F. A. Jr. 1970. Anatomy and function of expanded ribs
in certain edentates and primates. Journal of Mammalogy
51: 288301.
Jeuniaux, C. 1962. Research on polysaccharides in the stomach
of the sloth Choloepus hoffmanni Pet. (Mammalia, Edontata). Archives Internationales de Physiologie et de Biochimie
70: 4078.
Job, C. K., W. F. Kirchheimer, and R. M. Sanchez. 1983. Variable
lepromin response to Mycobacterium leprae in resistant armadillos. International Journal of Leprosy 51: 34753.
Job, C. K., R. M. Sanchez, and R. C. Hastings. 1985. Manifestations of experimental leprosy in the armadillo Dasypus
novemcinctus. American Journal of Tropical Medicine and
Hygiene 34: 15161.
336
Bibliography
Job, C. K., et al. 1984. Attempts to breed the nine-banded armadillo (Dasypus novemcinctus) in captivitya preliminary
report. International Journal of Leprosy 52: 36264.
Job, C. K., et al. 1986. A random survey of leprosy in wild ninebanded armadillos in Louisiana USA. International Journal
of Leprosy 54: 453457.
Job, C. K., et al. 1990. Early infection with M. leprae and antibodies to phenolic glycolipid-I in the nine-banded armadillo. Indian Journal of Leprosy 62: 193201.
Job, C. K., et al. 1991. Comparison of PCR techniques with other
methods for detection of M. leprae in tissues of wild armadillos. Leprosy Review 62: 36273.
Job, C. K., et al. 1992. The pathogenesis of leprosy in the ninebanded armadillo and the significance of IgM antibodies to
PGL-1. Indian Journal of Leprosy 64: 13751.
Job, C. K., et al. 1993. Armadillos (Dasypus novemcinctus) as
a model to test antileprosy vaccines: a preliminary report.
International Journal of Leprosy 61: 39497.
Johansen, K. 1961. Temperature regulation in the nine-banded
armadillo (Dasypus novemcinctus). Physiological Zoology
34: 12644.
Johnston, P. G., et al. 2002. Sex chromosome elimination and
reactivation in the southern brown bandicoot Isoodon obeselus (Marsupialia: Peramelidae). Cytogenetic and Genome
Research 99: 11924.
Jones-Rhoades, M. W., and D. P. Bartel. 2004. Computational
identification of plant microRNAs and their targets, including a stress-induced miRNA. Molecular Cell 14: 78799.
Jorge, W. 1982. Chromosome study in the Pichiciego Argentino, Chlamyphorus truncatus Harlan (Dasypodidae, Edentata). Mammalian Chromosomes Newsletter 23: 1017.
. 2000. Mitotic and meiotic chromosome studies in the
silky anteater Cyclopes didactylus (Myrmecophagidae: Xenarthra). Cytobios 101: 95100.
Jorge, W., R. C. Best, and R. M. Wetzel. 1985b. Chromosome studies on the silky anteater Cyclopes didactylus L.
(Myrmecophagidae: Xenarthra, Edentata). Caryologia 38:
32529.
Jorge, W., D. A. Meritt, and K. Benirschke. 1977. Chromosome
studies in Edentata. Cytobios 18: 15772.
Jorge, W., A. T. Orsi-Souza, and R. C. Best. 1985a. The somatic
chromosomes of Xenarthra. In G. G. Montgomery, ed., The
evolution and ecology of armadillos, sloths, and vermilinguas,
Jorge, W., and L. Pinder. 1990. Chromosome study on the
maned sloth Bradypus torquatus (Bradypodidae, Xenarthra). Cytobios 62: 2125.
Joseph, B. Z., L. J. Yoder, and R. R. Jacobson. 1985. Hansens
disease in native-born citizens of the United States. Public
Health Reports 100: 66671.
Jull, A. J. T., et al. 2004. Radiocarbon dating of extinct fauna in
the Americas recovered from tar pits. Nuclear Instruments
and Methods in Physics Research B 223/224: 66871.
Kalmbach, E. R. 1943. The armadillo: its relation to agriculture
and game. Texas Game, Fish and Oyster Commission, Austin.
Kalthoff, D. 2004. Dental microstructures in fossil and Recent
Xenarthra (Mammalia). Journal of Vertebrate Paleontology
24(3, Supplement): 77A.
Kaplan, W., J. R. Broderson, and J. N. Pacific. 1982. Spontaneous
systemic sporotrichosis in nine-banded armadillos (Dasypus novemcinctus). Sabouraudia 20: 28994.

Kasuga, T., J. W. Taylor, and T. J. White. 1999. Phylogenetic relationships of varieties and geographical groups of the human
pathogenic fungus Histoplasma capsulatum Darling. Journal
of Clinical Microbiology 37: 65363.
Kasuga, T., et al. 2003. Phylogeography of the fungal pathogen
Histoplasma capsulatum. Molecular Ecology 12: 3383401.
Kaszubiak, A., et al. 2004. Population structure and evolutionary origins of Microsporum canis, M. ferrugineum and M.
auouinii. Infection, Genetics and Evolution 4: 17986.
Kaufmann, P., M. Luckhardt, and W. Elger. 1985. The structure
of the tupaia placenta. II. Ultrastructure. Anatomy and Embryology 171: 21121.
Kay, R. F., and R. H. Madden. 1997. Mammals and rainfall: paleoecology of the middle Miocene at La Venta (Colombia,
South America). Journal of Human Evolution 32: 16199.
Keller, L. 1997. Indiscriminant altruism. Trends in Ecology and
Evolution 12: 99103.
Kellis, M., et al. 2003. Sequencing and comparison of yeast species to identify genes and regulatory elements. Nature 423:
24154.
Kendrick, B. 2000. The fifth kingdom. 3rd ed. Focus Publishing,
Newburyport, Mass.
Kielan-Jaworowska, Z., R. L. Cifelli, and Z-X. Luo. 2004. Mammals from the age of dinosaurs: origins, evolution, and structure. Columbia University Press, New York.
King, B. F., P. B. N. Pinheiro, and R. L. Hunter. 1982. The fine
structure of the placental labyrinth in the sloth, Bradypus
tridactylus. Anatomical Record 202: 1522.
Kinlaw, A. E. 2006. Burrow dispersion of central Florida armadillos. Southeastern Naturalist 5: 52334.
Kirchheimer, W. F. 1977. Occurrence of Mycobacterium leprae
in nature. Leprosy in India 49: 4447.
. 1979. Examination of North American armadillos for
mycobacteriosisa further report. Leprosy in India 51:
6064.
. 1982. Experimental leprosy: its present status. Leprosy
in India 54: 53035.
Kirchheimer, W. F., and R. M. Sanchez. 1978a. Examination of
North American armadillos for mycobacteriosis. Leprosy in
India 50: 15660.
. 1978b. Leprosy in the wild. Leprosy Scientific Memoranda L-966: 5.
Kirchheimer, W. F., and E. E. Storrs. 1971. Attempts to establish
the armadillo (Dasypus novemcinctus Linn.) as a model for
the study of leprosy. I. Report of lepromatoid leprosy in an
experimentally infected armadillo. International Journal of
Leprosy 39: 693702.
Kishino, H., J. L. Thorne, and W. J. Bruno. 2001. Performance of
a divergence time estimation method under a probabilistic
model of rate evolution. Molecular Biology and Evolution 18:
35261.
Kleiber, M. 1932. Body size and metabolism. Hilgardia 6:
31553.
Kleiman, D. G. 1977. Monogamy in mammals. Quarterly Review of Biology 52: 3969.
Klir, J. R., J. E. Heath, and N. Bennani. 1990. An infrared thermographic study of surface temperature in relation to external thermal stress in the Mongolian gerbil, Meriones un-
Bibliography
guiculatus. Comparative Biochemistry and Physiology 96A:

14146.
Kniffen, F. B., and S. B. Hillard. 1988. Louisiana: its land and
people. Louisiana State University Press, Baton Rouge.
Kopp-Valdivia, D. M. 2005. Abundancia de mamferos de uso
alimenticio en zonas de caza de Tsimane y zonas sin caza en
la Reserva de la Biosfera del Beni (Beni, Bolivia). I Congreso
Nacional de Mastozoologa en Bolivia, Cochabamba, Bolivia,
31 May3 June 2005.
Kovacs, A., et al. 2001. 2n = 65 karyotype in a two-toed sloth
(Choloepus didactylus L.). Annales de Gntic 44: 43.
Kraglievich, L. 1923. Descripcin de dos crneos y otros restos
del gnero Pliomorphus Ameghino procedentes de la Formacin Entrerriana de las barrancas del Rio Paran. Anales
del Museo Nacional de Historia Natural de Buenos Aires 33:
156.
. 1934. La antigedad Pliocena de las faunas de Monte
Hermoso y Chapadmalal deducidas de su comparacin con
las que le precedieron y sucedieron. El Siglo Ilustrado 398:
17136.
. 1940a. Los Megaterios miocenos y pliocenos de la Argentina. In A. J. Torcelli and C. A. Marelli, eds., Obras de
geologa y paleontologa. Vol. 3, 66989. Ministerio de Obras
Pblicas de la Provincia de Buenos Aires, La Plata.
. 1940b. Dos nuevas especies fsiles de oso hormigueros.
In A. J. Torcelli and C. A. Marelli, eds., Obras de geologa y
paleontologa. Vol. 3, 63136. Ministerio de Obras Pblicas
de la Provincia de Buenos Aires, La Plata.
Kraglievich, L., and S. Rivas. 1951. Orophodon Amegh., representante de una nueva superfamilia Orophodontoidea del
suborden Xenarthra. Comunicaciones del Instituto Nacional
de Investigacin de las Ciencias Naturales (Ciencias Zoolgicas) 2: 5582.
Krebs, J. R., and N. B. Davies. 1984. Behavioural ecology: an evolutionary approach. 2nd ed. Blackwell Scientific, London.
Kredel, F. E. 1928. Note on the temperature of the sloth. Journal
Kriegs, J. O., et al. 2006. Retroposed elements as archives for
the evolutionary history of placental mammals. PLoS Biology 4: e91.
Krishna, K., and D. A. Grimaldi. 2003. The first Cretaceous Rhinotermitidae (Isoptera): a new species, genus, and subfamily
in Burmese amber. American Museum Novitates 3390: 110.
Krogh, A. 1929. The progress of physiology. American Journal of
Physiology 90: 24351.
Krotoski, W. A., et al. 1984. Enzootic schistosomiasis in a Louisiana USA armadillo Dasypus novemcinctus. American Journal of Tropical Medicine and Hygiene 33: 26972.
Krumbiegel, I. 1940. Die Sugetiere der SdamerikaExpeditionen Prof. Dr. Kriegs. Zoologischer Anzeiger 131: 4973.
Labhsetwar, A. P., and A. C. Enders. 1968. Progesterone in
the corpus luteum and placenta of the armadillo, Dasy-
pus novemcinctus. Journal of Reproduction and Fertility 16:
38187.
Lacaz, C. S. 1994. Historical evolution of the knowledge on
paracoccidioidomycosis and its etiologic agent, Paracoccidioides brasiliensis. In M. F. Franco, C. S. Lacaz, A. Restrepo,
and G. Del Negro, eds., Paracoccidioidomycosis, 111. CRC
Press, Boca Raton, Fla.
Lainson, R., and J. J. Shaw. 1989. Leishmania (Viannia) naiffi sp.
337
n., a parasite of the armadillo, Dasypus novemcinctus (L.)

in Amazonian Brazil. Annales de Parasitologie Humaine et
Comparee 64: 39.
Lainson, R., J. J. Shaw, and M. Povoa. 1981. The importance of
edentates (sloths, anteaters) as primary reservoirs of Leishmania brasiliensis guyanensis. Transactions of the Royal Society of Tropical Medicine and Hygiene 75: 61112.
Lainson, R., et al. 1979. Leishmaniasis in Brazil: isolation of
Leishmania from armadillos (Dasypus novemcinctus) and
observation on the epidemiology of cutaneous leishmaniasis in north Para state. Transactions of the Royal Society of
Tropical Medicine and Hygiene 73: 23942.
Lander, E. S., and M. S. Waterman. 1988. Genomic mapping
by fingerprinting random clones: a mathematical analysis.
Genomics 2: 23139.
Lara-Ruiz, P. 2004. Tamanho corporal, dimorfismo sexual
e diversidade gentica da preguia-de-coleira, Bradypus
torquatus Illiger, 1811 (Xenarthra: Bradypodidae). Masters
thesis, Pontifica Universidade Catlica de Minas Gerais,
Brazil.
Lara-Ruiz, P., and A. G. Chiarello. 2005. Life-history traits and
sexual dimorphism of the Atlantic forest maned sloth Bradypus torquatus (Xenarthra: Bradypodidae). Journal of Zoology 267: 6373.
Laudermilk, J. D., and P. A. Munz. 1934. Plants in the dung of
Nothrotherium from Gypsum Cave, Nevada. Carnegie Institution of Washington 483: 2938.
. 1938. Plants in the dung of Nothrotherium from Rampart
and Muav Caves, Arizona. Carnegie Institution of Washington 487: 27181.
Layne, J. N., and D. Glover. 1977. Home range of the armadillo
in Florida. Journal of Mammalogy 58: 41113.
. 1978. Activity cycles of the nine-banded armadillo in
southern Florida. In R. Obrtel, C. Folk and J. Pellantova,
eds., Second international theriological congress, 140. Institute of Vertebrate Zoology, Brno, Czechoslovakia.
. 1985. Activity patterns of the common long-nosed armadillo Dasypus novemcinctus in south-central Florida. In
G. G. Montgomery, ed., The evolution and ecology of armadillos, sloths and vermilinguas, 40717. Smithsonian Institution Press, Washington, D.C.
Layne, J. N., and A. M. Waggener. 1984. Above-ground nests of
the nine-banded armadillo in Forida. Florida Field Naturalist 12: 5861.
Leclerc, M. C., H. Philippe, and E. Gueho. 1994. Phylogeny of
dermatophytes and dimorphic fungi based on large subunit
ribosomal RNA sequence comparisons. Journal of Medical
and Veterinary Mycology 32: 33141.
Leeuwenberg, F. 1994. Anlise etno-zoolgica e manejo da fauna
cinegtica na Reserva Indgena Xavante Rio das Mortes, aldeia Tenitipa, Mato Grosso, Brasil. Centro de Pesquisa Indgena, Brazil.
. 1997. Edentata as a food resource: subsistence hunting
by Xavante Indians, Brazil. Edentata 3: 45.
Leeuwenberg, F., and J. G. Robinson. 2000. Traditional management of hunting in a Xavante community in central Brazil: the search for sustainability. In J. G. Robinson and E. L.
Bennett, eds., Hunting for sustainability in tropical forests,
37594. Columbia University Press, New York.
Leffingwell, L. M., and S. U. Neill. 1989. Naturally acquired ra-
338
Bibliography
bies in an armadillo (Dasypus novemcinctus) in Texas. Journal of Clinical Microbiology 27: 17475.
Leigh, E. G., Jr., and S. J. Wright. 1990. Barro Colorado island
and tropical biology. In A. H. Gentry, ed., Four Neotropical
rainforests, 2847. Yale University Press, New Haven and
London.
Leite, K. C. E., et al. 2004. Estrutura de acasalamento e parentesco em uma populao de tamandu-bandeira (Myrmecophaga tridactyla, Myrmecophagidae) no Parque Nacional de Emas. In Annals of the 50th Congresso Brasileiro
de Gentica, 135. (available on CD ROM)
Leite, M. 1990. Stratigraphy and mammalian paleontology of
the Ash Hollow Formation (Upper Miocene) on the north
shore of Lake McConaughy, Keith County, Nebraska. University of Wyoming Contributions to Geology 28: 129.
Lessa, E. P., and R. A. Faria. 1996. Reassessment of extinction patterns among the late Pleistocene mammals of South
America. Palaeontology 39: 651662.
Lessa, E. P., B. Van Valkenburgh, and R. A. Faria. 1997. Testing
hypotheses of differential mammalian extinctions subsequent to the Great American Biotic Interchange. Palaeogeography, Palaeoclimatology and Palaeoecology 135: 15762.
Li, X., et al. 2003. Nuclear transfer: progress and quandaries.
Reproductive Biology and Endocrinology 1: 8489.
Lin, Y.-H., et al. 2002. Four new mitochondrial genomes and
the increased stability of evolutionary trees of mammals
from improved taxon sampling. Molecular Biology and Evolution 19: 206070.
Linares, O. J. 1998. Mamferos de Venezuela. Sociedad Conservacionista Audubon de Venezuela, Caracas.
. 2004. Bioestratigrafa de la fauna de mamferos de las
formaciones Socorro, Urumaco y Codore (Mioceno medioPlioceno temprano) de la regin de Urumaco, Falcn, Venezuela. Paleobiologa Neotropical 1: 126.
Lindblad-Toh, K., et al. 2005. Genome sequence, comparative
analysis and haplotype structure of the domestic dog. Nature 438: 80319.
Linderman, C. B., and K. S. Kanous. 1997. A model for flagellar
motility. International Review of Cytology 173: 172.
Lindsay, D. S., et al. 1996. Prevalence and identity of Sarcocystis
infections in armadillos (Dasypus novemcinctus). Journal of
Parasitology 82: 51820.
Linnaeus, C. 1758. Systema naturae per regna tria naturae, secoundum classes, ordines, genera, species, cum characteribus,
differentiis, synonymis, locis. Vol. 1, Regnum animale. Editio
decima, reformata. Stockholm.
Lins, Z. C., and M. L. Lopes. 1984. Isolation of Leptospira from
wild forest animals in Amazonian Brazil. Transactions of the
Royal Society of Tropical Medicine and Hygiene 78: 12426.
Lippert, K. J. 1994. Food habits, distribution and impact of the
nine-banded armadillo in Missouri. Masters thesis, Southwest Missouri State University, Springfield.
Liu, C., et al. 2005. NONCODE: an integrated knowledge
database of non-coding RNAs. Nucleic Acids Research 33:
D11215.
Lpez, P. M., and D. S. Jackson. 2004. Ground sloth predation
in the northern semiarid region of Chile. Current Research
in the Pleistocene 21: 1416.
Lpez Piero, J. M., and T. F. Glick. 1993. El megaterio de Bru y el
Presidente Jefferson. Una relacin insospechada en los albores
de la paleontologa. Cuadernos Valencianos de Historia de la

Medicina y de la Ciencia XLII, Serie A. Artes Grficas Soiler,
Valencia, Spain.
Lott, D. F. 1991. Intraspecific variation in the social systems of
wild vertebrates. Cambridge University Press, New York.
Loughry, W. J., and C. M. McDonough. 1994. Scent discrimination by infant nine-banded armadillos. Journal of Mammalogy 75: 103339.
. 1996. Are road kills valid indicators of armadillo population structure? American Midland Naturalist 135: 5359.
. 1998a. Comparisons between nine-banded armadillo
(Dasypus novemcinctus) populations in Brazil and the United States. Revista de Biologa Tropical 46: 117383.
. 1998b. Spatial patterns in a population of nine-banded
armadillos (Dasypus novemcinctus). American Midland
Naturalist 140: 16169.
. 2001. Natal recruitment and adult retention in a population of nine-banded armadillos. Acta Theriologica 46:
393406.
. 2002. Phenotypic variability within and between litters of nine-banded armadillos. Southeastern Naturalist 1:
28798.
Loughry, W. J., G. M. Dwyer, and C. M. McDonough. 1998.
Behavioral interactions between juvenile nine-banded armadillos in staged encounters. American Midland Naturalist
139: 12532.
Loughry, W. J., C. M. McDonough, and E. G. Robertson. 2002.
Patterns of anatomical damage in a population of ninebanded armadillos Dasypus novemcinctus (Xenarthra,
Dasypodidae). Mammalia 66: 11122.
Loughry, W. J., P. A. Prodhl, and C. M. McDonough. 2005.
The inadequacy of observation: understanding armadillo
biology with molecular markers. In S. G. Pandalai, ed., Recent research developments in ecology, 5573. Transworld
Research Network, Kerala, India.
Loughry, W. J., P. A. Prodhl, C. M. McDonough, and J. C.
Avise. 1998. Polyembryony in armadillos. American Scientist 86: 27479.
Loughry, W. J., P. A. Prodhl, C. M. McDonough, W. S. Nelson,
and J. C. Avise. 1998. Correlates of reproductive success in a
population of nine-banded armadillos. Canadian Journal of
Zoology 76: 181521.
Lovegrove, B. G. 2001. The evolution of body armor in mammals: plantigrade constraints of large body size. Evolution
55: 146473.
Lovegrove, B. G., G. Heldmaier, and T. Ruf. 1991. Perspectives in
endothermy revisited: the endothermic temperature range.
Journal of Thermal Biology 16: 18597.
Lubin, Y. D. 1983. Eating ants is no picnic. Natural History 83:
5459.
Lubin, Y. D., and G. G. Montgomery. 1981. Defenses of Nasutitermes termites (Isoptera, Termitidae) against Tamandua
anteaters (Edentata, Myrmecophagidae). Biotropica 13:
6676.
Luckett, W. P. 1968. Morphogenesis of the placenta and fetal
membranes of the tree shrews (Family Tupaiidae). American Journal of Anatomy 123: 385428.
Luederwaldt, H. 1918. Observaes sobre a preguia (Bradypus
tridactylus, L.) em liberdade e no captiveiro. Revista do Museu Paulista 10: 795812.
Bibliography
Lumpkin, L. R. III, G. F. Cox, and J. E. Wolf, Jr. 1983. Leprosy in

five armadillo handlers. Journal of the American Academy of
Dermatology 9: 899903.
. 1984. Leprosy in armadillo handlers. Journal of the
American Academy of Dermatology 10: 1073.
Lund, P. W. 1839. Coup doeil sur les espces eteintes de mammifres du Brsil: extrait de quelques mmoires presents
lAcadmies royal des Sciences deCopenhague. Annales des
Sciences Naturelles, Zoologie 2 11: 21434.
Lundy, W. E. 1952. The upside-down animal. Natural History
61: 11419.
Luo, Z.-X., and J. R. Wible. 2005. A Late Jurassic digging mammal and early mammalian diversification. Science 308:
1037.
Lydekker, R. 1894. Contributions to knowledge of the fossil
vertebrates of Argentina, Part II: The extinct edentates of
Argentina. Anales del Museo de La Plata (Paleontologa Argentina) 3: 1118.
Lynch, M. 1984. Destabilizing hybridization, general-purpose
genotypes and geographic parthenogenesis. Quarterly Review of Biology 59: 25790.
Ma, B., J. Tromp, and M. Li. 2002. PatternHunter: faster and
more sensitive homology search. Bioinformatics 18: 440
45.
Mace, G., and A. Balmford. 2000. Patterns and processes in
contemporary mammalian extinction. In A. Entwistle and
N. Dunstone, eds., Priorities for the conservation of mammalian diversity: has the panda had its day? 2752. Cambridge
Macedo, R. C. L., et al. 1999. Paracoccidioides brasiliensis
Infeco natural em tatus. Estudo em Serra da Mesa, Gois,
Brasil. Proceedings VII International Meeting on Paraco
ccidioidomycosis. Campos do Jordo, So Paulo, Brasil
D-06.
Machado, A. B. M., et al. 1998. Livro vermelho das espcies
ameaadas de extino de fauna de Minas Gerais. Fundao
Biodiversitas, Belo Horizonte.
Mackinnon, J. E., et al. 1969. Isolation of Sporothrix schenckii
from nature and considerations on its pathogenicity and
ecology. Sabouraudia 7: 3845.
MacPhee, R. D. E., and M. A. Iturralde-Vinent. 1994. First Tertiary land mammal from Great Antilles: an early Miocene
sloth (Xenarthra, Megalonychidae) from Cuba. American
Museum Novitates 3094: 113.
. 1995. Origin of the Greater Antillean land mammal
fauna 1: new Tertiary fossils from Cuba and Puerto Rico.
American Museum Novitates 3141: 130.
MacPhee, R. D. E., C. Flemming, and D. P. Lunde. 1999. Last
occurrence of the Antillean insectivore Nesophontes: new
radiometric dates and their interpretation. American Museum Novitates 3261: 119.
MacPhee, R. D. E., M. A. Iturralde-Vinent, and E. S. Gaffney.
2003. Domo de Zaza, an early Miocene vertebrate locality
in south-central Cuba, with notes on the tectonic evolution
of Puerto Rico and the Mona Passage. American Museum
Novitates 3394: 142.
MacPhee, R. D. E., J. L. White, and C. A. Woods. 2000. New
megalonychid sloths (Phyllophaga, Xenarthra) from the
Quaternary of Hispaniola. American Museum Novitates
3303: 132.
339
Madsen, O., et al. 2001. Parallel adaptive radiations in two major clades of placental mammals. Nature 409: 61014.
Maffei, L., E. Cullar, and A. Noss. 2004. One thousand jaguars
(Panthera onca) in Bolivias Chaco? Camera trapping in the
Kaaa-Iya National Park. Journal of Zoology 262: 295304.
Maffei, M. D. 1997. Management in National Wildlife Refuges. In D. Simberloff, D. C. Schmitz, and T. C. Brown, eds.,
Strangers in paradise: impact and management of nonindigenous species in Florida, 26774. Island Press, Washington,
D.C.
Maldonado, E. N., and E. B. Casanave. 1996. Electrocardiography in Chaetophractus villosus (Mammalia, Dasypodidae).
Revista de Cincias Biomdicas 17: 1722.
Maldonado, E. N., E. B. Casanave, and M. I. Aveldao. 2002.
Major plasma lipids and fatty acids in four HDL mammals.
Comparative Biochemistry and Physiology A, Molecular and
Integrative Physiology 132: 297303.
Mapelli, F. J., and M. C. Grierson. 2003. Estudio preliminar sobre uso de hbitat de Chaetophractus villosus (Mammalia,
Dasypodidae) en un agroecosistema del sudoeste bonaerense. Abstracts XVIII Jornadas de Mastozoologa 18: 61.
March, M. I. J. 1987. Los Lacandones de Mxico y su relacin
con los mamferos silvestres: un estudio etnozoolgico. Biotica 12: 4356.
Marennikova, S. S., et al. 1977. Outbreak of pox disease among
Carnivora (Felidae) and Edentata. Journal of Infectious Diseases 135: 35866.
Margulies, E. H., et al. 2005. An initial strategy for the systematic
identification of functional elements in the human genome
by low-redundancy comparative sequencing. Proceedings of
Marineros, L., and F. M. Gallegos. 1998. Gua de campo de los
mamferos de Honduras. Instituto Nacional de Ambiente y
Desarollo, Tegucigalpa.
Marinescu, V. D., I. S. Kohane, and A. Riva. 2005. The MAPPER
database: a multi-genome catalog of putative transcription
factor binding sites. Nucleic Acids Research 33: D9197.
Marshall, J. E. 1987. Effects of nest predation on hatching success
in gopher tortoises (Gopherus polyphemus Daudin 1802).
Masters thesis, University of South Alabama, Mobile.
Marshall, L. G., and T. Sempere. 1993. Evolution of the Neotropical land mammal fauna in its geochronologic, stratigraphic and tectonic context. In P. Goldblatt, ed., Biological
relationships between Africa and South America, 32992.
Yale University Press, New Haven.
Marshall, L. G., et al. 1984. Mammals and stratigraphy: geochronology of the continental mammal-bearing Quaternary of
South America. Paleovertebrata, Mmoire Extraordinaire,
Laboratoire de palontologie des vertbrs, Montpellier,
France.
Martin, P. S., B. E. Sabels, and D. Shutler, Jr. 1961. Rampart Cave
coprolite and ecology of the Shasta ground sloth. American
Journal of Science 259: 10227.
Martinez, A. R., et al. 1984. Lepra salvaje en Dasypus novemcinctus (Linneo 1758). Archivos Argentinus Dermatologa 34:
2130.
Massoia, E. 1970. Mamferos que contribuyen a deteriorar suelos y pasturas en la Repblica Argentina. IDIA December:
1417.
Matthew, W. D., and C. de Paula Couto. 1959. The Cuban eden-
340
Bibliography
tates. Bulletin of the American Museum of Natural History

117: 156.
McAdam, D. W., and J. S. Way. 1967. Olfatory discrimination in
the giant anteater. Nature 214: 31617.
McBee, K., and R. J. Baker. 1982. Dasypus novemcinctus. Mammalian Species 162: 19.
McCrane, M. P. 1966. Birth, behavior and development of a
hand reared two toed sloth, Choloepus didactylus. International Zoo Yearbook 6: 15363.
McDonald, D. R. 1977. Food taboos: a primitive environmental
protection agency (South America). Anthropos 72: 73448.
McDonald, H. G. 1977. Description of the osteology of the extinct gravigrade edentate Megalonyx with observations on
its ontogeny, phylogeny and functional anatomy. Masters
thesis, University of Florida, Gainesville.
. 1988. A systematic review of the Plio-Pleistocene Scelidotherine ground sloths (Mammalia: Xenarthra: Mylodontidae). Ph.D. diss., University of Toronto, Ontario.
. 1995. Gravigrade xenarthrans from the Early Pleistocene
Leisey shell pit 1A, Hillsborough County, Florida. Bulletin of
the Florida Museum of Natural History 37: 34573.
. 1996. Biogeography and paleoecology of ground sloths
in California, Arizona and Nevada. San Bernardino County
Museum Association Quarterly 43: 6165.
. 2003a. Sloth remains from North American caves and
associated karst features. In B. W. Schubert, J. I. Mead, and
R. W. Graham, eds., Ice Age cave faunas of North America,
116. Indiana University Press and Denver Museum of Nature and Science.
. 2003b. Xenarthran skeletal anatomy: primitive or derived? (Mammalia, Xenarthra). Senckenbergiana biologica
83: 517.
McDonald, H. G., and C. de Muizon. 1995. An aquatic sloth
from the Pliocene of Peru: another group of mammals returns to water. Journal of Vertebrate Paleontology 15, supplement to number 3: 42.
. 2002. The cranial anatomy of Thalassocnus (Xenarthra,
Mammalia), a derived nothrothere from the Neogene of the
Pisco Formation (Peru). Journal of Vertebrate Paleontology
22: 34965.
McDonald, H. G., C. R. Harington, and G. De Iuliis. 2000. The
ground sloth, Megalonyx, from Pleistocene deposits of the
Old Crow Basin, Yukon, Canada. Arctic 53: 21320.
McDonald, H. G., and D. Perea. 2002. The large scelidothere
Catonyx tarijensis (Xenarthra, Mylodontidae) from the
Pleistocene of Uruguay. Journal of Vertebrate Paleontology
22: 67783.
McDonough, C. M. 1994. Determinants of aggression in ninebanded armadillos. Journal of Mammalogy 75: 18998.
. 1997. Pairing behavior of the nine-banded armadillo
(Dasypus novemcinctus). American Midland Naturalist 138:
29098.
. 2000. Social organization of nine-banded armadillos
(Dasypus novemcinctus) in a riparian habitat. American
Midland Naturalist 144: 13951.
McDonough, C. M., and W. J. Loughry. 1995. Influences on
vigilance in nine-banded armadillos. Ethology 100: 5060.
. 1997a. Patterns of mortality in a population of ninebanded armadillos, Dasypus novemcinctus. American Midland Naturalist 138: 299305.
. 1997b. Influences on activity patterns in a population of

nine-banded armadillos. Journal of Mammalogy 78: 93241.
. 2005. Impacts of land management practices on a population of nine-banded armadillos in northern Florida. Wildlife Society Bulletin 33: 1198209.
McDonough, C. M., et al. 2000. Burrow characteristics and habitat associations of armadillos in Brazil and the United States
of America. Revista de Biologa Tropical 48: 10920.
McKenna, M. C. 1975. Toward a phylogenetic classification of the
Mammalia. In W. P. Luckett and P. S. Szalay, eds., Phylogeny of
the primates, 2146. Plenum, New York.
McKenna, M. C., and S. K. Bell. 1997. Classification of mammals
above the species level. Columbia University Press, New York.
McNab, B. K. 1978. Energetics of arboreal folivores: physiological problems and ecological consequences of feeding on an
ubiquitous food supply. In G. Montgomery, ed., The ecology
of arboreal folivores, 15362. Smithsonian Institution Press,
Washington, D. C.
. 1979. The influence of body size on the energetics and
distribution of fossorial and burrowing mammals. Ecology
60: 101021.
. 1980. Energetics and the limits to a temperate distribution
in armadillos. Journal of Mammalogy 61: 60627.
. 1982. The physiological ecology of South American marsupials. In M. A. Mares and H. H. Genoways, eds., Mammalian biology in South America, 187207. Special Publication
Series, Volume 6, Pymatuning Laboratory of Ecology, University of Pittsburgh, Pa.
. 1984. Physiological convergence amongst ant-eating and
termite-eating mammals. Journal of Zoology 203: 485510.
. 1985. Energetics, population biology, and distribution of
xenarthrans, living and extinct. In G. G. Montgomery, ed., The
evolution and ecology of armadillos, sloths and vermilinguas,
McNab, R. B. 2000. Wildlife management planUaxactn Integrated Forestry Concession. Wildlife Conservation Society,
Flores, Petn, Guatemala.
Medinaceli V., A. 2004. Recursos animales utilizados en la medicina tradicional en dos comunidades Tsimane que habitan en serranas (Alto Colorado) y llanos (Asuncin del
Quiquibey) de la RB-TCO Piln Lajas, La Paz/Beni, Bolivia.
Undergraduate thesis. Universidad Mayor de San Andrs, La
Paz, Bolivia.
Medri, . M. 2002. rea de vida e uso de hbitat de tamandubandeiraMyrmecophaga tridactyla Linnaeus, 1758nas Fazendas Nhumirim e Porto Alegre, Pantanal da Nhecolndia,
MS. Masters thesis, Universidade Federal de Mato Grosso do
Sul, Brazil.
Medri, . M., and G. Mouro. 2005a. Home range of giant anteaters (Myrmecophaga tridactyla) in the Pantanal wetland, Brazil. Journal of Zoology 266: 36575.
. 2005b. A brief note on the sleeping habits of giant anteatersMyrmecophaga tridactyla Linnaeus (Xenarthra, Myrmecophagidae). Revista Brasileira de Zoologia 22: 121315.
Medri, . M., G. Mouro, and A. Y. Harada. 2003. Dieta de tamandu-bandeira (Myrmecophaga tridactyla) no Pantanal da
Nhecolndia, Brasil. Edentata 5: 2934.
Mena Valenzuela, P., and R. Cueva. 2001. Cacera de subsistencia en tres comunidades de la zona de amortiguamiento del
Parque Nacional Yasun. In J. P. Jorgenson and M. C. Rodr-
Bibliography
guez, eds., Conservacin y desarrollo sostenible del Parque

Nacional Yasun y su rea de influencia: Memorias del seminario-taller 2001, 10127. WCS, Ministerio del Ambiente,
UNESCO, Quito, Ecuador.
Mena Valenzuela, P., et al. 2000. The sustainability of current
hunting practices by the Huaorani. In J. G. Robinson and E.
L. Bennett, eds., Hunting for sustainability in tropical forests,
Mendel, F. C. 1985a. Adaptations for suspensory behavior in
the limbs of two-toed sloths. In G. G. Montgomery, ed., The
. 1985b. Use of hands and feet of three-toed sloths (Bradypus variegatus) during climbing and terrestrial locomotion. Journal of Mammalogy 66: 35966.
Mendes, S. L., and M. P. Padovan. 2000. A estao biolgica de
Santa Lcia, Santa Teresa, Esprito Santo. Boletim do Museu
de Biologia Prof. Mello Leitao, Nova Srie 11/12: 734.
Mndez Alzola, R. 1950. Estudio sobre la obra cientfica de Larraaga, su iconografa paleomastozoolgica. Rex Press, Montevideo, Uruguay.
Mendoza, M., C. M. Janis, and P. Palmqvist. 2002. Characterizing complex craniodental patterns related to feeding behavior in ungulates: a multivariate approach. Journal of Zoology
258: 22346.
Mercer, J. B., and H. T. Hammel. 1989. Total calorimetry and
temperature regulation in the nine-banded armadillo. Acta
Physiologica Scandinavica 135: 57989.
Meritt, D. A. 1971a. The silky anteater Cyclopes didactylus in
captivity. International Zoo Yearbook 11: 19394.
. 1971b. The development of the LaPlata three-banded armadillo, Tolypeutes matacus at Lincoln Park Zoo, Chicago.
International Zoo Yearbook 11: 19596.
. 1973a. Some observations on the giant armadillo Priodontes giganteus. Zoologica 59: 103.
. 1973b. Observations on the giant armadillo, Priodontes
giganteus in captivity. In Proceedings of the Annual Conference of the American Association of Zoological Parks and
Aquariums, Wheeling, W. Va., 110.
. 1973c. The behavior of armadillos in nature. American
Philosophical Society Yearbook 19711972: 383384.
. 1976a. The nutrition of edentates. International Zoo
Yearbook 16: 3846.
. 1976b. The La Plata three-banded armadillo in captivity.
International Zoo Yearbook 16: 15356.
. 1977. The natural history, behavior, nutrition, physiology, reproduction, development and management of Hoffmanns sloth, Choloepus hoffmanni (Peters). Masters thesis,
Northeastern Illinois University, Chicago.
. 1985a. The two-toed Hoffmanns sloth, Choloepus hoffmanni Peters. In G. G. Montgomery, The evolution and ecology of armadillos, sloths, and vermilinguas, 33341. Smithsonian Institution Press, Washington, D.C.
. 1985b. Naked-tail armadillos, Cabassous sp.: behavior,
nutrition, development and management. In G. G. Montgomery, ed., The evolution and ecology of armadillos, sloths,
Washington, D.C.
. 1985c. The fairy armadillo, Chlamyphorus truncatus. In
G. G. Montgomery, ed., The evolution and ecology of arma-
341
dillos, sloths, and vermilinguas, 39396. Smithsonian Institution Press, Washington, D.C.
. 2006. Research questions on the behavior and ecology
of the giant armadillo (Priodontes maximus). Edentata 7:
3033.
Meritt, D. A., and K. Benirschke. 1973. The chromosomes of
Zaedyus pichiy Ameghino, 1889. Mammalian Chromosomes
Newsletter 14: 1089.
Merrett, P. K. 1983. Edentates. In Project for city and guilds:
animal management course, 3948. The Zoological Trust of
Guernsey, Guernsey.
Merriam, D. F. 2002. The armadillo [Dasypus novemcinctus
(Linnaeus)] invasion of Kansas. Transactions of the Kansas
Academy of Science 105: 4450.
Messias-Costa, A., et al. 2001. Order Xenarthra (Edentata)
(sloths, armadillos, anteaters). In M. E. Fowler and Z. S. Cubas, eds., Biology, medicine, and surgery of South American
wild animals, 23855. Iowa State University Press, Ames.
Meyers, D. M., A. C. Caparo, and J. P. Moreno. 1977. Isolation
of serotype hardjo and other leptospirae from armadillos in
Argentina. Bulletin of the Pan American Health Organization 11: 13139.
Meyers, W. M., et al. 1978. Indigenous leprosy in nine-banded
armadillos. In The armadillo as an experimental model in
biomedical research, 6772. Scientific publication 366, Pan
American Health Organization, Washington, D.C.
Mikich, S. B., and R. S. Bernils, eds. 2004. Livro vermelho da
fauna ameaada no estado do Paran. Instituto Ambiental
do Paran, Curitiba, Brazil.
Milton, K. 1979. Factors influencing leaf choice by howler monkeys: a test of some hypotheses of food selection by generalist herbivores. American Naturalist 114: 36278.
. 1991. Leaf change and fruit production in six neotropical
Moraceae species. Journal of Ecology 79: 126.
Milton, K., E. A. Laca, and M. W. Demment. 1994. Successional
patterns of mortality and growth of large trees in a Panamanian lowland forest. Journal of Ecology 82: 7987.
Minoprio, J. D. L. 1945. Sobre el Chlamyphorus truncatus Harlan. Acta Zoologica Lilloana 3: 558.
Miranda, G. H. B. 2004. Ecologia e conservao do tamandubandeira (Myrmecophaga tridactyla) no Parque Nacional
das Emas. Ph.D. diss., Universidade de Braslia, Brazil.
Miranda, G. H. B., et al. 2003. Giant anteater (Myrmecophaga
tridactyla) beehive foraging at Emas National Park, Brazil.
Edentata 5: 55.
Miranda, R. F., et al. 2004. Retrospective study of causes of
death in giant anteater (Myrmecophaga tridactyla) at Fundao Parque Zoolgico de So Paulo (FPZSP) from 1964 to
2003. In AAZV/AAWV, ed., Joint Conference AAZV/AAWV,
605. San Diego.
Mishima, M., et al. 1999. Diurnal variations of blood pressure in
dogs. Journal of Veterinary Medical Science 61: 64347.
Mitchell, C. L., and E. Raez Luna. 1990. The effects of hunting
by indigenous people on prey densities in Manu National
Park, Peru. Technical report, Wildlife Conservation Society,
New York.
Moeller, W. 1968. Allometrische Analyse der Grteltierschdel.
Ein Beitrag zur Phylogenie de Dasypodidae Bonaparte,
1938. Zoologische Jahrbcher 85: 411528.
Mohr, C. O. 1947. Table of equivalent populations of North
342
Bibliography
American mammals. American Midland Naturalist 37:

22349.
Moncrief, N. D. 1988. Absence of genic variation in a natural
population of nine-banded armadillos, Dasypus novemcinctus (Dasypodidae). Southwestern Naturalist 33: 22931.
Mones, A. 1986. Palaeovertebrata Sudamericana. Catlogo
sistemtico de los vertebrados fsiles de America del Sur, Parte
1. Lista preliminar y bibliografa. Courier Forschungsinstitut
Seneckenberg 82: 1625, E. Schweizerbart, Stuttgart.
Mones, A., and A. Ximenez. 1980. Hallazgo de dos mamferos
extintos del Uruguay. Revista de la Facultad de Humanidades y Ciencias, Serie Ciencias Biolgicas 1: 2016.
Monot, M., et al. 2005. On the origin of leprosy. Science 308:
104042.
Montgomery, G. G. 1979. El grupo alimenticio (feeding guild)
del oso hormiguero: convivencia y especializacin de las
presas de sustento de los osos hormigueros neotropicales
(Edentata, Myrmecophagidae). Concincia 6: 36.
. 1983. Cyclopes didactylus (tapacara, serafin de platanar,
silky anteater). In D. H. Janzen, ed., Costa Rican natural history, 46163. University of Chicago Press, Chicago.
, ed. 1985a. The evolution and ecology of armadillos, sloths,
and vermilinguas. Smithsonian Institution Press, Washington, D.C.
. 1985b. Movements, foraging and food habits of the four
extant species of neotropical vermilinguas (Mammalia;
Myrmecophagidae). In G. G. Montgomery, ed., The evolution and ecology of armadillos, sloths and vermilinguas, 365
. 1985c. Impact of vermilinguas (Cyclopes, Tamandua:
Xenarthra = Edentata) on arboreal ant populations. In G.
G. Montgomery, ed., The evolution and ecology of armadillos, sloths and vermilinguas, 35163. Smithsonian Institution
Montgomery, G. G., and Y. D. Lubin. 1977. Prey influences on
movements of neotropical anteaters. In R. L. Philips and C.
Jonkel, eds., Proceedings of the 1975 predator symposium,
10331. University of Montana, Missoula.
Montgomery, G. G., and M. E. Sunquist. 1974. Contact distress
calls of young sloths. Journal of Mammalogy 55: 211213.
. 1975. Impact of sloths on Neotropical forest energy flow
and nutrient cycling. In F. B. Golley and E. Medina, eds.,
Tropical ecological systems; trends in terrestrial and aquatic
research, 6998. Springer-Verlag, Berlin.
. 1978. Habitat selection and use by two-toed and threetoed sloths. In G. G. Montgomery, ed., The ecology of arboreal folivores, 32959. Smithsonian Institution Press, Washington, D.C.
Moore, D. M. 1978. Post-glacial vegetation in the south Patagonian territory of the giant ground sloth, Mylodon. Botanical
Journal of the Linnean Society 77: 177202.
Moraes-Barros, N. 2003. Estudo da variabilidade e estrutura
gentica em populaes de preguias do gnero Bradypus
(Xenarthra). Ph.D. diss., Universidade de So Paulo, Brazil.
Moraes-Barros, N., J. S. Morgante, and C. Y. Miyaki. 2002. Genetic diversity in different populations of sloths assessed by
DNA fingerprinting. Brazilian Journal of Biology 62: 5038.
Moraes-Barros, N., et al. 2006. Comparative phylogeography
of the Atlantic forest endemic sloth (Bradypus torquatus)
and the widespread three-toed sloth (Bradypus variegatus)

(Bradypodidae, Xenarthra). Genetica 126: 18998.
Moran, E. F. 1981. Developing the Amazon. Indiana University
Press, Bloomington.
Morgan, G. S., and C. A. Woods. 1986. Extinction and the zoogeography of West Indian land mammals. Biological Journal
of the Linnean Society 28: 167203.
Moser, H. G., and K. Benirschke. 1962. The fetal zone of the adrenal gland in the nine-banded armadillo, Dasypus novemcinctus. Anatomical Record 143: 4760.
Mossman, H. W. 1987. Vertebrate fetal membranes. MacMillan,
Houndmills, United Kingdom.
Mota, D. L. da, et al. 1992. An immunohistochemical study
on the pancreatic endocrine cells of the three-toed sloth,
Bradypus variegatus. Archives of Histology and Cytology 55:
2039.
Mouro, G., and . M. Medri. 2002. A new way of using inexpensive large-scale assembled GPS to monitor giant anteaters in short time intervals. Wildlife Society Bulletin 30:
10291032.
Mouse Genome Sequencing Consortium. 2002. Initial sequencing and comparative analysis of the mouse genome. Nature
420: 52062.
Mhlbauer, M., et al. 2006. Fecal estradiol and progesterone
metabolite levels in the three-toed sloths (Bradypus variegatus). Brazilian Journal of Medical and Biological Research
39: 28995.
Muizon, C. de, and H. G. McDonald. 1995. An aquatic sloth
from the Pliocene of Peru. Nature 375: 22427.
Muizon, C. de, et al. 2003. A new early species of the aquatic
sloth Thalassocnus (Mammalia, Xenarthra) from the late
Miocene of Peru. Journal of Vertebrate Paleontology 23:
88694.
Muizon, C. de, et al. 2004a. The youngest species of the aquatic
sloth Thalassocnus and a reassessment of the relationships
of the nothrothere sloths (Mammalia, Xenarthra). Journal
Muizon, C. de, et al. 2004b. The evolution of feeding adaptations of the aquatic sloths Thalassocnus. Journal of Vertebrate Paleontology 24: 398410.
Muoz-Rivas, G. 1978. Notes on granulomatosis in armadillos
innoculated with M. leprae. In The armadillo as an experimental model in biomedical research, 99102. Scientific publication 366, Pan American Health Organization, Washington, D.C.
Murata, K., and R. Masuda. 1996. Gender determination of the
Linns two-toed sloth (Choloepus didactylus) using SRY
amplified from hair. Journal of Veterinary Medical Sciences
58: 115759.
Murphy, M. G., and J. L. OLeary. 1973. Hanging and climbing
functions in racoon and sloth after total cerebellectomy. Archives of Neurology 28: 11117.
Murphy, W. J., E. Eizirik, W. E. Johnson, et al. 2001. Molecular
phylogenetics and the origins of placental mammals. Nature
409: 61418.
Murphy, W. J., E. Eizirik, S. J. OBrien, et al. 2001. Resolution of
the early placental mammal radiation using Bayesian phylogenetics. Science 294: 234851.
Murphy, W. J., P. A. Pevzner, and S. J. OBrien. 2004. Mam-
Bibliography
malian phylogenomics comes of age. Trends in Genetics 20:

63139.
Murphy, W. J., R. Stanyon, and S. J. OBrien. 2001. Evolution of
mammalian genome organization inferred from comparative gene mapping. Genome Biology 2: Reviews0005Reviews0008.
Murphy, W. J., et al. 2005. Dynamics of mammalian chromosome evolution inferred from multispecies comparative
maps. Science 309: 61317.
Murray, P. 1984. Furry egg-layers: the monotreme radiation.
In M. Archer and G. Clayton, eds., Vertebrate zoogeography
and evolution in Australasia, 57183. Hesperian Press, Carlisle, Western Australia.
Nagy, F., and R. H. Edmonds. 1973a. Morphology of the reproductive system of the armadillo. The spermatogonia. Journal
of Morphology 140: 30720.
. 1973b. Some observations of the fine structure of armadillo spermatozoa. Journal of Reproduction and Fertility
34: 55153.
Nagy, K. A., and G. G. Montgomery. 1980. Field metabolic
rate, water flux, and food consumption in three-toed sloths
(Bradypus variegatus). Journal of Mammalogy 61: 46572.
Naiff, R. D., and T. Barret. 1989. Novos registros de Paracoccidioides brasiliensis em tatus (Dasypus novemcinctus). In
Proceedings Congresso Brasileiro de Parasitologia, Rio de
Janeiro, 197.
Naiff, R. D., W. Y. Mok, and M. F. Naiff. 1985. Distribution of
Histoplasma capsulatum in Amazonian wildlife. Mycopathologia 89: 16568.
Naiff, R. D., et al. 1986. Enzootic paracoccidioidomycosis in
armadillos (Dasypus novemcinctus) in the state of Para.
Revista do Instituto de Medicina Tropical de So Paulo 28:
1927.
Naiff, R. D., et al. 1996. New records of Histoplasma capsulatum
from wild animals in the Brazilian Amazon. Revista do Instituto de Medicina Tropical de So Paulo 38: 27377.
Nakakura, K., et al. 1982. Fetal-maternal gradients of steroid
hormones in the nine-banded armadillo (Dasypus novemcinctus). Journal of Reproduction and Fertility 66: 63543.
Naples, V. L. 1985. The superficial facial musculature of the
sloths and anteaters. In G. G. Montgomery, The evolution
. 1986. The morphology and function of the hyoid region
in the tree sloths, Bradypus and Choloepus. Journal of Mammalogy 67: 71224.
. 1987. Reconstruction of cranial morphology and analysis of function in Nothrotheriops shastensis. Contributions
in Science, Los Angeles County Museum of Natural History
389: 121.
. 1989. The feeding mechanism in the Pleistocene ground
sloth, Glossotherium. Contributions in Science, Los Angeles
County Museum of Natural History 415: 123.
. 1999. Morphology, evolution and function of feeding
in the giant anteater (Myrmecophaga tridactyla). Journal of
Zoology 249: 1941.
. 2001. Anteaters. In D. M. MacDonald, ed., The new encyclopedia of mammals, 78891. Oxford University Press,
New York.
343
National Human Genome Research Institute. 2002. Sequencing/resequencing workshop summary. http://www.genome.
gov/10005564. Accessed 2005 May.
. 2003. ENCODE project target selection process and
target regions. http: //www.genome.gov/10506161. Accessed
2005 May.
Navarro, G., and A. Fuentes. 1999. Geobotnica y sistemas
ecolgicos de paisaje en el Gran Chaco de Bolivia. Revista
Boliviana de Ecologa y Conservacin Ambiental 5: 2550.
Neill, M. A., A. W. Hightower, and C. V. Broome. 1985. Leprosy
in the United States, 19711981. Journal of Infectious Disease
152: 106469.
Neris, N., et al. 2002. Gua de mamferos medianos y grandes del
Paraguay. Distribucin, tendencia poblacional y utilizacin.
Secretaria del Ambiente, Madame Lynch 3.5000 C/Premer
Presidente, Asuncin, Paraguay.
Nesbitt, S. A., et al. 1977. Foods of the nine-banded armadillo in
Florida. Proceedings of the Annual Conference of Southeastern Fish and Wildlife Agencies 31: 5761.
Newman, H. H. 1912. The ovum of the nine-banded armadillo:
growth of the oocytes, maturation and fertilization. Biological Bulletin 23: 100141.
. 1917. The biology of twins. University of Chicago Press,
Chicago.
Newman, H. H., and J. T. Patterson. 1909. A case of normal
identical quadruplets in the nine-banded armadillo, and its
bearing on the problems of identical twins and of sex determination. Biological Bulletin 17: 18187.
. 1910. The development of the nine-banded armadillo
from primitive streak to birth, with special reference to
the question of polyembryony. Journal of Morphology 21:
359423.
Nickerson, D. A., V. O. Tobe, and S. L. Taylor. 1997. PolyPhred:
automating the detection and genotyping of single nucleotide substitutions using fluorescence-based resequencing.
Nucleic Acids Research 25: 274551.
Nikaido, M., A. P. Rooney, and N. Okada. 1999. Phylogenetic
relationships among cetartiodactyls based on insertions of
short and long interspersed elements: hippopotamuses are
the closest extant relatives of whales. Proceedings of the National Academy of Sciences, U.S.A. 96: 1026166.
Nikaido, M., et al. 2003. Ancient SINEs from African endemic
mammals. Molecular Biology and Evolution 20: 52227.
Nill, A. 1907. Die Fortpflanzung des Groen Ameisenbren
(Myrmecophaga jubata) in Nills Zoologischem Garten
Stuttgart. Zoologische Beobachtungen 4S: 14551.
Nishihara, H., et al. 2005. A retroposon analysis of Afrotherian
phylogeny. Molecular Biology and Evolution 22: 182333.
Nodot, L. 1857. Description dun nouveau genre ddent fossile renfermant plusieurs espces voisines des Glyptodonts,
et classification mthodique de treize espces appartenant
des deux genres. Mmoires de lAcadmie Impriale des
Science, Arts et Belles-Lettres de Dijon 2: 1170, Atlas 112
lminas.
Noss, A. 1998. El monitoreo comunitario de cacera en el Izozog: datos preliminares. Ecologa en Bolivia 31: 5366.
. 2000. La sostenibilidad de la cacera de subsistencia Izocea. In E. Cabrera, C. Mercolli, and R. Resquin, eds., Manejo de fauna silvestre en Amazona y Latinoamrica, 53544.
344
Bibliography
CITES Paraguay, Fundacin Moises Bertoni, University of

Florida, Asuncin, Paraguay.
Noss, A., and E. Cullar. 2000. ndices de abundancia para
fauna terrestre en el Chaco boliviano: huellas en parcelas y
en brechas barridas. In E. Cabrera, C. Mercolli, and R. Resquin, eds., Manejo de fauna silvestre en Amazona y Latinoamrica, 7382. CITES Paraguay, Fundacin Moiss Bertoni,
University of Florida, Asuncin, Paraguay.
Noss, A., E. Cullar S., and R. L. Cullar S. 2003. Hunter selfmonitoring as a basis for biological research: data from the
Bolivian Chaco. Mastozoologa Neotropical 10: 4967.
. 2004. An evaluation of hunter self-monitoring in the
Bolivian Chaco. Human Ecology 32: 685702.
Noss, A., I. Oetting, and R. L. Cullar. 2005. Hunter self-monitoring by the Isoseo-Guaran in the Bolivian Chaco. Biodiversity and Conservation 14: 267993.
Noss, A., and M. D. Painter. 2004. Community-based wildlife
management in the Gran Chaco, Bolivia. In K. M. Silvius,
R. E. Bodmer, and J. M. V. Fragoso, eds., People in nature:
wildlife conservation in South and Central America, 5975.
Columbia University Press, New York.
Noss, A., R. Pea, and D. I. Rumiz. 2004. Camera trapping Priodontes maximus in the dry forests of Santa Cruz, Bolivia.
Endangered Species Update 21: 4352.
Novacek, M. J. 1992. Mammalian phylogeny: shaking the tree.
Nature 356: 12125.
Novacek, M. J., and A. R. Wyss. 1986. Higher-level relationships
of the recent eutherian orders: morphological evidence.
Cladistics 2: 25787.
Nowak, R. M. 1991. Walkers mammals of the world. 5th ed.
Johns Hopkins University Press, Baltimore.
. 1999. Walkers mammals of the world. 6th ed. Johns
Hopkins University Press, Baltimore.
OLeary, M. 1999. Parsimony analysis of total evidence from extinct and extant taxa and the cetacean-artiodactyl question
(Mammalia, Ungulata). Cladistics 15: 31530.
OLeary, M., and J. H. Geisler. 1999. The position of Cetacea
within Mammalia: phylogenetic analysis of morphological
data from extinct and extant taxa. Systematic Biology 48:
45590.
Oliveira, E. V., and L. P. Bergqvist. 1998. A new Paleocene armadillo (Mammalia, Dasypodoidea) from the Itaborai Basin,
Brazil. Asociacin Paleontolgica Argentina, Publicacin
Especial 5: 3540.
Oliveira, L. H. A., C. P. da Costa, and S. E. Huggins. 1980. Cardiac mass, blood temperature and ventricular fibrillation: a
study of the comparative physiology of the three-toed sloth
and the domestic cat. Comparative Biochemistry and Physiology 67A: 48390.
Ono, M. A., et al. 2001. Canine paracoccidioidomycosis: a seroepidemiologic study. Medical Mycology 39: 27782.
Oria, J. 1941. Bao mieloide e megacariocitos nos bradipodideos, dasipodideos e mirmecofagideos adultos. Anais da
Faculdade de Medicina da Universidade de Sao Paulo 17:
13961.
Orr, C. M. 2005. Knuckle-walking anteater: a convergence test
of adaptation for purported knuckle-walking features of African Hominidae. American Journal of Physical Anthropology 128: 63958.
Osborn, D. A., M. I. Nelson, and R. J. Warren. 2000. Armadillo
diets among seasons and between habitats on Cumberland

Island, Georgia. Proceedings of the Annual Conference of the
Southeastern Association of Fish and Wildlife Agencies 54:
28291.
Ostenrath, F. 1974. Haltung von Riesengrteltieren (Priodontes giganteus) im Zoo Duisburg. Zeitschrift des Klner Zoo
4: 14546.
Owen, R. 1838. Note on the Glyptodon. In W. Parish, ed., Buenos
Ayres and the provinces of the Ro de la Plata: their present
state, trade, and debt; with some account from original documents of the progress of geographical discovery in those parts
of South America during the last sixty years, 178a178e. John
Murray Press, London.
. 1839. Description of a tooth and part of the skeleton
of the Glyptodon clavipes, a large quadruped of the Edentate order, to which belongs the tesselated bony armour
described and figures by Mr. Clift in the former Volume of
the Transactions of the Geological Society; with a consideration of the question whether the Megatherium possessed
an analogous dermal armour. Transactions of the Geological
Society of London 2: 81106.
. 1842. Description of the skeleton of an extinct gigantic
sloth, Mylodon robustus, Owen, with observations on the osteology, natural affinities, and probable habits of the megatherioid quadruped in general. R. and J. E. Taylor, London.
. 1851. On the Megatherium (Megatherium americanum,
Blumenbach). I. Preliminary observations on the exogenous
processes of vertebrae. Philosophical Transactions of the Royal Society of London 141: 71964.
. 1856. On the Megatherium (Megatherium americanum
Cuvier and Blumenbach). III. The skull. Philosophical Transactions of the Royal Society of London 146: 57189.
Pacheco, J., and C. J. Naranjo. 1978. Field ecology of Dasypus
sabanicola in the flood savanna of Venezuela. In The armadillo as an experimental model in biomedical research, 1317.
Scientific publication 366, Pan American Health Organization, Washington, D.C.
Paige, C. F., D. T. Scholl, and R. W. Truman. 2002. Prevalence
and incidence density of Mycobacterium leprae and Trypanossoma cruzi infections within a population of wild
nine-banded armadillos. American Journal of Tropical Medicine and Hygiene 67: 52832.
Painter, M., and Noss, A. 2000. Conservacin de fauna con organizaciones comunales: experiencia con el pueblo Izoceo
en Bolivia. In E. Cabrera, C. Mercolli, and R. Resquin, eds.,
Manejo de fauna silvestre en Amazona y Latinoamrica,
16780. CITES Paraguay, Fundacin Moises Bertoni, University of Florida, Asuncin, Paraguay.
Painter, T. S. 1925. Chromosome number in mammals. Science
61: 42324.
Pang, K. C., et al. 2005. RNAdba comprehensive mammalian noncoding RNA database. Nucleic Acids Research 33:
D12530.
Parera, A. 2002. Los mamferos de la Argentina y la regin austral de Amrica del Sur. Editorial El Ateneo, Buenos Aires.
Parker, G. A. 1982. Why so many tiny sperm? The maintenance
of two sexes with internal fertilization. Journal of Theoretical
Biology 96: 28194.
. 1984. Sperm competition and the evolution of animal
mating strategies. In R. L. Smith, ed., Sperm competition and
Bibliography
the evolution of animal mating strategies, 160. Academic

Press, Orlando.
Parra, G., et al. 2003. Comparative gene prediction in human
and mouse. Genome Research 13: 10817.
Partridge, J. 1991. The two-toed sloth. International Zoo News
34: 2328.
Pascual, R. 1960. Una nueva superfamilia Entelopsoidea:
descripcin de la nueva especie Entelops parodii. Acta Geologica Lilloana 3: 12746.
Pascual, R., and E. Ortiz Jaureguizar. 1990. Evolving climates
and mammal faunas in Cenozoic South America. Journal of
Human Evolution 19: 2360.
Pascual, R., E. Ortiz Jaureguizar, and J. L. Prado. 1996. Land
mammals: paradigm for Cenozoic South American geobiotic evolution. Mnchner Geowissenschaftliche Abhandlungen (A) 30: 265319.
Patterson, B., and R. Pascual. 1968. Evolution of mammals on
the southern continents. Quarterly Review of Biology 43:
40951.
. 1972. The fossil mammal fauna of South America. In
A. Keast, F. C. Erk, and B. Glass, eds., Evolution, mammals,
and southern continents, 247309. State University of New
York Press, Albany.
Patterson, B., W. Segall, and W. D. Turnbull. 1989. The ear region in xenarthrans (= Edentata, Mammalia). Part I. Cingulates. Fieldiana, Geology, n.s. 18: 146.
Patterson, B., et al. 1992. The ear region in xenarthrans (=
Edentata: Mammalia). Part II. Pilosa (sloths and anteaters),
palaeanodonts, and a miscellany. Fieldiana Geology, n.s. 24:
179.
Patterson, J. T. 1913. Polyembryonic development in Tatusia
novemcincta. Journal of Morphology 24: 559684.
Patzl, M., et al. 1998. Monitoring ovarian cycle and pregnancy
in the giant anteater (Myrmecophaga tridactyla) by faecal
progestagen and oestrogen analysis. Animal Reproduction
Science 53: 20919.
Paula Couto, C. de. 1979. Tratado de paleomastozoologia. Academia Brasileira de Cincias, Rio de Janeiro.
. 1980. Fossil Pleistocene to Sub-recent mammals from
northeastern Brazil. I. Edentata Megalonychidae. Academia
Brasileira de Ciencias Anais 52: 14351.
Pedrosa, M. A. C., et al. 2002. The effect of feeding on the respiratory activity of the sloth. Brazilian Journal of Medical and
Biological Research 35: 85154.
Pea, R., and E. Cullar. 2005. Utilidad del monitoreo con
trampas-cmara en evaluaciones de edentados en el Chaco,
Santa Cruz, Bolivia. Technical Report Number 154, Wildlife
Conservation Society, Bolivia.
Peppler, R. D. 1979. Reproductive parameters in the nine-banded armadillo. Anatomical Record 193: 64950.
. 1985. Reproductive parameters in the nine-banded armadillo: correlation of follicular development and plasma
progesterone concentration with age. Anatomical Record
211: 351.
. 1999. Armadillo. In E. Knobil and J. Neill, eds., Encyclopedia of reproduction, 27480. Academic Press, San Diego.
Peppler, R. D., and J. Canale. 1980. Quantitative investigation
of the annual pattern of follicular development in the ninebanded armadillo (Dasypus novemcinctus). Journal of Reproduction and Fertility 59: 19397.
345
Peppler, R. D., F. E. Hossler, and S. C. Stone. 1986. Determination of reproductive maturity in the female armadillo
(Dasypus novemcinctus). Journal of Reproduction and Fertility 76: 14146.
Peppler, R. D., and S. C. Stone. 1976. Plasma progesterone level
in the female armadillo during delayed implantation and
gestation: preliminary report. Laboratory Animal Science
26: 5014.
. 1980a. Plasma progesterone level during delayed implantation, gestation and the postpartum period in the female armadillo. Laboratory Animal Science 30: 18891.
. 1980b. Clomiphene-induced ovulation in the 9banded
armadillo (Dasypus novemcinctus). Laboratory Animals 14:
32930.
. 1981. Annual pattern in plasma testosterone in the male
armadillo (Dasypus novemcinctus). Animal Reproduction
Science 4: 4953.
Peralta De Galmarini, I., and E. Martinez Carretero. 1995. Guas
botnicas para la Provincia de Mendoza. II. Reserva Natural
Telteca. INCA Editorial, Mendoza, Argentina.
Perea, D. 1992. Aproximacin a la sistemtica filogentica de
Mylodontinae (Xenarthra). Boletn Sociedad Zoologa Uruguay 7: 3334.
Peredo, B. 1999. Bolivias trade in hairy armadillos. TRAFFIC
Bulletin 18: 4145.
Pereira, H. R. J., W. Jorge, and M. E. L. T. da Costa. 2004. Chromosome study of anteaters (Myrmecophagidae/Xenarthra):
a preliminary report. Genetics and Molecular Biology 27:
39194.
Peres, C. A. 1999. The structure of nonvolant mammal communities in different amazonian forest types. In J. F. Eisenberg
and K. H. Redford, eds., Mammals of the Neotropics. Vol. 3,
56481. University of Chicago Press, Chicago.
Prez, L. M. 2001. Morphology and function of the hyoid apparatus of fossil xenarthrans. Journal of Morphology 248: 271.
Prez, L. M., G. J. Scillato-Yan, and S. F. Vizcano. 2000. Estudio morfofuncional del aparato hioideo de Glyptodon cf.
clavipes Owen (Cingulata: Glyptodontidae). Ameghiniana
37: 29399.
Prez Zapata, A., A. D. Vitullo, and O. A. Reig. 1987. Karyotypic
and sperm distinction of Calomys hummelincki from Calomys laucha (Rodentia: Cricetidae). Acta Cientfica Venezolana 38: 9093.
Perrogn, E. 2001. Estudio de la abundancia y frecuencia de uso
de hbitat de la familia Dasypodidae en el Chaco Boliviano
(Izozog, Provincia Cordillera, Santa Cruz, Bolivia). Technical Report Number 77, Kaa-Iya Project (WCS, CABI), Santa
Cruz, Bolivia.
Peterson, E. A., and W. C. Heaton. 1968. Peripheral auditory
responses in representative edentates. Journal of Auditory
Research 8: 17184.
Pfeiffer, L., and R. W. Truman. 2003. Hansens disease in the
United States and Louisiana in 2001. Louisiana Morbidity
Report 14: 13.
Philippe, H. 1997. Rodent monophyly: pitfalls of molecular
phylogenies. Journal of Molecular Evolution 45: 71215.
Pictet, F. 1853. Trait de palontologie ou histoire naturelle des
animaux fossiles considers dans leurs rapports zoologiques et
gologiques. J.-B. Baillire, Paris.
Piggins, D., and W. R. A. Muntz. 1985. The eye of the three-toed
346
Bibliography
sloth. In G. G. Montgomery, ed., The evolution and ecology

Pinder, L. 1985. Observaes preliminares sobre a preguia de
coleira (Bradypus torquatus) (Illiger, 1811) (Edentata, Bradypodidae). In Universidade Estadual de Campinas, ed.,
Anais de Resumos do XII Congresso Brasileiro de Zoologia,
Campinas, 29091. Universidade Estadual de Campinas,
Campinas, So Paulo.
. 1993. Body measurements, karyotype, and birth frequencies of maned sloth (Bradypus torquatus). Mammalia
57: 4348.
Pisani, D. 2004. Identifying and removing fast-evolving sites
using compatibility analysis: an example from the arthropoda. Systematic Biology 53: 97889.
Platt, S. G., and T. R. Rainwater. 2003. Above-ground nest of
the nine-banded armadillo in Belize. Florida Field Naturalist 31: 89.
Platt, S. G., T. R. Rainwater, and S. W. Brewer. 2004. Aspects of
the burrowing ecology of nine-banded armadillos in northern Belize. Mammalian Biology 69: 21724.
Plese, T., and S. Moreno. 2005. Protocolos de rehabilitacin y
reintroduccin o reubicacin de osos perezosos de dos y tres
uas. Fundacin UNAU, Corantioquia, Colombia.
Plotnick, R., and T. K. Baumiller. 2000. Invention by evolution:
functional analysis in paleobiology. Supplement to Paleobiology 26: 30521.
Pocock, R. I. 1924. The external characters of the South American edentates. Proceedings of the Zoological Society of London 65: 9831031.
Poinar H. N., et al. 1998. Molecular coproscopy: dung and diet
of the extinct ground sloth Nothrotheriops shastensis. Science 281: 4026.
Poinar H. N., et al. 2003. Nuclear gene sequences from a late
Pleistocene sloth coprolite. Current Biology 13: 115052.
Polanco-Ochoa, R., ed. 2003. Manejo de fauna silvestre en Amazona y Latinoamrica: seleccin de trabajos V Congreso Internacional. Fundacin Natura, Bogot, Colombia.
Polini, N. N., R. Camina, and E. B. Casanave. 1999. Morphological and morphometric study of blood leucocytes from
Chaetophractus villosus, Mammalia, Dasypodidae. Comparative Haematology International 9: 16267.
Poljak, S., M. Lizarralde, and M. S. Merani. 2004. Caractersticas ecolgico poblacionales en tres especies de armadillos
de la provincia de Buenos Aires. In Congreso Nacional de
Conservacin de la Biodiversidad, Temaikn, Escobar, Provincia de Buenos Aires, Argentina, 11516.
Posey, D. A. 1983a. Indigenous ecological knowledge and development of the Amazon. In E. F. Moran, ed., The dilemma of
Amazonian development, 22557. Westview Press, Boulder,
Colo.
. 1983b. Indigenous knowledge and development: an ideological bridge to the future. Cincia e Cultura 35: 87794.
Prabhakaran, K., et al. 1984. An alternative route for infecting armadillos with Mycobacterium leprae. Microbios 39:
8386.
Prevosti, F. J., and S. F. Vizcano. 2006. Paleoecology of the large
carnivore guild from the late Pleistocene of Argentina. Acta
Palaeontologica Polonica 51: 40722.
Prodhl, P. A., et al. 1996. Molecular documentation of polyem-
bryony and the micro-spatial dispersion of clonal sibships

in the nine-banded armadillo, Dasypus novemcinctus. Proceedings of the Royal Society, London, Series B 263: 164349.
Prodhl, P. A., et al. 1998. Genetic maternity and paternity in
a local population of armadillos assessed by microsatellite
DNA markers and field data. American Naturalist 151: 719.
Provincia de Mendoza, Argentina. 1998. Ley provincial 6599.
<http://www.tribunet.com.ar/tribunet/ley/6599.htm>. Accessed 10 April, 2006.
Puertas, P. E. 1999. Hunting effort analysis in northeastern Peru:
the case of the Reserva Comunal Tamshiyacu-Tahuayo.
Masters thesis, University of Florida, Gainesville.
Pujos, F. 2000. Scelidodon chiliensis (Xenarthra, Mammalia) du
Plistocene terminal de Pampa de los Fsiles (nord-Prou).
Quaternaire 11: 197206.
. 2002. Contribution la connaissance des Tardigrades
(Mammalia: Xenarthra) du Plistocne pruvien: systmatique, phylognie, anatomie fonctionnelle et extinction.
Ph.D. diss., Musum National dHistoire Naturelle, Paris.
. 2006. Megatherium celendinense sp. nov. from the Pleistocene of the Peruvian Andes and the phylogenetic relationships of megatheriines. Palaeontology 49: 285306.
Pujos, F., and G. De Iuliis. 2007. Late Oligocene Megatherioidea
fauna (Edentata: Xenarthra) from Salla-Luribay (Bolivia):
new data on basal sloth radiation and Cingulata-Phyllophaga split. Journal of Vertebrate Paleontology 27: 13244.
Pujos, F., and R. Salas. 2004a. A new species of Megatherium
(Mammalia: Xenarthra: Megatheriidae) from the Pleistocene of Sacaco and Tres Ventanas, Peru. Palaeontology 47:
579604.
. 2004b. A systematic reassessment and paleogeographic
review of fossil Xenarthra from Peru. Bulletin de lInstitut
Franais dtudes andines 33: 33177.
Pujos, F., R. Salas, and J. Mattos. 2002. Andean lineage of Pleistocene Megatherium: geographical implications. Journal of
Vertebrate Paleontology 22: 97A.
Pujos, F., et al. 2007. A peculiar climbing Megalonychidae from
the Pleistocene of Peru and its implications for sloth history.
Zoological Journal of the Linnean Society 148: 179235.
Purtilo, D. T., et al. 1975. The immune system of the nine-banded armadillo (Dasypus novemcinctus Linn.). Anatomical
Record 181: 72534.
Purvis, A., J. L. Gittleman, and T. M. Brooks. 2005. Phylogeny
and conservation. In A. Purvis, J. L. Gittleman and T. M.
Brooks, eds., Phylogeny and conservation, 117. Cambridge
Queiroz, H. L. de. 1995. Preguias e guaribas: os mamferos folivoros arborcolas do Mamirau. Conselho Nacional de Desenvolvimento Cientfico e Tecnolgico e Sociedade Civil
Mamirau, Braslia, DF.
Quinn, T. W., et al. 1987. DNA marker analysis detects multiple
maternity and paternity in single broods of the lesser snow
goose. Nature 326: 39294.
Radinsky, L. B. 1987. The evolution of vertebrate design. University of Chicago Press, Chicago.
Ralls, K. 1976. Mammals in which females are larger than males.
Quarterly Review of Biology 51: 24576.
Ramrez Rozzi, F., and I. Podgorny. 2001. La metamorfosis del
megaterio. Ciencia Hoy 11: 1219.
Ramprashad, F., et al. 1985. Functional cochlear morphology
Bibliography
of the two-toed sloth (Choloepus sp). In G. G. Montgomery, ed., The evolution and ecology of armadillos, sloths and
Ramsey, P. R., and B. A. Grigsby. 1985. Protein variation in
populations of Dasypus novemcinctus and comparisons to
D. hybridus, D. sabanicola and Chaetophractus villosus. In G.
G. Montgomery, ed., The evolution and ecology of armadillos, sloths, and vermilinguas, 13141. Smithsonian Institution
Ramsey, P. R., et al. 1981. Blood chemistry and nutritional balance of wild and captive armadillos (Dasypus novemcinctus
L.). Comparative Biochemistry and Physiology A 69: 51721.
Rappaport, A. B., and H. Hochman. 1988. Cystic calculi as a
cause of recurrent rectal prolapse in a sloth (Choloepus sp.).
Journal of Zoo Animal Medicine 19: 23536.
Ratajszczak, R., and E. Trzesowska. 1997. Management and
breeding of the larger hairy armadillo, Chaetophractus villosus, at Pozan Zoo. Zoologischer Garten 67: 22028.
Redford, K. H. 1985a. Food habits of armadillos (Xenarthra:
Dasypodidae). In G. G. Montgomery, ed., The evolution
and ecology of armadillos, sloths and vermilinguas, 42937.
. 1985b. Feeding and food preference in captive and wild
giant anteaters (Myrmecophaga tridactyla). Journal of Zoology 205: 55972.
. 1986. Dietary specialization and variation in two mammalian myrmecophages. Revista Chilena de Historia Natural
59: 2018.
Redford, K. H., and J. G. Dorea. 1984. The nutritional value of
invertebrates with emphasis on ants and termites as food for
mammals. Journal of Zoology 203: 38595.
Redford, K. H., and J. F. Eisenberg. 1992. Mammals of the Neotropics. Volume 2. The southern cone: Chile, Argentina,
Uruguay, Paraguay. University of Chicago Press, Chicago.
Redford, K. H., and J. A. Mansour, eds. 1996. Traditional peoples
and biodiversity conservation in large tropical landscapes.
America Verde Publications, The Nature Conservancy,
Washington, D.C.
Redford, K. H., and A. M. Stearman. 1989. Local peoples and
the Beni Biosphere Reserve, Bolivia. Vida Silvestre Neotropical 2: 4956.
Redford, K. H., A. Taber, and J. A. Simonetti. 1990. There is
more to biodiversity than the tropical rain forests. Conservation Biology 4: 32830.
Redford, K. H., and R. M. Wetzel. 1985. Euphractus sexcinctus.
Mammalian Species 252: 14.
Reed, K. D., et al. 1993. Zoonotic transmission of sporotrichosis: case report and review. Clinical Infectious Diseases 16:
38487.
Rees, R. J. 1983. Progress in the preparation of an antileprosy
vaccine from armadillo-derived Mycobacterium leprae. International Journal of Leprosy 51: 51518.
Reichel-Dolmatoff, G. 1976. Cosmology as ecological analysis:
a view from the forest. Man 11: 30718.
Reid, F. A. 1997. A field guide to the mammals of Central America
and southeast Mexico. Oxford University Press, Oxford.
Reinhardt, J. T. 1879. Beskrivelse af Hovedskallen af et Kaempedovendyr, Grypotherium darwinii fra La Plata-Landenes,
plejstocene Dannelser. Kongelige Danske videnskabernes
347
Selskabs Skrifter. Naturvidenskabelig og mathematisk Afdeling 5: 35380.

Reiss, K. Z. 1996. Evolution of the feeding apparatus in ant-eating mammals: when do we stop believing in convergence?
American Zoologist 35: 85A.
. 1997. Myology of the feeding apparatus of myrmecophagid anteaters (Xenarthra: Myrmecophagidae).
. 2001. Using phylogenies to study convergence: the case
of the ant-eating mammals. American Zoologist 41: 50725.
Resoagli, J. P., et al. 1985. Patologa del armadillo en cautividad
2. Etiologa y diagnstico de las afecciones respiratorias.
Veterinaria Argentina 2: 92530.
Restrepo, A. 1985. The ecology of Paracoccidioides brasiliensis:
a puzzle still unsolved. Journal of Medical and Veterinary
Mycology 23: 32334.
Restrepo, A., et al. 2000. Clues to the presence of pathogenic
fungi in certain environments. Medical Mycology 38: 67
77.
Restrepo-Moreno, A. 1994. Ecology of Paracoccidioides brasiliensis. In M. F. Franco, C. S. Lacaz, A. Restrepo, and G. Del
Negro, eds., Paracoccidioidomycosis, 12128. CRC Press,
Boca Raton, Fla.
Reyes, A., et al. 2004. Congruent mammalian trees from mitochondrial and nuclear genes using Bayesian methods. Molecular Biology and Evolution 21: 397403.
Richard-Hansen, C., and E. Taube. 1997. Note on the reproductive behavior of the three-toed sloth, Bradypus tridactylus,
in French Guiana. Mammalia 61: 25963.
Richard-Hansen, C., et al. 1999. Body measurements on 40 species of mammals from French Guiana. Journal of Zoology
247: 41928.
Rideout, B. A., et al. 1985. Stress-induced adrenal changes and
their relation to reproductive failure in captive nine-banded
armadillos (Dasypus novemcinctus). Zoo Biology 4: 12937.
Ridewood, W. G. 1901. On the structure of the hairs of Mylodon
listai and other South American edentata. Quarterly Review
of Microscopical Science n.s. 44: 393411.
Ringuelet, R. A. 1957. Restos de probables huevos de nematodes en el estiercol del edentado extinguido Mylodon listai
(Ameghino). Ameghiniana 1: 1516.
Rippon, J. W. 1988. Medical mycology: the pathogenic fungi and
the pathogenic Actinomycetes. 3rd ed. Saunders, Philadelphia.
Roberts, M., L. Newman, and G. Peterson. 1982. The management and reproduction of the large hairy armadillo Chaetophractus villosus at the National Zoological Park. International Zoo Yearbook 22: 18594.
Robinson, J. G., and E. L. Bennett, eds. 2000. Hunting for sustainability in tropical forests. Columbia University Press,
New York.
Rodrigues, F. H. G., and J. S. Marinho-Filho. 2003. Diurnal rest
sites of translocated lesser anteaters (Tamandua tetradactyla) in the Cerrado of Brazil. Edentata 5: 4446.
Rodrigues, F. H. G., J. S. Marinho-Filho, and H. G. dos Santos.
2001. Home ranges of translocated lesser anteaters Tamandua tetradactyla in the Cerrado of Brazil. Oryx 35: 16669.
Rodrguez, C. A., and M. C. van der Hammen. 2003. Manejo
indgena de fauna en el medio y bajo ro Caquet (Amazona colombiana). Tradicin, transformaciones y desafos
348
Bibliography
para su uso sostenible. In R. Polanco-Ochoa, ed., Manejo

de fauna silvestre en Amazona y Latinoamrica: seleccin de
trabajos V Congreso Internacional, 32538. Fundacin Natura, Bogot, Colombia.
Rodrguez, R., and G. Hernndez. 1992. Bioqumica de algunos
desdentados extinguidos de Cuba. Archaeofauna 1: 1058.
Roig, V. G. 1971. Observaciones sobre la termorregulacin en
Zaedyus pichiy. Acta Zoolgica Lilloana 28: 1318.
. 1995. Situacin de conservacin, biologa y ecologa de
Chlamyphorus truncatus. Edentata 2: 19.
Rojas-Surez, C. 2002. Estimacin de edad de fetos de corechi
Tolypeutes matacus colectados en Isoso. Technical Report
Number 72, Wildlife Conservation Society, Bolivia.
Rojas-Surez, C., and L. Maffei. 2003. Reproduccin de Dasypus novemcinctus en el Izozog, Santa Cruz, Bolivia. Edentata
5: 4754.
Roldan, E. R. S., M. Gomendio, and A. D. Vitullo. 1992a. The
evolution of eutherian spermatozoa and underlying selective forces: female selection and sperm competition. Biological Reviews 67: 55193.
. 1992b. Sperm shape and size: evolutionary processes in
mammals. In B. Baccetti, ed., Comparative spermatology, 20
years after, 100110. Raven Press, New York.
Roldan, E. R. S., et al. 1985. Cross fertilization in vivo and in
vitro between three species of vesper mice, Calomys (Rodentia, Cricetidae). Journal of Experimental Zoology 233:
43342.
Romanovsky, A. A., and S. R. Petersen. 2003. The spleen: another mystery about its function. American Journal of PhysiologyRegulatory, Integrative and Comparative Physiology
284: R137879.
Romero-Muoz, A., and J. C. Prez-Zubieta. 2005. Evaluacin
preliminar del comercio de mamferos silvestres en el mercado La Pampa, Cochabamba, Bolivia. I Congreso Nacional de Mastozoologa en Bolivia, Cochabamba, Bolivia, 31
May3 June 2005.
Rose, K. D., and R. J. Emry. 1993. Relationships of Xenarthra,
Pholidota, and fossil edentates: the morphological evidence. In F. S. Szalay, M. J. Novacek, and M. C. McKenna,
eds., Mammal phylogeny: placentals, 81102. Springer-Verlag, New York.
Rose, K. D., et al. 2005. Xenarthra and Pholidota. In K. D. Rose
and J. D. Archibald, eds., The rise of placental mammals:
origins and relationships of the major extant clades, 10626.
Johns Hopkins University Press, Baltimore.
Ross, E. B. 1978. Food taboos, diet, and hunting strategy: the
adaptation to animals in Amazon cultural ecology. Current
Anthropology 19: 136.
Rossetti, D. de F., et al. 2004. Reconstructing habitats in central
Amazonia using megafauna, sedimentology, radiocarbon,
and isotope analyses. Quaternary Research 61: 289300.
Roth, E. E., et al. 1964. Serologic analysis of two related leptospiral serotypes isolated in Louisiana. Zoonoses Research 3:
3138.
Rouse, G. W., and S. K. Robson. 1986. An ultrastructural study
of megachiropteran (Mammalia: Chiroptera) spermatozoa:
implications for chiropteran phylogeny. Journal of Submicroscopic Cytology and Pathology 18: 13752.
Rovereto, C. 1914. Los estratos araucanos y sus fsiles. Anales
del Museo Nacional de Historia Natural de Buenos Aires 25:

1247.
Rozen, S., and H. Skaletsky. 2000. Primer3 on the WWW for
general users and for biologist programmers. Methods in
Molecular Biology 132: 36586.
Rumiz, D. I., et al. 2001. Logging and hunting in community forests and corporate concessions: two contrasting case studies
in Bolivia. In R. A. Fimbel, A. Grajal and J. G. Robinson,
eds., The cutting edge: conserving wildlife in logged tropical
forests, 33358. Columbia University Press, New York.
Rummel, R. G. 1988. Arboreal activity in a captive giant anteater (Myrmecophaga tridactyla). Animal Keepers Forum 15:
1617.
Saez, F. A., M. E. Drets, and N. Brum. 1964. Chromosomes of
the mulita (Dasypus hybridus Desmarest): a mammalian
edentate of South America. In C. Pavan, ed., Mammalian
cytogenetics and related problems in radiobiology, 16370.
Pergamon Press, New York.
Salas, R., et al. 2004. Los vertebrados de los humedales pleistocnicos de Aguada de Lomas y Sacaco (Arequipa, Per):
interpretaciones paleo ecolgicas. XII Congreso Peruano de
Geologa, Resmenes Extendidos 1: 52932.
Salazar, M. E., A. Restrepo, and D. A. Stevens. 1988. Inhibition
by estrogens of conidium-to-yeast conversion in the fungus Paracoccidioides brasiliensis. Infection and Immunity 56:
71113.
San Blas, G., G. Nino-Vega, and T. Iturriaga. 2002. Paracoccidioides brasiliensis and paracoccidioidomycosis: molecular approaches to morphogenesis, diagnosis, epidemiology,
taxonomy and genetics. Medical Mycology 40: 22542.
Sanborn, C. C. 1930. Distribution and habits of the three-banded armadillo (Tolypeutes). Journal of Mammalogy 11: 6169.
Sano, A., J. Defaveri, et al. 1999. Pathogenicity and Gp43kDa
gene of three Paracoccidioides brasiliensis isolates originated
from a nine-banded armadillo (Dasypus novemcinctus). Mycopathologia 144: 6165.
Sano, A., R. Tanaka, et al. 1999. Comparison between human
and armadillo Paracoccidioides brasiliensis by random amplified polymorphic DNA analysis. Mycopathologia 143:
16569.
Santos, I. B., et al. 1994. The rediscovery of the Brazilian three
banded armadillo and notes on its conservation status.
Edentata 1: 1115.
Santos, J. C. R., A. Rancy, and J. Ferigolo. 1993. Octodontobradyinae, uma nova subfamlia de Orophodontidae (Edentata, Tardigrada) do Mioceno Superior-Plioceno do Estado
do Amazonas, Brasil. Ameghiniana 30: 25564.
Santos, M. S. B., et al. 1998a. Monitorizaco a longo prazo da
frequncia respiratria em preguias, Bradypus variegatus.
Resumos da XIII Reunio Anual da Federao das Sociedades
de Biologia Experimental 11.045: 284.
Santos, M. S. B., et al. 1998b. Padro respiratrio em preguias
Bradypus variegatus) no anestesiadias. Resumos da XIII
Reunio Anual da Federao das Sociedades de Biologia Experimental 11.077: 384.
Saraiva, P. E. S., and B. Magalhes-Castro. 1975. Sensory and
motor representation in the cerebral cortex of the three-toed
sloth (Bradypus tridactylus). Brain Research 90: 18193.
Sarich, V. M. 1985. Xenarthran systematics: albumin immuno-
Bibliography
logical evidence. In G. G. Montgomery, The evolution and

ecology of armadillos, sloths, and vermilinguas, 7781. Smithsonian Institution Press, Washington, D.C.
Schaller, G. B. 1983. Mammals and their biomass on a Brazilian
ranch. Arquivos de Zoologia 31: 136.
Schell, P. T. 1994. Home range, activity period, burrow use, and
body temperatures of the nine- banded armadillo (Dasypus
novemcinctus) on the northern edges of its range. Masters
thesis, Southwest Missouri State University, Springfield.
Schmidt, E. E., I. C. MacDonald, and A. C. Groom. 1993. Comparative aspects of splenic microcirculatory pathways in
mammals: the region bordering the white pulp. Scanning
Microscopy 7: 61328.
Schmidt, G. D., D. W. Dusznynski, and P. S. Martin. 1992. Parasites of the extinct Shasta ground sloth, Nothrotheriops shastensis, in Rampart Cave, Arizona. Journal of Parasitology 78:
81116.
Schmidt, O. 1887. Les mammifres dans leurs rapports avec leurs
anctres gologiques. F. Alcan, Paris.
Schmitz, J., C. Roos, and H. Zischler. 2005. Primate phylogeny:
molecular evidence from retroposons. Cytogenetic and Genome Research 108: 2637.
Schmitz, J., and H. Zischler. 2003. A novel family of tRNA-derived SINEs in the colugo and two new retrotransposable
markers separating dermopterans from primates. Molecular
Phylogenetics and Evolution 28: 34149.
Schnell, C. R., and J. M. Wood. 1993. Measurement of blood
pressure and heart rate by telemetry in conscious, unrestrained marmosets. American Journal of Physiology (Heart
Circulation Physiology) 264: H150916.
Scholander, P. F., and J. Krog. 1957. Concurrent heat exchange
and vascular bundles in sloths. Journal of Applied Physiology
10: 40511.
Schubert, B. W., et al. 2004. Latest Pleistocene paleoecology of
Megalonyx jeffersonii and Cervalces scotti. Quaternary Research 61: 23140.
Schulte-Hostedde, A. I., J. S. Millar, and H. L. Gibbs. 2002.
Female-biased sexual size dimorphism in the yellow-pine
chipmunk (Tamias amoenus): sex-specific patterns of
an-nual reproductive success and survival. Evolution 56:
251929.
Scillato-Yan, G. J. 1976a. Sobre algunos restos de Mylodon(?)
listai (Edentata, Tardigrada) procedentes de la cueva Las
Buitreras (provincia de Santa Cruz, Argentina). Relaciones
de la Sociedad Argentina de Antropologa n.s. 10: 30912.
. 1976b. Sobre un Dasypodidae de edad Riochiquense
(Paleoceno superior) de Itabora (Brasil). Annais Academia
Brasileira de Cincias 48: 52730.
. 1977. Sur quelques Glyptodontidae nouveaux (Mammalia, Edentata) du Dsadien (Oligocne infrieur) de Patagonie (Argentine). Bulletin du Musum National dHistoire
Naturelle 3e sr. 487, Sciences de la Terre 64: 24962.
. 1979. El ms antiguo Megalonychinae (Edentata, Tardigrada, Megalonychidae) conocido: Holomegalonyx menendezi n. gen., n. sp., del Colhuehuapense (Oligoceno tardo)
de Gaiman (Pcia. de Chubut, Argentina). Ameghiniana 14:
31112.
. 1980. Catlogo de los Dasypodidae fsiles (Mammalia,
Edentata) de la Repblica Argentina. Actas del Segundo Con-
349
greso Argentino de Paleontologa y Biostratigrafa y Primer

Congreso Latinoamericano de Paleontologa 3: 736.
Scillato-Yan, G. J., and C. C. Carlini. 1998. Nuevos xenarthra del
Friasense (Mioceno medio) de Argentina. Studia Geolgica
Salmanticensia 34: 4367.
Scollard, D. M. 2000. Endothelial cells and the pathogenesis of
lepromatous neuritis. Insights from the armadillo model. Microbes and Infection 2: 183543.
Scott, J. A., and W. A. Untereiner. 2004. Determination of keratin
degradation by fungi using keratin azure. Medical Mycology
42: 23946.
Scott, K. 1990. Postcranial dimensions of ungulates as predictors
of body mass. In J. Damuth and B. J. MacFadden, eds., Body
size in mammalian paleobiology: estimation and biological implications, 30135. Cambridge University Press, Cambridge.
Scott, W. B. 19031904. Mammalia of the Santa Cruz Beds. Vol. V,
Paleontology. Part I, Edentata. In W. B. Scott, ed., Reports of
the Princeton University expeditions to Patagonia, 18961899,
1364. E. Schweizerbartsche Verlagshandlung (E. Ngele),
Stuttgart.
Shaw, C. A., and H. G. McDonald. 1987. First record of giant
anteater (Xenarthra, Myrmecophagidae) in North America.
Science 236: 18688.
Shaw, J. H., T. S. Carter, and J. C. Machado-Neto. 1985. Ecology
of the giant anteater Myrmecophaga tridactyla in Serra da
Canastra, Minas Gerais, Brazil: a pilot study. In G. G. Montgomery, ed., The evolution and ecology of armadillos, sloths,
Washington, D.C.
Shaw, J. H., J. Machado-Neto, and T. S. Carter. 1987. Behavior of
free-ranging anteaters (Myrmecophaga tridactyla). Biotropica
19: 25559.
Shepard, C. C. 1960. The experimental disease that follows the
injection of human leprosy bacilli into footpads of mice.
Journal of Experimental Medicine 112: 44554.
Shimodaira, H., and M. Hasegawa. 1999. Multiple comparisons
of log-likelihoods with applications to phylogenetic inference. Molecular Biology and Evolution 16: 111416.
Short, L. E. 1975. A zoogeographic analysis of the South American Chaco avifauna. Bulletin of the American Museum of
Shoshani, J. 1986. Mammalian phylogeny: comparison of morphological and molecular results. Molecular Biology and Evolution 3: 22242.
Shoshani, J., and M. C. McKenna. 1998. Higher taxonomic relationships among extant mammals based on morphology,
with selected comparisons of results from molecular data.
Sikes, R. S., G. A. Heidt, and D. A. Elrod. 1990. Seasonal diets
of the nine-banded armadillo (Dasypus novemcinctus) in a
northern part of its range. American Midland Naturalist 123:
38389.
Silla, L. M. D. R., and N. L. Stephens. 1976. Mechanical properties of diaphragm of three-toed sloth. Comparative Biochemistry and Physiology 55A: 39397.
Silva, A. V., et al. 2006. Study of Toxoplasma infection in Brazilian wild mammals: serological evidence in Dasypus novemcinctus Linnaeus, 1758 and Euphractus sexcinctus Wagler,
1830. Veterinary Parasitology 135: 8183.
350
Bibliography
Silva, E. M. 1996. Estdo do eletrocardiograma e do ritmo cardiaco na preguia, Bradypus variegatus. Masters thesis, Universidade Federal de Pernambuco, Pernambuco, Brazil.
Silva, E. M., D. P. F. Duarte, and C. P. da Costa. 2005. Electrocardiographic studies of the three-toed sloth Bradypus variegatus. Brazilian Journal of Medical and Biological Research
38: 188588.
Silva, G. E. G., M. J. F. Santiago, and M. C. C. de Albuquerque.
1975. Aspectos da contrao muscular no Bradypus tridactylus. Neurobiologia Recife 38: 7588.
Silva, L. C. L., et al. 1997. Coccidioidomicose pulmonar aguda.
Primeiro surto epidmico descrito no Cearsegundo no
Brasil. Jornal Brasileiro de Medicina 72: 4966.
Silva-Vergara, M. L., and R. Martinez. 1999. Role of the armadillo Dasypus novemcinctus in the epidemiology of paracoccidioidomycosis. Mycopathologia 144: 13133.
Silva-Vergara, M. L., et al. 2000. Isolation of Paracoccidioides
brasiliensis from armadillos (Dasypus novemcinctus) in an
area where the fungus was recently isolated from soil. Medical Mycology 38: 19399.
Silveira, E. K. P. 1968. Notas sobre a histria natural do tamandu mirim (Tamandua tetradactyla chiriquensis J. A. Allen
1904, Myrmecophagidae), com referncias fauna do Istmo
do Panam. Vellozia 6: 931.
. 1969. Histria natural do tamandu-bandeira, Myrmecophaga tridactyla Linn. 1758, Myrmecophagidae. Vellozia 7: 3443.
Silveira, L. 2004. Ecologia comparada e conservao da onapintada (Panthera onca) e ona-parda (Puma concolor), no
Cerrado e Pantanal. Ph.D. diss., Universidade de Braslia,
Brazil.
Silveira, L., et al. 1999. Impact of wildfires on the megafauna of
Emas National Park, central Brazil. Oryx 33: 10814.
Silvius, K. M., R. E. Bodmer, and J. M. V. Fragoso, eds. 2004.
People in nature: wildlife conservation in South and Central
America. Columbia University Press, New York.
Simpson, G. G. 1945. The principles of classification and a classification of mammals. Bulletin of the American Museum of
. 1980. Splendid isolation: the curious history of South
American mammals. Yale University Press, New Haven.
Singletary, J. L. 1981. Feeding habits of the nine-banded armadillo, Dasypus novemcinctus, in southern Mississippi, with
notes on its natural history. Masters thesis, University of
Southern Mississippi, Hattiesburg.
Sivinski, J. 1984. Sperm in competition. In R. L. Smith, ed.,
Sperm competition and the evolution of animal mating systems, 86115. Academic Press, Orlando.
Skinsnes, O. K. 1976. Leprosy in wild armadillos. International
Journal of Leprosy 443: 37677.
Smith, D. A. 2005. Garden game: shifting cultivation, indigenous hunting and wildlife ecology in western Panama. Human Ecology 33: 50537.
Smith, J. H., et al. 1978. Leprosy-like disease of wild armadillos
in French Acadiana, Louisiana USA. Journal of the Reticuloendothelial Society 24: 70520.
Smith, J. H., et al. 1983. Leprosy in wild armadillos Dasypus
novemcinctus of the Texas gulf coast USA: epidemiology
and mycobacteriology. Journal of the Reticuloendothelial
Society 34: 7588.
Smith, K. K., and K. H. Redford. 1990. The anatomy and function of the feeding apparatus in two armadillos (Dasypoda):
anatomy is not destiny. Journal of Zoology 222: 2747.
Smith, L. L., and R. W. Doughty. 1984. The amazing armadillo:
geography of a folk critter. University of Texas Press, Austin.
Smith, P. A. 1992. Population ecology of the nine-banded armadillo (Dasypus novemcinctus) in west central Texas. Masters
thesis, Angelo State University, San Angelo, Texas.
Smith, S. A., L. W. Robbins, and J. G. Steiert. 1998. Isolation
and characterization of a chitinase from the nine-banded
armadillo, Dasypus novemcinctus. Journal of Mammalogy
79: 48691.
Smith, T. T., and R. Yanagimachi. 1990. The viability of hamster
spermatozoa stored in the isthmus of the oviduct: the importance of sperm-epithelium contact for sperm survival.
Biology of Reproduction 42: 45057.
Soibelzon, L. 2004. Revisin sistemtica de los Tremarctinae
(Carnivora, Ursidae) fsiles de Amrica del Sur. Revista del
Museo Argentino de Ciencias Naturales 6: 10531.
Soilbelzon, E., et al. 2003. Descripcin de la dieta del pichi
llorn Chaetophractus vellerosus (Mammalia, Dasypodidae)
en la zona de Pipinas, provincia de Buenos Aires. Abstracts
XVIII Jornadas de Mastozoologa 18: 59.
Solounias, N., and S. M. Moelleken. 1993a. Dietary adaptation
of some extinct ruminants determined by premaxillary
shape. Journal of Mammalogy 74: 105971.
. 1993b. Tooth microwear and premaxillary shape of an
archaic antelope. Lethaia 26: 26168.
Solounias, N., M. Teaford, and A. Walker. 1988. Interpreting the
diet of extinct ruminants: the case of a non-browsing giraffid. Paleobiology 14: 287300.
Soma, H. 1976. In chase of trophoblast: a comparison between
human and other mammalian placentas. Journal of Tokyo
Medical College 34: 87383. (in Japanese)
Sonta, S., et al. 1980. Karyotype and sex-determining in the
two-toed sloth, Choloepus didactylus. Chromosomal Information Service 28: 1517.
Spencer, J. S., et al. 2002. Antigenic specificity of the Mycobacterium leprae homologue of ESAT-6. Infection and Immunology 70: 101013.
Springer, M. S., and W. W. de Jong. 2001. Which mammalian
supertree to bark up? Science 291: 17091711.
Springer, M. S., et al. 2003. Placental mammal diversification
and the Cretaceous-Tertiary boundary. Proceedings of the
National Academy of Sciences, U.S.A. 100: 105661.
Springer, M. S., et al. 2004. Molecules consolidate the placental
mammal tree. Trends in Ecology and Evolution 19: 43038.
Springer, M. S., et al. 2005. Molecular evidence for major placental clades. In K. D. Rose and J. D. Archibald, eds., The
rise of placental mammals, 3749. Johns Hopkins University
Press, Baltimore.
St-Andr, P.-A., and G. De Iuliis. 2001. The smallest and most
ancient representative of the genus Megatherium Cuvier,
1796 (Xenarthra, Tardigrada, Megatheriidae), from the Pliocene of the Bolivian Altiplano. Geodiversitas 23: 62545.
Staller, E. L. 2001. Identifying predators and fates of Northern
Bobwhite nests using miniature video cameras. Masters
thesis, University of Georgia, Athens.
Stallknecht, D. E., et al. 1987. Surveillance for naturally acquired
Bibliography
leprosy in a nine-banded armadillo population. Journal of

Wildlife Diseases 23: 30810.
Steadman, D. W., et al. 2005. Asynchronous extinction of late
Quaternary sloths on continents and islands. Proceedings of
Stehli F. G., and S. D. Webb, eds. 1985. The great American biotic
interchange. Plenum Press, New York.
Stevenson, H. M., and R. L. Crawford. 1974. Spread of the armadillo into the Tallahassee-Thomasville area. Florida Field
Naturalist 2: 810.
Stock, C. 1925. Cenozoic gravigrade edentates of western North
America with special reference to the Pleistocene Megalonychinae and Mylodontidae of Rancho La Brea. Carnegie
Institute of Washington 331: 1206.
Stockard, C. R. 1921. A probable explanation of polyembryony
in the armadillo. American Naturalist 55: 6268.
Storch, G. 1981. Eurotamandua joresi, ein Myrmecophagidae
aus dem Eozn der Grube Messel bei Darmstadt (Mammalia, Xenarthra). Senckenbergiana lethaia 61: 24789.
. 2003. Fossil old world edentates (Mammalia). Senckenbergiana biologica 83: 5160.
Storrs, E. E. 1971. The nine-banded armadillo: a model for leprosy and other biomedical research. International Journal of
Leprosy 39: 70314.
. 1974. Growing points in leprosy research. I. The armadillo as an experimental model for the study of human leprosy. Leprosy Review 45: 814.
Storrs, E. E., H. P. Burchfield, and R. J. W. Rees. 1988. Superdelayed parturition in armadillos: a new mammalian survival
strategy. Leprosy Review 59: 1115.
Storrs, E. E., and R. J. Williams. 1968. A study of monozygotic
quadruplet armadillos in relation to mammalian inheritance. Proceedings of the National Academy of Science, U.S.A.
60: 91014.
Storrs, E. E., et al. 1974. Leprosy in the armadillo: new model for
biomedical research. Science 183: 85152.
Strahl, H. 1913. Ueber den Bau der Plazenta von D. novemcinctus. Anatomischer Anzeiger 44: 44047.
. 1917. Beitrge zur Entwicklungsgeschichte von Tatusia
novemcincta L. Die Zeitschrift fr angewandte Anatomie und
Konstitutionslehre 2: 139.
Suga, N. 1967. Hearing in some arboreal edentates in terms of
cochlear microphonics and neural activity. Journal of Auditory Research 7: 26770.
Sunquist, M. E., and G. G. Montgomery. 1973. Activity patterns
and rates of movement of two-toed and three-toed sloths
(Choloepus hoffmanni and Bradypus infuscatus). Journal of
Mammalogy 54: 94654.
Superina, M. 2000. Biologie und Haltung von Grteltieren (Dasypodidae). Doctoral thesis, Universitt Zrich, Switzerland.
. 2005. Armadillo bibliography. <http://www.edentata.
org/biblio.htm>. Accessed 20 May 2005.
. 2006. New information on population declines in pink
fairy armadillos. Edentata 7: 4850.
Suttkus, R. D., and C. Jones. 1999. Observations on the ninebanded armadillo, Dasypus novemcinctus, in southern Louisiana. Tulane Studies in Zoology and Botany 31: 122.
Svartman, M., G. Stone, and R. Stanyon. 2006. The ancestral
eutherian karyotype is present in Xenarthra. PLoS Genetics
2: 100611.
351
Swofford, D. L. 2002. PAUP*. Phylogenetic analysis using parsimony (* and other methods). Version 4.0b10. Sinauer Associates, Sunderland, Mass.
Szalay, F. S., and F. Schrenk. 1998. The middle Eocene Eurotamandua and a Darwinian phylogenetic analysis. Kaupia
7: 97186.
Szeplaki O., E., J. Ochoa G., and J. Clavijo A. 1988. Stomach
contents of the greater long-nosed armadillo (Dasypus kappleri) in Venezuela. Mammalia 52: 42225.
Taber, A., G. Navarro, and M. A. Arribas. 1997. A new park in
the Bolivian Gran Chacoan advance in tropical dry forest
conservation and community-based management. Oryx 31:
18998.
Taber, F. W. 1945. Contribution on the life history and ecology
of the nine-banded armadillo. Journal of Mammalogy 26:
21126.
Tablin, J., K. Chamberlain, and L. Weiss. 2002. The microanatomy of the mammalian spleen: mechanisms of splenic
clearance. In A. J. Bowdler, ed., The complete spleen: structure, function and clinical disorders. 2nd ed., 1122. Humana
Press, Totowa, N.J.
Talmage, R. V., and G. D. Buchanan. 1954. The armadillo Dasypus novemcinctusa review of its natural history, ecology,
anatomy and reproductive physiology. Rice Institute Pamphlet, Houston 41: 1135.
Talmage, R. V., et al. 1954. The presence of a functional corpus
luteum during delayed implantation in the armadillo. Journal of Endocrinology 11: 4449.
Tanaka, Y. 1986. A morphological study of hematopoietic organs of the platypus Ornithorhynchus anatinus. Journal of
the Kyushu Hematological Society 34: 110.
. 1987. A morphological study of the spleen of the mole,
with special reference to its vascular architecture. Acta Haematologica Japonica 50: 93948.
. 1990. Intermediate zone of mammalian spleens: light
and electron microscopic study of three primitive mammalian species (platypus, shrew, and mole) with special reference to intrasplenic arteriovenous communication. American Journal of Anatomy 187: 31337.
Taube, E., et al. 1999. Distribution of two sympatric species of
sloths (Choloepus didactylus and Bradypus tridactylus) along
the Sinnamary River, French Guiana. Biotropica 31: 68691.
Taube, E., et al. 2001. Reproductive biology and postnatal development in sloths, Bradypus and Choloepus: review with
original data from the field (French Guiana) and from captivity. Mammal Review 31: 17388.
Tauber, A. A. 1997. Paleoecologa de la Formacin Santa Cruz
(Mioceno inferior) en el extremo sudeste de la Patagonia.
Taulman, J. F., and L. W. Robbins. 1996. Recent range expansion and distributional limits of the nine-banded armadillo
(Dasypus novemcinctus) in the United States. Journal of Biogeography 23: 63548.
Taylor, B. K. 1978. The anatomy of the forelimb in anteater (Tamandua) and its functional implications. Journal of Morphology 157: 34768.
. 1985. Functional anatomy of the forelimb in vermilinguas (anteaters). In G. G. Montgomery, The evolution
352
Bibliography
Taylor, J. W. 1995. Molecular phylogenetic classification of fungi. Archives of Medical Research 26: 30714.
Taylor, K. I. 1972. Sanuma (Yanomamo) food prohibitions: the
multiple classification of society and fauna. Ph.D. diss., University of Wisconsin, Madison.
The Chimpanzee Sequencing and Analysis Consortium. 2005.
Initial sequence of the chimpanzee genome and comparison
with the human genome. Nature 437: 6987.
The Encode Project Consortium. 2004. The ENCODE (ENCyclopedia Of DNA Elements) project. Science 306: 63640.
The Nature Conservancy, Fundacin Vida Silvestre Argentina,
Fundacin para el Desarrollo Sustentable del Chaco, and
the Wildlife Conservation Society. 2005. Evaluacin ecorregional del Gran Chaco Americano/Gran Chaco Americano
ecorregional evaluation. Fundacin Vida Silvestre, Buenos
Aires, Argentina.
Theodoro, R. C., et al. 2005. Molecular detection of Paracoccidioides brasiliensis in soil. Medical Mycology 43: 72529.
Thomas, D. A., et al. 1987. Armadillo exposure among Mexicanborn patients with lepromatous leprosy. Journal of Infectious
Diseases 156: 99093.
Thomas, W. D. 1980. Summer movement of a male armadillo in
central Texas. Texas Journal of Science 32: 36365.
Thomaz, L. D. 1996. Florstica e fitossociologia da floresta Atlantica na Estaco Biolgica de Santa Lcia, Santa Teresa,
E.S. Ph.D. diss., Universidade Estadual Paulista, Rio Claro,
So Paulo.
Thomaz, L. D., and R. Monteiro. 1997. Composio florstica
da Mata Atlntica de encosta da Estao Biolgica de Santa
Lcia, municpio de Santa Teresa, E.S. Boletim do Museu de
Biologia Prof. Mello Leito, Nova Srie 7: 348.
Thompson, E. A., and T. R. Meagher. 1987. Parental and sib
likelihoods in genealogy reconstruction. Biometrics 43:
585600.
Thompson, R. H. 1972. Algae from the hair of the sloth, Bradypus. Journal of Phycology 8 (Supplement): 8.
Thompson, R. S., et al. 1980. Shasta ground sloth (Nothrotheriops shastense Hoffstetter) at Shelter Cave, New Mexico:
environment, diet, and extinction. Quaternary Research 14:
36076.
Thorne, J. L., and H. Kishino. 2002. Divergence time and evolutionary rate estimation with multilocus data. Systematic
Biology 51: 689702.
Thorne, J. L., H. Kishino, and I. S. Painter. 1998. Estimating the
rate of evolution of the rate of molecular evolution. Molecular Biology and Evolution 15: 164757.
Tinbergen, N. 1963. On aims and methods of ethology.
Zeitschrift fr Tierpsychologie 20: 41033.
Tirira S., D., ed. 2001. Libro rojo de los mamferos del Ecuador.
Sociedad para la Investigacin y Monitoreo de la Biodiversidad Ecuatoriana (SIMBIOE)/Ecociencias/Ministerio del
Ambiente/UICN. Serie Libros Rojos del Ecuador, Tomo
1. Publicacin Especial sobre los Mamferos del Ecuador.
Quito, Ecuador.
Tirler, H. 1966. A sloth in the family. The Harvill Press, London.
Tito, G., and G. De Iuliis. 2003. Morphofunctional aspects and
paleobiology of the manus in the giant ground sloth Ere-
motherium Spillmann 1948 (Mammalia, Xenarthra, Megatheriidae). Senckenbergiana biologica 83: 7994.
Toledo, P. M. de 1996. Locomotory patterns within the Pleistocene sloths. Ph.D. diss., University of Colorado, Boulder.
Tonni, E. P. 1974. Un nuevo cariamido (Aves, Gruiformes) del
Plioceno superior de la provincia de Buenos Aires. Ameghiniana 11: 366372.
Tonni, E. P., et al. 1992. El Piso/Edad Montehermosense (Plioceno) y sus mamferos. Actas Terceras Jornadas Geolgicas
Bonaerenses 1: 113118.
Toole, J. F. 1972. Blood chemistry of the sloth. Laboratory Animal Science 22: 118121.
Toole, J. F., and J. H. Bullock. 1973. Neuromuscular responses of
sloths. Journal of Comparative Neurology 149: 25970.
Torres, C. N., H. P. Godinho, and B. P. Setchell. 1981. Frequency
and duration of the stages of the cycle of the seminiferous
epithelium of the nine-banded armadillo (Dasypus novemcinctus). Journal of Reproduction and Fertility 61: 33540.
Torrico, O., M. Viscarra, and M. da Silva. 2005. Uso de mastofauna silvestre en el carnaval de Oruro y sus implicaciones. I
Congreso Nacional de Mastozoologa en Bolivia, Cochabamba,
Bolivia, 31 May3 June 2005.
Townsend, W. R. 1995. Living on the edge: Sirion hunting and
fishing in lowland Bolivia. Ph.D. diss., University of Florida,
Gainesville.
. 1996. Nyao it: caza y pesca de los Sirion. Instituto de
Ecologa, La Paz, Bolivia.
. 2000. The sustainability of subsistence hunting by the
Sirion Indians of Bolivia. In J. G. Robinson and E. L. Bennett, eds., Hunting for sustainability in tropical forests, 267
Toyama, Y., F. U. Calderon, and R. Quesada. 1990. Ultrastructural study of crystalloids in Sertoli cells of the three-toed
sloth (Bradypus tridactylus). Cell and Tissue Research 259:
599602.
Truman, R. W. 2005. Leprosy in wild armadillos. Leprosy Review 76: 198208.
Truman, R. W., C. K. Job, and R. C. Hastings. 1990. Antibodies to the phenolic glycolipid-1 antigen for epidemiologic
investigations of enzootic leprosy in armadillos (Dasypus
novemcinctus). Leprosy Review 61: 1924.
Truman, R. W., M. J. Morales, et al. 1986. Evaluation of monitoring antibodies to PGL-1 in armadillos experimentally infected
with M. leprae. International Journal of Leprosy 54: 55659.
Truman, R. W., and R. Sanchez. 1993. Armadillos: models for
leprosy. Lab Animal 22: 2832.
Truman, R. W., E. J. Shannon, et al. 1986. Evaluation of the
origin of Mycobacterium leprae infections in the wild armadillo, Dasypus novemcinctus. American Journal of Tropical
Medicine and Hygiene 35: 58893.
Truman, R. W., et al. 1991. Seasonal and spatial trends in the
detectability of leprosy in wild armadillos. Epidemiology and
Infection 106: 54960.
Trusk, A., et al. 1992. Evaluation of tamandua diets in zoos in
North and South America. Milwaukee County Zoo, Milwaukee.
Ulloa, A., H. Rubio, and C. Campos. 1996. Trua Wuandra: estrategias para el manejo de fauna con comunidades embera
Bibliography
en el Parque Nacional Natural Utra, Choc, Colombia.

Organizacin Regional Indgena Embera Wounaan, Fundacin Natura, Ministerio de Medio Ambiente, OEI, Bogot, Colombia.
Ulrich, M., et al. 1976. Immunological characteristics of the
armadillo, Dasypus sabanicola. Clinical and Experimental
Immunology 25: 17076.
Untereiner, W. A., et al. 2004. The Ajellomycetaceae, a new
family of vertebrate-associated Onygenales. Mycologia 96:
81261.
Valencia, J. G. R. 1999. Hbitos alimenticios en poca hmeda
a partir de contenido estomacal de Dasypus novemcinctus
en el Chaco Boliviano. Field practical (pasanta), UAGRM,
Wildlife Conservation Society, Bolivia.
Valentinuzzi, M. E., E. D. V. Ruiz, and C. P. da Costa. 1984.
Ventricular fibrillation threshold in the three-toed sloth
(Bradypus tridactylus). Acta Physiologica Latino Americana
34: 31322.
Van Deelen, T. R., J. D. Parrish, and E. J. Heske. 2002. A ninebanded armadillo (Dasypus novemcinctus) from Central Illinois. Southwestern Naturalist 47: 48991.
Van Dijk, M. A. M., et al. 1999. The virtues of gaps: Xenarthran
(Edentate) monophyly supported by a unique deletion in
A-crystallin. Systematic Biology 48: 94106.
Vaughan, C., and S. Shoenfelder. 1999. Dry season activity,
movement, habitat and den utilization of nine-banded armadillo (Dasypus novemcinctus) in Neotropical dry forest,
Costa Rica. Revista de Biologa Tropical 47: 111719.
Veselovsky, Z. 1966. A contribution to the knowledge of the
reproduction and growth of the two-toed sloth Choloepus
didactylus at Prague Zoo. International Zoo Yearbook 6:
14753.
Vickers, W. T. 1984. The fauna components of lowland South
American hunting kills. Interciencia 9: 36676.
. 1991. Hunting yields and game composition over ten
years in an Amazon Indian territory. In J. G. Robinson and
K. H. Redfod, eds., Neotropical wildlife use and conservation,
5381. University of Chicago Press, Chicago.
. 1994. From opportunism to nascent conservation: the
case of the Siona-Secoya. Human Nature 5: 30737.
Villarroel, C. A. 2000. Un nuevo Mylodontinae (Xenarthra,
Tardigrada) en la fauna de La Venta, Mioceno de Colombia:
el estado actual de la familia Orophodontoidae. Revista de
la Academia Colombiana de Ciencias Exactas, Fsicas y Naturales 24: 11727.
Vizcano, S. F. 1995. Identificacin especfica de las mulitas,
gnero Dasypus L. (Mammalia, Dasypodidae), del noroeste
argentino. Descripcin de una nueva especie. Mastozoologa
Neotropical 2: 513.
. 2000. Vegetation partitioning among Lujanian (late
Pleistocene-early Holocene) armored herbivores in the
pampean region. Current Research in the Pleistocene 17:
13537.
Vizcano, S. F., and M. S. Bargo. 1998. The masticatory apparatus of the armadillo Eutatus (Mammalia, Cingulata) and
some allied genera: paleobiology and evolution. Paleobiology 24: 37183.
Vizcano, S. F., M. S. Bargo, and G. Cassini. 2006. Dental oc-
353
clusal surface area in relation to body mass, food habits and

other biologic features in fossil Xenarthrans. Ameghiniana
43: 1126.
Vizcano, S. F., M. S. Bargo, and R. F. Kay. 2004. Functional
remarks on the humerus of anteaters (Mammalia, Vermilingua) from Santa Cruz Formation (early-middle Miocene).
Ameghiniana 41 supplement: 22R.
Vizcano, S. F., M. S. Bargo, et al. 2004. Myrmecophagidae
(Mammalia, Xenarthra) de edad Santacrucense (Mioceno
temprano-medio). Ameghiniana 41 supplement: 67R.
Vizcano, S. F., and G. De Iuliis. 2003. Evidence for advanced
carnivory in fossil armadillos (Mammalia: Xenarthra: Dasypodidae). Paleobiology 29: 12338.
Vizcano, S. F., G. De Iuliis, and M. S. Bargo. 1998. Skull shape,
masticatory apparatus, and diet of Vassallia and Holmesina
(Mammalia: Xenarthra: Pampatheriidae): when anatomy
constrains destiny. Journal of Mammalian Evolution 5: 291
322.
Vizcano, S. F., and R. A. Faria. 1997. Diet and locomotion of
the armadillo Peltephilus: a new view. Lethaia 30: 7986.
Vizcano, S. F., R. A. Faria, and G. Mazzetta. 1999. Ulnar dimensions and fossoriality in armadillos and other South
American mammals. Acta Theriologica 44: 30920.
Vizcano, S. F., R. A. Faria, et al. 2004a. Functional and phylogenetic assessment of the masticatory adaptations in Cingulata (Mammalia, Xenarthra). Ameghiniana 41: 65164.
Vizcano, S. F., R. A. Faria, et al. 2004b. Palaeoecological implications of the mid-Pliocene faunal turnover in the Pampean region (Argentina). Palaeogeography, Palaeoclimatology, Palaeoecology 213: 10113.
Vizcano, S. F., and A. Giallombardo. 2001. Armadillos del Noroeste Argentino (Provincias de Jujuy y Salta). Edentata 4:
59.
Vizcano, S. F., and N. Milne. 2002. Structure and function in
armadillo limbs (Mammalia: Xenarthra: Dasypodidae).
Journal of Zoology 257: 11727.
Vizcano, S. F., N. Milne, and M. S. Bargo. 2003. Limb reconstruction of Eutatus seguini (Mammalia, Dasypodidae): paleobiological implications. Ameghiniana 40: 89101.
Vizcano, S. F., A. Rinderknecht, and A. Czerwonogora. 2003.
An enigmatic Cingulata (Mammalia: Xenarthra) from the
late Miocene of Uruguay. Journal of Vertebrate Paleontology
23: 98183.
Vizcano, S. F., and G. J. Scillato-Yan. 1995. Eocene tardigrade
(Mammalia, Xenarthra) from Seymour Island, West Antarctica. Antarctic Science 7: 4078.
Vizcano, S. F., et al. 1997. Youngest record of fossil land mammals from Antarctica: its significance on the evolution of the
terrestrial environment of the Antarctic Peninsula during
the late Eocene. Journal of Paleontology 71: 34850.
Vizcano, S. F., et al. 2001. Pleistocene burrows in the Mar del
Plata area (Buenos Aires Province, Argentina) and their
probable builders. Acta Paleontologica Polonica 46: 15769.
Vrijenhoek, R. C. 1984. Ecological differentiation among clones:
the frozen niche variation model. In K. Whrmann and V.
Loeschcke, eds., Population biology and evolution, 21731.
Springer-Verlag, Berlin.
Vrijenhoek, R. C., et al. 1989. A list of the known unisexual ver-
354
Bibliography
tebrates. In R. M. Dawley and J. P. Bogart, eds., Evolution

and ecology of unisexual vertebrates, 1923. New York State
Museum Bulletin 466, Albany.
Waage, J. K., and R. C. Best. 1985. Arthropod associates of
sloths. In G. G. Montgomery, ed., The evolution and ecology
Waddell, P. J., N. Okada, and M. Hasegawa. 1999. Towards resolving the interordinal relationships of placental mammals.
Systematic Biology 48: 15.
Waddell, P. J., and S. Shelley. 2003. Evaluating placental interordinal phylogenies with novel sequences including RAG1,
gamma-fibrinogen, ND6, and mt-tRNA, plus MCMC-driven nucleotide, amino acid, and codon models. Molecular
Phylogenetics and Evolution 28: 197224.
Walls, E. W. 1939. Myrmecophaga jubata: an embryo with placenta. Journal of Anatomy 73: 31117.
Walsh, G. P., W. M. Meyers, and C. H. Binford. 1986. Naturally
acquired leprosy in the nine-banded armadillo: a decade
of experience 19751985. Journal of Leukocyte Biology 40:
64556.
Walsh, G. P., et al. 1975. Leprosy-like disease occurring naturally in armadillos. Journal of the Reticuloendothelial Society
18: 347351.
Walsh, G. P., et al. 1977. Naturally acquired leprosy-like disease
in the nine-banded armadillo Dasypus novemcinctus: recent
epizootiologic findings. Journal of the Reticuloendothelial
Society 22: 36368.
Wampler, S. N. 1969. Husbandry and health problems of armadillos. Laboratory Animal Care 19: 39193.
Wang, M., J. Buhler, and M. R. Brent. 2003. The effects of evolutionary distance on TWINSCAN, an algorithm for pairwise comparative gene prediction. Cold Spring Harbor Symposia of Quantitative Biology 68: 12530.
Wanke, B., and A. Londero. 1994. Epidemiology and paracoccidioidomycosis infection. In M. Franco, C. Lacaz, A.
Restrepo, and G. Del Negro, eds., Paracoccidioidomycosis,
10920. CRC Press, Boca Raton, Fla.
Wanke, B., et al. 1999. Investigation of an outbreak of endemic
coccidioidomycosis in Brazils northeastern state of Piau
with review of the occurrence and distribution of Coccidioides immitis in three other Brazilian states. Mycopathologia
148: 5767.
Ward, P. I. 1998. Intraspecific variation in sperm size characters.
Heredity 80: 65559.
Weaker, F. J. 1977a. Spermatogonia and the cycle of the seminiferous epithelium in the nine-banded armadillo. Cell Tissue
Research 179: 97109.
. 1977b. The fine structure of the interstitial tissue of the
testis of the nine-banded armadillo. Anatomical Record 18:
1128.
. 1980. Morphology of the prostate gland in the ninebanded armadillo. Acta Anatomica 106: 40514.
Webb, S. D. 1978. A history of savanna vertebrates in the New
World. Part II. South America and the Great Interchange.
Annual Review of Ecology and Systematics 9: 393426.
. 1985a. The interrelationships of the tree sloths and
ground sloths. In G. G. Montgomery, ed., The evolution

. 1985b. Late Cenozoic dispersals between the Americas.
In F. G. Stehli and S. D. Webb, eds., The great American biotic
interchange, 35786. Plenum, New York.
. 1989. Osteology and relationships of Thinobadistes segnis, the first mylodont sloth in North America. In K. H. Redford and J. F. Eisenberg, eds., Advances in Neotropical mammalogy, 469532. Sandhill Crane Press, Gainesville, Fla.
Webb, S. D., and S. Perrigo. 1985. New megalonychid sloths
from El Salvador. In G. G. Montgomery, ed., The evolution
Weber, M. 2000. Effects of hunting on tropical deer populations in southeastern Mexico. Masters thesis, University of
London, U.K.
Weiser, R. S. 1975. Natural leprosy-like disease in armadillos: a
boon to leprosy research. Journal of the Reticuloendothelial
Society 18: 31516.
Weiss, C. S. 1830. Ueber das sdliche Ende des Gebirgszuges
von Brasilien in der Provinz S. Pedro do Sul und der Banda
Oriental oder Staate von Monte Video: nach den Samlungen des Herrn Fr. Sellow. Abhandlungen der physikalischen
Klasse der Kniglichen Akademie der Wissenschaften zu Berlin 217293, plates 15.
Weiss, L. 1989. New trends in spleen research: reticuloendothelial basis of the clearance of blood by the spleen. In C.
Pochedly, R. H. Sills, and A. D. Schwartz, eds., Disorders of
the spleen: pathophysiology and management, 43153. Marcel
Dekker, New York.
Weiss, L. P., and G. B. Wislocki. 1956. Seasonal variations in
hematopoiesis in the dermal bones of the nine-banded armadillo. Anatomical Record 126: 14363.
Weitzman, I., and R. C. Summerbell. 1995. The dermatophytes.
Clinical Microbiology Reviews 8: 24059.
Wenker, C. J., et al. 1998. Sporotrichosis in a nine-banded armadillo (Dasypus novemcinctus). Journal of Zoo and Wildlife
Medicine 29: 47478.
West, B. C., et al. 1988. Leprosy in six isolated residents of northern Louisiana: time-clustered cases in an essentially nonendemic area. Archives of Internal Medicine 148: 198792.
Wetzel, R. M. 1975. The species of Tamandua Gray (Edentata,
Myrmecophagidae). Proceedings of the Biological Society of
Washington 88: 95112.
. 1982. Systematics, distribution, ecology and conservation of South American edentates. In M. A. Mares and H.
H. Genoways, eds., Mammalian biology in South America,
34575. Special Publication Series, Vol. 6. Pymatuning Laboratory of Ecology, University of Pittsburgh, Pa.
. 1985a. The identification and distribution of recent Xenarthra (= Edentata). In G. G. Montgomery, ed., The evolution and ecology of armadillos, sloths, and vermilinguas, 521.
. 1985b. Taxonomy and distribution of armadillos. In G.
G. Montgomery, ed., The evolution and ecology of armadillos, sloths, and vermilinguas, 2346. Smithsonian Institution
Wetzel, R. M., and F. D. d. Avila-Pires. 1980. Identification and
Bibliography
distribution of the recent sloths of Brazil (Edentata). Revista

Brasileira de Biologia 40: 83136.
Wetzel, R. M., and E. Mondolfi. 1979. The subgenera and species of long-nosed armadillos, genus Dasypus L. In J. F.
Eisenberg, ed., Vertebrate ecology in the northern Neotropics,
White, J. L. 1993. Indicators of locomotor habits in xenarthrans:
evidence for locomotor heterogeneity among fossil sloths.
Journal of Vertebrate Paleontology 13: 23042.
. 1997. Locomotor adaptations in Miocene xenarthrans.
In R. F. Kay, R. H. Madden, R. L. Cifelli, and J. J. Flynn, eds.,
Vertebrate paleontology in the neotropics: the Miocene fauna
of La Venta, Colombia, 24664. Smithsonian Institution
White, J. L., and R. D. E. MacPhee. 2001. The sloths of the
West Indies: a systematic and phylogenetic review. In C. A.
Woods and F. E. Sergile, eds., Biogeography of the West Indies: patterns and perspectives. 2nd ed., 20135. CRC Press,
New York.
White, L. L. 1992. Seasonal dietary habits of the nine-banded
armadillo in southern Alabama and the surrounding area.
Masters thesis, University of South Alabama, Mobile.
Wible, J. R., and T. J. Gaudin. 2004. On the cranial osteology of
the yellow armadillo Euphractus sexcinctus (Dasypodidae,
Xenarthra, Placentalia). Annals of the Carnegie Museum of
Wilson, E. D., et al. 2003. Taurine deficiency associated dilated
cardiomyopathy in giant anteaters (Myrmecophaga tridactyla): preliminary results and diagnostics. Abstract, Proceedings of the American Association of Zoo Veterinarians,
15559.
Wilson, G. T., et al. 1984. Absence of leprosy-like disease in
the nine-banded armadillo Dasypus novemcinctus in and
around Taylor County Texas USA. Texas Journal of Science
36: 7379.
Wimsatt, W. A., A. C. Enders, and H. W. Mossman. 1973. A
re-examination of the chorio-allantoic placental membrane
of a shrew, Blarina brevicauda: resolution of a controversy.
American Journal of Anatomy 138: 20733.
Winge, H. 1915. Jordfunde og nulevende gumlere (Edentata) fra
Lagoa santa, Minas Geraes, Brasilien. E Museo Lundii 3 II.
Wingender, E., et al. 2001. The TRANSFAC system on gene expression regulation. Nucleic Acids Research 29: 28183.
Wirtz, W. O., D. H. Austin, and G. W. Dekle. 1985. Food habits
of the common long-nosed armadillo Dasypus novemcinctus in Florida, 19601961. In G. G. Montgomery, ed., The
Wislocki, G. B. 1925. On the placentation of the sloth (Bradypus
griseus). Contributions to Embryology, Carnegie Institution
of Washington 16: 719.
. 1926. Further observations upon the placentation of the
sloth (Bradypus griseus). Anatomical Record 32: 4551.
. 1927. On the placentation of the tridactyl sloth (Bradypus griseus) with a description of some characters of the
fetus. Contributions to Embryology, Carnegie Institution of
Washington 19: 21128.
. 1928a. Observations on the gross and microscopic anat-
355
omy of the sloths (Bradypus griseus griseus Gray and Choloepus hoffmanni Peters). Journal of Morphology 46: 31798.
. 1928b. On the placentation of the two-toed anteater (Cyclopes didactylus). Anatomical Record 39: 6983.
. 1928c. Further observations upon the minute structure
of the labyrinth in the placenta of the sloths. Anatomical Record 40: 38595.
Wolda, H. 1985. Seasonal distribution of sloth moths Cryptoses
choloepi Dyar (Pyralidae; Chrysauginae) in light traps in
Panama. In G. G. Montgomery, ed., The evolution and ecology of armadillos, sloths and vermilinguas, 31318. Smithsonian Institution Press, Washington, D.C.
Woods, J. P. 2003. Knocking on the right door and making a
comfortable home: Histoplasma capsulatum intracellular
pathogenesis. Current Opinion in Microbiology 6: 32731.
Wooley, D. M. 1979. Interpretations of the pattern of sperm tail
movements. In D. W. Fawcett and J. M. Bedford, eds., The
spermatozoon, 6979. Urban and Schwarzenberg, Baltimore
and Munich.
World Health Organization. 2005. Global leprosy situation,
2004. Weekly Epidemiology Record 80: 11824.
World Health Organization, Immunology of Mycobacterial
Diseases (IMMYC) Steering Committee. 1995. Analysis of
vaccines prepared from armadillo-derived M. leprae: results of an inter-laboratory study coordinated by the World
Health Organization. International Journal of Leprosy 63:
4855.
Wujek, D. E., and J. M. Cocuzza. 1986. Morphology of hair of
two- and three-toed sloths (Edentata: Bradypodidae). Revista de Biologa Tropical 34: 24346.
Wyss, A. R., et al. 1990. A new Tertiary mammal fauna from
central Chile: implications for Andean stratigraphy and tectonics. Journal of Vertebrate Paleontology 10: 51822.
Wyss, A. R., et al. 1993. South Americas earliest rodent and recognition of a new interval of mammalian evolution. Nature
365: 43437.
Xie, X., et al. 2005. Systematic discovery of regulatory motifs
in human promoters and 3 UTRs by comparison of several
mammals. Nature 434: 33845.
Yaeger, R. G. 1988. The prevalence of Trypanosoma cruzi infection in armadillos collected at a site near New Orleans,
Louisiana. American Journal of Tropical Medicine and Hygiene 38: 32326.
Yeo, M., et al. 2005. Origins of Chagas disease: Didelphis species are natural hosts of Trypanosoma cruzi I and armadillos
hosts of Trypanosoma cruzi II, including hybrids. International Journal for Parasitology 35: 22533.
Young, R. J., C. M. Coelho, and D. R. Wieloch. 2003. A note on
the climbing abilities of giant anteaters, Myrmecophaga tridactyla (Xenarthra, Myrmecophagidae). Boletim do Museu
de Biologia Mello Leito 15: 4146.
Zapata, A. G., and E. L. Cooper. 1990. The immune system: comparative histophysiology. John Wiley and Sons, Chichester,
England.
Zapata Ros, G. 2001. Sustentabilidad de la cacera de subsistencia: el caso de cuatro comunidades quichuas en la Amazona nororiental ecuatoriana. Mastozoologa Neotropical 8:
5966.
356
Bibliography
Zrate, M. A., et al. 1998. Estructuras biognicas en el Cenozoico tardo de Mar del Plata (Argentina) atribuibles a grandes
mamferos. Revista de la Sociedad Argentina de Sedimentologa 5: 95103.
Zeledon, R., C. de Ponce, and E. Ponce. 1975. Isolation of Leishmania herreri sp. n. from sloths in Costa Rica. American
Journal of Tropical Medicine and Hygiene 24: 7067.
Zeller, U., J. R. Wible, and M. Elsner. 1993. New ontogenetic
evidence on the septomaxilla of Tamandua and Choloepus
(Mammalia, Xenarthra), with a reevaluation of the homology of the mammalian septomaxilla. Journal of Mammalian
Evolution 1: 3146.
Zhao, F., et al. 2005. TRED: a Transcriptional Regulatory Element Database and a platform for in silico gene regulation
studies. Nucleic Acids Research 33: D1037.
Zimmerman, J. W. 1982. The common long-nosed armadillo
(Dasypus novemcinctus) in northcentral Oklahoma. Masters thesis, Oklahoma State University, Stillwater.
. 1990. Burrow characteristics of the nine-banded armadillo, Dasypus novemcinctus. Southwestern Naturalist 35:
22627.
Zuleta, G., and M. L. Bolkovic. 1994. Conservation ecology of
armadillos in the Chaco region of Argentina. Edentata 1:
1617.
Contributors
Emma B. Casanave is professor of animal physiology

in the Departamento de Biologa, Bioqumica y Farmacia, Universidad Nacional del Sur, San Juan 670, 8000
Baha Blanca, Argentina, and a researcher with the
Consejo Nacional de Investigaciones Cientficas y Tcnicas (CONICET).
Agustn M. Abba is a researcher with the Vertebrate Division of the Facultad de Ciencias Naturales y Museo de
La Plata, Universidad Nacional de La Plata, Paseo del
Bosque s/n, B1900FWA La Plata, Argentina.
James E. Adams is a post-doctoral research associate
with the GWL Hansens Disease Program, USPHS,
Louisiana State University, Baton Rouge, LA 70803.
Marcelo H. Cassini is professor of biology in the Departamento de Ciencias Bsicas, Universidad Nacional de
Lujn, Rutas 5 y 7, 6700 Lujn, Argentina, and senior
researcher for the Argentinean Research Council .
John M. Aguiar is coordinator for the IUCN/SSC Primate and Edentate Specialist Groups, CABS-Conservation International, 2011 Crystal Drive, Suite 500, Arlington, Virginia 22202.
Pablo D. Cetica is professor of biochemistry, Facultad

de Ciencias Veterinarias, Universidad de Buenos Aires,
Av. Chorroarn 280, C1427CWO Buenos Aires, Argentina.
Eduardo Bagagli is associate professor in the Departmento de Microbiologia e Imunologia, Instituto de

Biociencias, Universidade Estadual PaulistaUNESP,
Botucatu, SP, Brazil, 18618-000.
Jean L. Chang is a systems administrator at the Broad

Institute of MIT and Harvard, 7 Cambridge Center,
Cambridge, MA 02142.
M. Susana Bargo is associate researcher in the Divisin

Paleontologa de Vertebrados, Museo de La Plata, Paseo
del Bosque s/n, B1900FWA La Plata, Argentina.
Kurt Benirschke is professor emeritus of Pathology and
Reproductive Medicine at the University of California,
San Diego, CA 92103.
Adriano Garcia Chiarello is professor in the Programa

de Mestrado em Zoologia de Vertebrados, Pontifcia
Universidade Catlica de Minas Gerais, Av. Dom Jos
Gaspar 500, Prdio 41, Belo Horizonte, MG, 30.535-610,
Brazil.
Patrice Boily is assistant professor in the Department

of Biology and Environmental Science, Western Connecticut State University, 181 White Street, Danbury, CT
06810.
Erika Cullar is a Ph.D. student at Oxford University

and affiliated with the Wildlife Conservation SocietyBolivia, calle Bumberque #349, Casilla 6272, Santa
Cruz, Bolivia, 591-3-332-8681.
Sandra de Moraes Gimenes Bosco is assistant professor

in the Departamento de Microbiologia e Imunologia,
Instituto de Biocincias, Universidade Estadual PaulistaUNESP, Botucatu, SP, Brazil, 18618-000.
Rosa Leny Cullar is a biologist in community wildlife management with the Wildlife Conservation Society-Bolivia, calle Bumberque #349, Casilla 6272, Santa
Cruz, Bolivia, 591-3-332-8681 and a Ph.D. student in
the Wildlife and Fisheries Sciences Department, Texas
A&M University, College Station, TX 77843-2258.
Constana Camilo-Alves is an ecologist at Fazenda

Bela Vista, Estrada Parque km 26, CP 18, Corumb, MS,
CEP 79301-970, Brazil.
Gerardo De Iuliis is professor in the Faculty of Health

Sciences, George Brown College, Toronto, ON, Canada
M5A 1J5.
357
358
Contributors
Frdric Delsuc is a CNRS researcher in evolutionary biology at the Institut des Sciences de lEvolution
(UMR5554-CNRS), Universit Montpellier II, 34095
Montpellier Cedex, France.
Wilham Jorge is professor of the genetics post-graduate course, Departmento de Biologa Geral, Instituto
de Ciencias Biolgicas, Universidade Federal de Minas
Gerais, Belo Horizonte, Brazil.
Emmanuel J. P. Douzery is professor of evolutionary biology at the Institut des Sciences de lEvolution
(UMR5554-CNRS), Universit Montpellier II, 34095
Montpellier Cedex, France.
W. J. Loughry is professor in the Department of Biology,

Valdosta State University, Valdosta, GA 31698-0015.
Denia Fittipaldi Duarte is associate professor in the Departamento de Fisiologia e Farmacologia, Universidade
Federal de Pernambuco, Cidade Universitria, Recife
PE 50.670-901, Brazil.
Allen C. Enders is professor emeritus of Cell Biology
and Human Anatomy, University of California, Davis,
CA 95616.
Richard A. Faria is assistant professor in the Departamento de Paleontologia, Facultad de Ciencias, Universidad de la Republica, Igua 4225, 11400 Montevideo,
Uruguay.
Juan Carlos Fernicola is associate researcher in the Seccin Paleontologa de Vertebrados, Museo Argentino
de Ciencias Naturales Bernardino Rivadavia, Av. Angel Gallardo 470, C1405DJR, Buenos Aires, Argentina.
Gustavo A. B. da Fonseca is professor of zoology at the
Federal University of Minas Gerais, Brazil, and Team
Leader, Natural Resources, Global Environment Facility, 1818 H Street NW, G 6-602, Washington DC 20433.
Elena J. Galndez is chief of lab practice, Laboratorio de
Histologa Animal, Universidad Nacional del Sur, Departamento de Biologa, Bioqumica y Farmacia, San
Juan 670, 8000 Baha Blanca, Argentina.
Timothy J. Gaudin is UC Foundation professor, Department of Biological and Environmental Sciences,
University of Tennessee at Chattanooga, 615 McCallie
Ave., Chattanooga, TN 37403-2598.
Desmond Gilmore is senior lecturer in physiology,
Room 133 Thomson Building, Laboratory of Human
Anatomy, University of Glasgow G12 8QQ, Scotland.
H. Gregory McDonald is senior curator of natural history for the National Park Service, Fort Collins, CO
80521.
Colleen M. McDonough is professor in the Department of Biology, Valdosta State University, Valdosta,
GA 31698-0015.
sis M. Medri is a Ph.D. student in the Programa de
Ps-Graduao em Ecologia, Departamento de Ecologia, Instituto de Cincias Biolgicas, Universidade de
Braslia, Asa Norte, Braslia, DF, CEP 70910-900.
Mara Susana Merani is a CONICET principal investigator at the Centro de Investigaciones en Reproduccin, Facultad de Medicina, Universidad de Buenos
Aires, Paraguay 2155 10, C1121ABG Buenos Aires, Argentina.
Dennis A. Meritt Jr. is adjunct professor in the Department of Biological Sciences, DePaul University, Chicago, IL 60614.
Flvia Miranda is director of Projeto Tamandu, Av.
Agua Fria 269, apt. 133B, Barrio Agua Fria, So Paulo
SP, CEP 02333-000, Brazil.
Guilherme H. B. de Miranda is a federal forensics expert at the Instituto Nacional de Criminalstica, Diretoria Tcnico-Cientfica, Departamento de Polcia
Federal, SAIS Q. 7, Lt. 23, Sala A113, Braslia, DF, CEP
70610-200, Brazil.
Guilherme Mouro is a wildlife ecologist with Embrapa
Pantanal, Corumb, MS, Brazil 79320-900.
Andrew J. Noss is project director with the Wildlife
Conservation Society in Ecuador, CP 17-21-168, Quito,
Ecuador.
Contributors
Richard D. Peppler is associate dean for faculty and academic affairs, College of Medicine, University of Central Florida, 12201 Research ParkwaySuite 338, P.O.
Box 160116, Orlando, FL 32816.
Helio Rubens Jacintho Pereira Jr. is adjunct professor at
the Centro Universitrio Nossa Senhora do Patrocinio,
Salto/SP 13320-902, Brazil.
Carlos Peres da Costa is professor of physiology in the
Departamento de Fisiologia e Farmacologia, Universidade Federal de Pernambuco, Cidade Universitria,
Recife PE 50.670-901, Brazil.
Tinka Plese is director of the Fundacin UNAU, Cir. 1a
# 73-24, Medelln, Colombia.
Paulo A. Prodhl is reader in population and evolutionary genetics in the School of Biological Sciences at
Queens University Belfast, Belfast BT9 1NN, Northern
Ireland, U.K.
Franois Pujos is associate researcher with the Departamento de Geologa y Paleontologa, IANIGLACRICYT-CONICET, Avda. Ruiz Leal s/n, Parque Gral.
San Martn, 5500 Mendoza, Argentina and is affiliated
359
with the Institut Franais dEtudes Andines, Casilla 181217, Av. Arequipa 4595, Lima 18, Peru.
Flvio H. G. Rodrigues is professor of ecology at the
Departamento de Biologia Geral, Instituto de Cincias
Biolgicas, Universidade Federal de Minas Gerais, Av.
Antnio Carlos 6627, CP 486, Pampulha, Belo Horizonte, MG, CEP 31270-901, Brazil and a researcher with
the Instituto para Conservao dos Carnvoros NeotropicaisPr-Carnivoros.
Mariella Superina is a post-doctoral researcher at the
Laboratorio de Reproduccin y Lactancia, IMBECU,
CRICYT-CONICET, Casilla de Correos 855, 5500
Mendoza, Argentina, and deputy chair of the IUCN
Edentate Specialist Group.
Richard W. Truman is director of the GWL Hansens
Disease Program, USPHS, Louisiana State University,
Baton Rouge, LA 70803.
Sergio F. Vizcano is professor of vertebrate zoology at
the Universidad Nacional de La Plata and associate researcher for CONICET in the Divisin Paleontologa
de Vertebrados, Museo de La Plata, Paseo del Bosque
s/n, B1900FWA La Plata, Argentina.
Taxonomic Index
Acantholippia seriphioides, 52
Acinetobacter hinshawii, 104
Acratocnus, 31, 43, 5153
Acratocnus odontrigonus, 54
Acremylodon, 41, 43, 48
Adiastaltus habilis, 70, 7778
Adiastaltus procerus, 70, 78
Afrotheria, 1315, 20, 27, 88
Agamofilaria oxyura, 52
Agouti paca, 109
Ajellomyces, 105, 109
Ajellomyces capsulatus, 109
Ajellomyces dermatitidis, 109
Amerhippus, 61, 63
Amphibradys, 44
Amphiocnus, 43
Anacardium excelsum, 273
Analcimorphus, 44
Analcitherium, 44, 46
Anathitus revelator, 70, 78
Ancylostoma, 104
Apiales, 52
Apis mellifera, 260
Apocinaceae, 273
Archeococcidia, 52
Argyromanis patagonica, 70, 78
Artiodactyla, 125
Ascaris, 104
Ascobolus, 109
Blastomyces dermatitidis, 105
Boreoeutheria, 13, 20
Bradypodidae, 11, 16, 19697, 270
Bradypus, 16, 19, 2425, 3031, 33, 39, 40,
43, 99, 131, 13336, 13840, 16667,
197200, 218, 230, 23943, 26971, 27476,
27879
Bradypus griseus, 16869
Bradypus pygmaeus, 16, 130, 198, 21519, 270,
274
Bradypus torquatus, 16, 22, 130, 14041, 168,
19899, 210, 21519, 27079
Bradypus tridactylus, 16, 22, 90, 130, 13234,
136, 14041, 16871, 192, 19899, 21718,
27073, 27678
Bradypus variegatus, 16, 104, 130, 132, 13541,
16869, 19899, 210, 21719, 240, 242, 250,
253, 27078
Brunellia sibundoya, 241
Burmeisteria, 21
Cabassous, 1619, 143, 145, 148, 150, 197, 2023,

224, 226, 234, 28283, 286, 296
Cabassous centralis, 107, 202, 217, 22526,
253, 282
Cabassous chacoensis, 202, 217, 22526, 253,
272, 282, 296
Cabassous tatouay, 197, 202, 217, 22526,
28283, 292
Cabassous unicinctus, 145, 192, 197, 202,
22526, 250, 253, 28283, 28788, 292
Calyptophractus, 17, 21, 225, 297
Carnivora, 12, 8687
Catagonus wagneri, 247, 294
Catonyx, 43, 50
Catonyx chiliensis, 54
Cavia porcellus, 183
Cecropia, 241, 273
Chaetophractus, 1619, 29, 143, 145, 148, 150,
201, 22324, 248, 286
Chaetophractus nationi, 202, 215, 217, 223, 244,
24748, 282
Chaetophractus vellerosus, 12021, 123, 125,
145, 147, 217, 223, 235, 247, 250, 25253,
28184, 286, 288, 296, 298, 300312, 314,
318
Chaetophractus villosus, 12021, 123, 145, 148,
180, 195, 2012, 217, 22324, 23334, 236,
247, 250, 25253, 28284, 288, 290, 292,
296, 300312, 314
Chasicobradys, 44
Chlamyphorus, 12, 14, 17, 21, 27, 202, 227, 230,
286
Chlamyphorus retusus, 17, 21, 202, 217, 225,
227, 235, 246, 282, 297, 3067
Chlamyphorus truncatus, 21, 91, 197, 202, 217,
22527, 233, 235, 283
Choloepidae, 270
Choloepus, 16, 19, 2425, 3033, 3940, 43,
13136, 13839, 197, 200201, 232, 23943,
26972, 27576, 27880
Choloepus didactylus, 16, 22, 104, 109, 130,
133, 136, 138, 14042, 16568, 197, 200, 203,
21718, 221, 232, 242, 253, 270, 277
Choloepus hoffmanni, 16, 130, 13233, 135,
14042, 16567, 197, 200, 21718, 232, 242,
27071, 27475, 27778
Chondrichthyes, 120
Chrysauginae, 131
Chrysocyon brachyurus, 299
Chubutherium, 44, 46
361
Chuquiraga ruscifolia, 52
Cingulata, 1, 11, 14, 16, 18, 2427, 29, 32, 34, 73,
79, 8082, 88
Clitoria, 273
Clypeotherium, 83
Coccidioides immitis, 105, 110
Cochlosperma, 273
Coleoptera, 150, 28586, 304
Collembola, 286
Copernicia tectorum, 264
Cornitermes cumulans, 259
Coussapoa, 27374
Cryptococcus neoformans, 105
Ctenomys, 148, 253
Cuvieronus hyodon, 62
Cyclopedidae, 11, 28, 6465, 72, 75, 196, 202,
258
Cyclopes, 16, 1920, 28, 6571, 7475, 202,
220, 263
Cyclopes didactylus, 22, 67, 73, 161, 164, 192,
197, 203, 217, 219, 221, 238, 250, 253, 25758
Dasypodidae, 2, 11, 16, 19, 88, 108, 120, 14344,
14950, 160, 184, 19697, 201
Dasypodinae, 17, 2012
Dasyprocta variegata, 247
Dasypus, 2, 56, 1617, 1921, 2829, 8081,
83, 12629, 143, 145, 148, 17273, 17980,
19293, 197, 206, 208, 22729, 287, 29192
Dasypus hybridus, 107, 109, 12023, 125, 127,
129, 145, 173, 19697, 2012, 217, 22829,
283, 300305
Dasypus kappleri, 20, 127, 202, 217, 22829,
250, 253, 284
Dasypus novemcinctus, 1, 35, 20, 22, 1034,
1078, 11011, 125, 12728, 145, 151, 161,
17273, 17981, 18385, 18788, 190, 19294,
19697, 2012, 20510, 217, 227, 22930,
233, 23536, 247, 250, 25354, 28184,
28693, 297, 30612, 318
Dasypus pilosus, 6, 202, 215, 217, 22829, 282
Dasypus sabanicola, 202, 217, 22829, 283,
285, 28789, 293
Dasypus septemcinctus, 110, 145, 197, 2012,
217, 22829, 253, 282
Dasypus yepesi, 202, 21617, 22829, 253, 282
Deseadognathus, 4445
Dideilotherium, 65
Didelphis marsupialis, 109
Diheterocnus, 43, 48
362
Taxonomic Index
Diptera, 285
Doedicuridae, 79, 8384
Doedicurinae, 27, 79, 8284
Doedicurini, 27
Doedicurus, 25, 7980, 82, 84
Doedicurus clavicaudatus, 91, 95
Echinops telfairi, 183
Edentata, 12, 216, 231
Eleutherocercus, 27
Emmonsia, 105
Entelops, 33
Enterobacter aerogenes, 104
Eomanis, 30
Eomegatherium, 47
Eomegatherium nanum, 47
Eosclerocalyptus, 83
Ephedra breana, 52
Epitheria, 1315, 20, 26, 88
Equus, 61
Eremotherium, 31, 43, 47, 4951, 5659, 61,
63, 97
Eremotherium eomigrans, 59
Eremotherium laurillardi, 40, 5154, 56, 59, 61,
63, 93, 97
Eremotherium sefvei, 59, 61, 63
Erinaceus europaeus, 183
Escherichia coli, 104
Euarchontoglires, 13, 15, 27, 88
Eucholoeops, 30, 44, 4647
Eucinepeltus, 82
Euphorbiacea, 273
Euphractinae, 11, 1722, 28, 2012, 286, 314
Euphractus, 1619, 120, 143, 145, 148, 197, 201,
224
Euphractus sexcinctus, 22, 104, 145, 147, 197,
2012, 217, 224, 247, 250, 253, 264, 28283,
28586, 28890, 292, 297, 30612, 314
Eurotamandua, 6, 2425, 28, 30, 32, 65
Eutardigrada, 30
Eutatus, 8283
Eutatus seguini, 92, 95
Fabales, 52
Fabiana peckii, 51
Felis catus, 183
Felis concolor, 296
Ficus, 264, 27273
Ficus trigonata, 273
Folivora, 2, 11, 14, 16, 2426, 31, 40, 57, 88
Fruitafossor, 64
Glires, 13, 27
Globigerina ampliapertura, 45
Glossotheriopsis, 41, 44, 48
Glossotherium, 32, 43, 4850, 63, 98
Glossotherium robustum, 91, 94, 96
Glyptatelinae, 27, 79, 83, 85
Glyptatelus, 80, 83
Glyptodon, 8384, 98
Glyptodon clavipes, 91, 93

Glyptodon reticulatus, 91
Glyptodontidae, 2, 27, 79, 83, 88, 91
Glyptodontinae, 2728, 63, 79, 8284
Glyptodontini, 28
Glyptodontoidea, 81
Glyptotheriini, 28
Glyptotherium, 28
Guatteria, 273
Habanocnus, 43
Hapaloides, 44
Hapalops, 31, 33, 40, 44, 4647
Heydyosmium goudotianum, 241
Hippidion, 63
Hippocrateaceae, 273
Histoplasma capsulatum, 105, 109
Holmesina, 29
Holmesina cf. paulacoutoi, 63
Holomegalonyx, 44, 46
Homo sapiens, 181
Hoplophoridae, 79, 8384
Hoplophorinae, 2728, 63, 79, 8384
Hoplophorini, 2728
Hoplophractus, 27, 83
Huilabradys, 44
Hymenoptera, 28586
Hyperlleptus, 44
Imagocnus, 44, 48
Imagocnus zazae, 48
Insectivora, 123
Isoodon obesulus, 200
Isoptera, 28586
Junellia seriphioides, 52
Lacmellea, 273
Lamiales, 52
Laurasiatheria, 13, 15, 27, 65
Leishmania, 109
Leptospira interrogans, 104
Lestodon, 32, 43, 50
Lestodon armatus, 91, 94, 96
Lipotyphla, 27
Litopterna, 56
Lomaphorini, 2728, 83
Loxodonta africana, 183
Luehea cymulosa, 27172
Macroeuphractus, 94
Macroeuphractus outesi, 95
Macropus eugenii, 20
Malpighiales, 52
Mandevilla, 273
Manis, 27, 258
Manis pentadactyla, 183
Mazama gouazoubira, 247
Megabradys, 32, 43, 48
Megalocninae, 32, 48
Megalocnus, 32, 41, 43, 5152

Megalocnus rodens, 5354
Megalonychidae, 4546, 57, 88, 196, 270
Megalonychops, 43
Megalonychotherium, 44, 46
Megalonyx, 41, 43, 4849, 51, 53
Megalonyx jeffersonii, 52
Megalonyx obtusidens, 49
Megatheria, 2425, 30
Megathericulus, 41, 44, 47, 57, 62
Megatheriidae, 2, 16, 2425, 3031, 46, 57, 88
Megatheriinae, 5659, 6163, 92
Megatheriodea, 24
Megatheriops, 47, 57
Megatherium, 5759, 6163, 80, 8688, 9798
Megatherium (Megatherium) altiplanicum,
59, 6263
Megatherium (Megatherium) americanum,
5960, 62
Megatherium (Pseudomegatherium) celendinense, 59, 62
Megatherium (Pseudomegatherium) elenense,
59, 6162
Megatherium (Pseudomegatherium) medinae,
59, 62
Megatherium (Pseudomegatherium) tarijense,
5960, 62
Megatherium (Pseudomegatherium) urbinai,
59, 62
Megatherium americanum, 5659, 63, 80,
8688, 9093, 9597, 99
Megatherium antiquum, 47
Megatherium sundti, 59, 61
Meizonyx, 43, 49
Miconia tonduzii, 241
Microcnus, 32
Microphilus venulosa, 274
Miocnus, 43
Monodelphis domestica, 20
Moraceae, 27273
Mus musculus, 181
Mycobacterium leprae, 4, 104, 108, 111, 11319,
127, 192
Mylodon, 32, 50, 5354
Mylodon darwini, 16, 5254, 91, 9598
Mylodontidae, 2, 16, 2425, 3031, 41, 46,
57, 88
Mylodontinae, 32, 63
Mylodontopsis, 43
Myotis lucifugus, 183
Myrmecobius fasciatus, 64, 258
Myrmecophaga, 16, 19, 25, 28, 57, 6566,
6872, 7475, 202, 220, 230, 261
Myrmecophaga caroloameghinoi, 73
Myrmecophaga tridactyla, 22, 67, 73, 7576,
91, 161, 203, 210, 217, 22021, 232, 237, 239,
244, 247, 250, 253, 25758, 262, 295
Myrmecophagidae, 11, 28, 6465, 70, 72, 75,
77, 19697, 202, 204, 258
Myrtaceae, 274
Taxonomic Index
Nasutitermes, 260
Nematherium, 44, 46
Neocnus, 43, 53
Neocnus comes, 54
Neocnus dousman, 54
Neoglyptatelus, 83
Neohapalops, 43
Neomesocnus, 43
Neonematherium, 44
Neosclerocalyptus, 7980, 8385
Neosclerocalyptus ornatus, 91
Neotamandua, 28, 7072, 7576
Neotamandua borealis, 73, 76
Neotamandua conspicua, 67, 73, 75
Neotamandua greslebini, 73, 76
Neotamandua magna, 71, 76
Neotamandua? australis, 70, 73, 75
Neothoracophorini, 2728
Neuryurini, 2728
Nocardia brasiliensis, 104
Nothofagus, 42
Nothropus, 43
Nothrotheriidae, 2, 2425, 3031, 33, 4647,
57, 88
Nothrotheriops, 3032, 41, 43, 5154
Nothrotheriops shastensis, 16, 5253
Nothrotherium, 32, 41, 43, 50
Nothrotherium maquinense, 54
Notongulata, 56
Nunezia caroloameghinoi, 71, 73, 76
Nycticebus tardigradus, 137
Ocnopus, 43
Ocnotherium, 43
Ocotea aciphyla, 274
Octodontobradys, 44
Octodontotherium, 42, 4446, 82
Octomylodon, 44
Onohippidion, 63
Onygena, 109
Onygenales, 103, 105, 10910
Oreomylodon, 43
Ornithorhynchus anatinus, 20
Orophodon, 45, 8182
Orthoarthrus mixtus, 70, 78
Ortotherinae, 32
Orycteropus afer, 258
Oryctolagus cuniculus, 183
Otolemur garnettii, 183
Pachyarmatherium, 27, 83
Paenungulata, 13
Palaeanodonta, 26, 30, 3233
Palaeohoplophorini, 27
Palaeomyrmidon, 28, 71, 74, 76
Palaeomyrmidon incomptus, 67, 71, 73, 76
Palaeopeltis, 8182
Paleolama, 61
Pampatheriidae, 2, 63, 88
Pampatherium, 61, 8183
Panochthini, 2728, 83
Panochthus, 63, 7981, 8385
Panochthus tuberculatus, 91
Panthera onca, 262, 296, 308
Paracoccidioides brasiliensis, 10310
Paralouatta marianae, 48
Paramylodon, 3132, 43, 51, 98
Paramylodon harlani, 5254, 97
Paranabradys, 43
Parapelecyodon, 44
Parocnus, 43
Parocnus brownii, 54
Parocnus serus, 54
Passeriformes, 150
Paulocnus, 43
Pelecyodon, 44
Peltephilus, 2829, 77, 82, 92
Peltephilus pumilus, 95
Perissodactyla, 125
Pholidota, 1213, 2627, 30, 3233, 6465, 74
Phyllophaga, 2, 11, 2426, 3032, 35, 40,
5657, 88
Pilobolus crystalinus, 109
Pilosa, 2, 11, 14, 16, 2426, 30, 32, 34, 66, 73
Piranhea trifoliata, 27374
Placentalia, 2627, 32
Plagiocoelus? latus, 78
Plagiocoelus obliquus, 70, 78
Planops, 44, 4647, 57
Planopsinae, 46
Plesiomegatherium, 43, 4748, 57
Plesiomegatherium halmyronomum, 47
Plesiomegatherium hansmeyeri, 47
Pleurolestodon, 32
Pliomegatherium, 47
Pliometanastes, 43, 46, 4849
Pliomorphus, 4243
Plohophorini, 2728, 83, 85
Plohophorus, 27, 79, 8384
Poulsenia, 273
Prepotherium, 46
Priodontes maximus, 6, 16, 91, 147, 215, 217,
222, 233, 244, 248, 250, 253, 258, 282, 298,
3067
Procavia capensis, 183
Proeutatus, 29
Prolestodon, 32
Promegatherium, 47
Promylodon, 32
Promyrmephagus dolichoarthrus, 70, 77
Promyrmephagus euryarthrus, 70, 77
Pronothrotherium, 43, 48
Propalaehoplophoridae, 79, 8384
Propalaehoplophorinae, 79, 8384
Propalaehoplophorus, 2829, 8182
Proplatyarthrus, 42
Proscelidodon, 43, 48
Prosopis, 317
Protamandua, 28, 6972, 74, 77
Protamandua rothi, 67, 70, 73, 77
363
Proteles cristatus, 64, 258

Protobradys, 42
Prototheria, 123
Prunus, 273
Pseudalopex vetulus, 258
Pseudobombax munguba, 274
Pseudoglyptodon, 33, 40, 42, 4446
Pseudoplohophorus, 83
Pseudoprepotherium, 41, 44, 4849
Pseudorophodon, 82
Pteropus vampyrus, 183
Pyramiodontherium, 43, 47, 49, 57
Quercus humboldtii, 241
Ranculcus, 32, 43, 48
Rhizophora mangle, 274
Riostegotherium, 80
Rodentia, 123, 148
Salmonella, 104, 239
Sapindales, 52
Sapotaceae, 274
Scaeopus torquatus, 198
Scelidodon, 50
Scelidodon chiliensis, 54, 63
Scelidotherium, 43, 48, 50, 98
Scelidotherium leptocephalum, 9091, 94, 96
Schismotheriinae, 33, 46
Schmismotherium, 44
Sclerocalyptidae, 79, 83
Sclerocalyptinae, 27, 79, 83
Sclerocalyptini, 83, 85
Scotaeops, 65
Senecio, 52
Sibyllotherium, 44
Sorex araneus, 183
Spermophilis tridecemlineatus, 183
Sphenotherus, 32, 43, 48
Sporothrix schenckii, 110
Stegomastodon waringi, 62
Stegotherium, 28, 65, 74, 81
Stegotherium tessellatum, 95
Streptococcus, 104
Stromaphorus, 83
Strongyloides, 104
Strongyloides shastensis, 52
Synhapalops, 43
Synocnus, 43
Tachyglossus, 258
Tamandua, 16, 19, 2728, 6566, 6870, 72,
74, 77, 161, 164, 2024, 220, 237, 257, 260,
263, 296
Tamandua longicaudata, 202
Tamandua mexicana, 71, 73, 77, 2034, 217,
22021, 237, 25758
Tamandua tetradactyla, 22, 67, 71, 73, 77, 104,
164, 2024, 217, 22021, 232, 237, 247, 250,
253, 25758, 296
364
Taxonomic Index
Tapirus terrestris, 247

Tardigrada, 2, 11, 2426, 31, 40, 42, 57, 88, 9798
Tayassu pecari, 247
Tayassu tajacu, 247
Thalassocnus, 41, 43, 4748, 53, 9495
Thinobadistes, 32, 43, 4849
Tolypeutes, 1619, 22, 29, 68, 143, 146, 148, 150,
225, 234, 286, 298
Tolypeutes matacus, 93, 14647, 197, 202, 217,
22425, 23233, 23536, 247, 250, 25253,
28283, 28586, 288, 290, 292, 295, 298,
304, 30612
Tolypeutes tricinctus, 215, 217, 22425, 254,
258, 283, 290
Tolypeutinae, 11, 1720, 22, 202, 286, 287

Toxoplasma gondii, 104
Trichuris, 104
Trypanosama, 109
Trypanosoma cruzi, 113
Tubulidentata, 12, 64
Tupaia belangeri, 183
Tupinambis, 253
Urotherium, 7980, 8384
Urumacotherium, 43, 48
Valgipes, 43
Vassallia, 28
Vassallia maxima, 95
Vermilingua, 2, 11, 14, 16, 2426, 28, 30, 32, 35,
6474, 88, 202, 236, 238, 264
Xenocnus, 43, 50
Xyophorus, 44
Zaedyus, 1619, 143, 145, 148, 150, 201
Zaedyus pichiy, 5, 91, 12024, 145, 197, 201, 217,
22728, 283, 286, 28890, 292, 304, 313,
315, 318
Zaglossus, 258
Zazamys veronicae, 48
Subject Index
Ach, 24850, 253

Acrosome. See Sperm
Activity patterns. See Behavior
Adrenal glands: armadillos, 113, 154, 157, 160;
sloths, 139, 16566
Affinitive behavior. See Behavior
Aggression. See Behavior
Agonistic interactions: anteaters, 263, 266
Agriculture, 22425, 228, 318
Algae (in sloth hair), 131
Algorithm, 185, 191
Allantoic, 161, 164, 167
Alleles, 185, 190, 194, 207
Allometric equations, 9192, 94
Altruism, 207, 291
Amazon, 107, 131, 217, 219, 222, 225, 228, 230,
27072, 274, 277
Amnion, 164, 167, 170, 174, 176
Anatomy, 33, 89, 9298; anteaters, 6469;
armadillo spleen, 12124; comparative, 92,
97, 123, 14448; sloths, 57, 59, 13140
Andes, 6, 50, 56, 59, 6163, 217, 228, 314
Andoke, 250
Anesthetics, 113
Anesthetized: armadillos, 152; sloths, 13639
Animalivorous, 94. See also Diet
Antarctica, 42, 44
Anteaters: giant, 16, 22, 25, 7374, 91, 104,
16165, 2023, 210, 22022, 232, 23739,
247, 249, 25253, 25868, 29596; northern
tamandua, 25860, 26367; silky, 22, 161,
192, 197, 202, 21920, 230, 249, 252, 25859,
26367; southern tamandua, 25859,
26365, 267, 296. See also Anatomy;
Behavior; Diet; Ecology; Evolution; Fossils;
Physiology
Anthropic activity/impact, 262, 266, 305
Antibodies, 11417, 162, 192
Antillean megalonychids, 30, 32, 4344, 46, 53
Ants, 3, 6466, 220, 238, 25860, 263, 26667,
28687, 304, 317
Aquatic sloths. See Sloths
Arboreal/arboreality, 53, 6869, 7172, 74, 92,
135, 263, 265. See also Locomotion
Argentina, 40, 4547, 4950, 52, 54, 57, 59,
6162, 6971, 7578, 80, 90, 9394, 99,
107, 109, 115, 11718, 120, 184, 197, 22228,
24849, 253, 258, 28385, 290, 294, 296,
300305, 31317
Armadillos. See Anatomy; Behavior; Burrows;
Diet; Ecology; Evolution; Physiology
Artisanal use of xenarthrans, 245, 250, 253

Atlantic forest, 13132, 203, 210, 21920,
27072, 274
Atresia, 15354
August Krogh principle, 126
Autosomes. See Chromosomes
Aymara, 250
Ayoreode, 245, 249
Bacteria/bacterial, 4, 104, 114; fermentation,
271
Bacterial artificial chromosome (BAC),
18485, 189, 19495
Barro Colorado Island (BCI), 26364, 27178
Bayesian inference, 13, 1819
Behavior
species: anteaters, 239, 25768, 29596;
armadillos, 92, 11213, 209, 236, 28193,
29698, 300318; sloths, 59, 92, 97, 132, 137,
14042, 24243, 26980
types: activity, 132, 134, 26466, 27779,
283, 296, 3023, 30810, 317; affinitive,
288, 29091; aggression, 234, 236, 23839,
243, 28890, 317; captive, 11213, 23243;
circadian, 137, 27779, 3023, 30910,
317; feeding, 59, 97, 137, 25860, 27174,
28287, 304, 3089, 317; locomotory, 92;
reproductive, 14042, 209, 26667, 29091,
308, 317; social, 28993; space use, 26164,
27475, 28788, 3012, 309, 312, 315; thermoregulation, 26566
Bicornuate. See Uterus
Bilophodont, 4041
Biochemical analysis of fossil bone, 53
Biodiversity, 294
Biomechanics, 8889, 91, 94, 99
Bipedal/bipedalism, 59, 80, 9293
Birth, 112, 12728, 14142, 161, 165, 206, 235,
239, 242, 267, 292, 3078, 317
Bite forces, 94
Blancan, 43, 49
Blastocyst, 151, 155, 166, 168, 17374, 17880
Blood gases: sloths, 138
Blood pressure: sloths, 13637, 279
Blood sinuses, 161, 174, 17778
Bocas del Toro, 16, 130, 219
Body mass, 40, 53, 80, 91, 94, 98, 135, 270
Body size, 2, 40, 53, 5859, 88, 9192, 9799,
198, 209, 219, 265, 288, 291, 299
Body temperature, 13, 113, 11920, 127, 130,
134, 138, 158, 26566, 31415, 317; core, 53
365
Bolivia, 42, 44, 50, 57, 59, 6163, 220, 22224,

227, 24454, 28385, 294, 297, 30612
Bootstrap, 1718
Bowel impaction, 113
Brain anatomy: sloths, 131
Brazil, 18, 5052, 54, 61, 63, 71, 75, 77, 8081,
105, 1078, 110, 132, 141, 169, 198, 200, 203,
210, 21920, 222, 22425, 25253, 259,
26164, 27071, 27780, 283, 286, 294,
296
Breathing patterns: armadillos, 127; sloths, 138
Breeding/mating season: anteaters, 237, 261,
266; armadillos, 15354, 209, 236, 28990,
308, 314, 317; sloths, 141
Browsers, 94, 98
Bugl, 250
Burdigalian, 48
Burrows, 22425, 247, 253, 264, 29698,
317; in captivity, 112, 236; hypoxia, 121;
paleoburrows, 53, 94; pathogens in, 10810;
sharing, 209, 288, 29293, 298; structure of,
292, 3025, 318
Camera traps, 252, 308, 31012
Caniniforms, 4041, 46
Caprilic acid, 53
Caprylic acid, 53
Captive breeding, 128, 242. See also Behavior
Captive maintenance, 11213, 23243
Captivity, 3, 6, 110, 114, 119, 126, 14042, 152,
157, 260, 266, 295, 298, 31718
Carapace, 3, 6, 16, 21, 2527, 34, 8081, 83, 85,
92, 11213, 120, 123, 125, 127, 234, 24749,
295, 29798, 312, 31415, 317
Cardiovascular performance: sloths, 13537
Caribbean, 32, 40, 4546, 5154, 115, 118, 218,
222
Carrion-eating, 94, 98, 224, 282, 286, 297, 304,
317. See also Diet
Caudal: armor, 27; tube, 80, 82, 8485
C-banding, 200201
Cenozoic, 28, 40, 49, 56, 91, 98
Census with dogs, 253, 308, 311
Cerebellectomy, 132
Chaco, 222, 22425, 227, 230, 248, 252; Bolivian, 220, 224, 24454, 30612; Paraguayan,
220, 223, 224, 227, 29499
Chagas disease, 104
Chapadmalalan, 43, 4849, 73
Charango, 223, 248
Chile, 40, 44, 5254, 59, 6163, 313
366
Subject Index
Chiquitano, 245, 24850

Chorion, 164, 16768
Chromosomes, 16, 187, 19394, 196204; acrocentric, 19798, 200204; autosomes, 197,
200, 204; metacentric, 19798, 200204;
painting, 22, 197; sex, 162, 185, 19798, 200,
202204; submetacentric, 198, 2012, 204;
telocentric, 198
Circadian rhythms, 137, 152, 279. See also
Behavior
Cladistic, 14, 1718, 21, 26, 28, 71, 8182, 85, 192
Clomiphene, 15758
Clones: armadillos, 12729, 208, 29192;
use in genome analysis, 18487, 18992;
vertebrates, 292
Coevolution, 150
Cold-induced fever, 127
Colhuehuapian, 44, 46, 69, 73
Colloncuran, 44, 47, 70, 73
Colombia, 6163, 70, 76, 107, 115, 200, 203,
217, 228, 250, 25253, 28485
Colonial insects, 6465, 72, 74
Commercial use of xenarthrans, 224, 245,
24850
Conservation, 5, 6, 142, 232, 243, 24546, 254,
26768, 287, 3045; genetics, 22, 210; status
of extant xenarthrans, 21531, 301. See also
Management; Population
Conserved genomic elements, 182, 184, 192
Contig(s), 18689, 191, 19394
Contractile cells, 123
Corpus luteum, 15155, 157, 179
Cotyledons of placenta, 16669, 171. See also
Placenta
Courtship, 140, 142, 266, 293. See also Behavior; Mating; Reproduction/reproductive
Coverage: of genome sequence, 18190, 19294
Craniodental traits, 1718, 28, 30, 32, 8182,
91, 95
Cuba, 44, 48, 54
Cuvier, G., 57, 80, 8788
Cytochrome-b, 210
Cytogenetics, 2122, 196204
Darwin, Charles, 81, 87, 304
Dasypodids, 8083, 85, 14350
Database(s) for genomic analyses, 182, 184,
191, 19395
Decidua, 16162, 16471, 17980. See also
Placenta
Delayed implantation, 112, 151, 15457, 160,
165, 173, 17980
Dentition, 2, 26, 3334, 40, 42, 44, 46, 48, 65,
72, 89, 92
Dermatophytic, 105
Deseadan, 42, 44, 45, 57, 73
Development: embryonic, 135, 16080, 267;
follicular, 140, 15255; post-natal, 13942;
splenic, 12223
Diet: anteaters, 6465, 74, 25860; armadil-
los, 28287, 304, 3079, 317; captive, 112,

23435, 23841; fossil sloths, 41, 48, 52, 95,
9798; living sloths, 134, 27174. See also
Behavior; Nutrition
Digestibility of food items, 27172
Digestion: sloths, 134
Digging, 21, 66, 68, 9294, 282, 297, 3024,
318. See also Fossoriality
Dimorphic fungi, 10310
Diploid number, 197203
Disease(s), 10419, 23435, 239
D-loop, 210
DNA, 108, 114, 148, 165, 18194, 198, 20510;
ancient, 16, 21, 30, 52; microsatellite, 22,
128, 19294, 207, 20910, 291; mitochon
drial (mtDNA), 5, 1214, 1618, 2021,
2627, 184, 210; use in phylogenetics, 1123.
See also Genetic(s); Genome/genomic
Dwarfism, 219, 274
Ear: anatomy, 14, 27, 132, 315; tissue in leprosy
screening, 11417
Ecology: anteaters, 25768, 29596; armadillos, 28193, 296318; sloths, 13031,
26980. See also Paleoecology
Ecomorphology. See Morphology
Ecuador, 6163, 197, 200, 222, 228, 248,
25051, 253
Edentate Conservation Fund, 231
Edentates, 12, 216
Edentate Specialist Group, 6, 216, 247
Electrocardiogram (ECG), 136
Electroencephalogram (EEG), 13132
Ellipsoidal sheath. See Spleen
Emas National Park, 210, 25859, 26162,
264, 267
Ember, 250
Embryo(s), 112, 151, 157, 168, 17377, 17980
Embryonic shield, 17476, 180
Enclosures, 112, 234, 23740, 314
Endemism/endemic, 16, 63, 130, 224, 226
Endometrium, 161, 16466, 168, 171, 17374,
17778
Endothelium, 162, 16771
Energy expenditure, 92, 125, 130, 13435, 139,
265, 275. See also Metabolism/metabolic rate
Ensenadan, 43, 49, 6263, 73
Entepicondylar foramen, 4042, 66
Environmental contamination hypothesis, 115.
See also Leprosy
Eocene, 1920, 2728, 42, 57, 69, 72, 80
Epitrichium, 169
Escape behavior, 287, 298. See also Behavior
Estradiol, 105, 140, 152, 15456
Estrogen, 155, 157, 161
Estrus, 140, 161, 236, 239, 242, 26668
Euchromatic genome, 18688, 194
Euphractines, 1819, 28, 92
Eutatines, 28, 94
Evolution: anteater, 6478; armadillo spleen,
125; convergent, 3, 13, 21, 6465, 74, 258;

dasypodid sperm, 14450; glyptodonts,
8085; karyotype, 19798, 201; pathogenic
fungi, 1045; polyembryony, 5, 17980, 283,
292; significant units for conservation, 6,
22, 21720, 230; sloth, 4063; xenarthran,
12, 1136, 216
Exhibits of captive xenarthrans, 23243
Extinction, 12, 5355, 61, 87, 110, 198, 22223,
228, 230, 270, 301, 304
Eyes: sloths, 133
Facial musculature, 97, 134
Farms, 220, 241, 305, 308
Fat deposits, 68, 139, 247, 249, 298, 317
Feces, 105, 1079, 113, 131, 13436, 142, 259,
27677, 284, 286, 304, 314
Feeding: morphology, 59, 6566, 9498,
134; physiological consequences in sloths,
13738. See also Behavior; Diet
Feeding selection, 27274
Femur, 3435, 4042, 4546, 5861, 6870,
7577, 80, 92, 94; index of distal shape, 68;
transverse diameter, 92
Fertilization, 144, 149, 173
Fetus/fetal, 140, 161, 16566, 16869, 171, 176,
3078; adrenal glands, 160, 165; blood vessels, 16164, 167, 171, 179
Field education, 22021
Fire, 217, 220, 222, 268
Flood, 25859, 262, 264, 27172, 274, 27778,
292, 304, 308
Florida, 48, 51, 59, 107, 110, 11517, 128, 151,
189, 206, 210, 283, 291
Fluorescent in-situ hybridization (FISH), 22,
193, 197
Folivore/folivorous, 40, 57, 88, 134, 23940,
27374. See also Diet; Feeding
Folklore, 225, 230, 245, 295, 297, 313
Follicle stimulating hormone (FSH), 152, 155,
159
Follicular cells, 122
Follicular development. See Development
Foraging theory, 260, 284
Forearm, 3, 66, 68, 9294
Form-function correlation approach, 89
Fosmid(s), 18591, 194
Fossils
paleobiology, 8999
taxa: anteaters, 6978; cingulates, 8085;
sloths, 3963
Fossoriality, 14, 18, 66, 6869, 74, 121, 125, 225,
233, 286, 29697, 311; index of fossorial
ability, 66, 92. See also Locomotion
Founder effect, 210
Friasian, 44, 70, 73, 75
Functional morphology. See Morphology
Fundus: of eye, 133; of uterus, 151, 16162,
16569, 174, 176, 17880
Fungi, 10310
Subject Index
Gastrointestinal tract, 104, 108, 134
G-banding, 21, 197, 202
Gene(s)/genic, 56, 12, 15, 1721, 182, 184,
18994, 205. See also DNA; Nuclear
Genetic(s): conservation, 22; marker(s), 56,
2021, 128, 19294, 2067, 20910, 291;
molecular, 1123, 18194; population, 6, 22,
19294, 20511. See also Cytogenetics
Genitalia, 140, 142, 144, 149, 26667, 276, 291
Genital tract, 14950. See also Reproduction/
reproductive
Genome/genomic, 56, 1213, 16, 2022, 119,
18195, 197, 200, 210
elements: conserved, 18283, 192; functional, 182
library(ies), 18487, 18995
sequence: alignment, 182, 18688, 19193;
assembly, 18494; gaps, 18687, 189, 191;
linkage, 18789, 191, 194; quality, 18485,
18890, 194. See also DNA; Gene(s)/genic
Geographic distribution: of extant xenarthrans, 21729, 270, 28687, 29598, 3012,
305; of fossil anteaters, 72, 7578; of fossil
sloths, 4344, 5051, 53, 5759, 6163; of
leprosy, 11518; limits to, 127, 265, 287
Gestation, 141, 151, 15457, 161, 16466, 168,
171, 173, 239, 24647, 267, 317. See also
Pregnancy
Giant anteater. See Anteaters
Giant sloth. See Fossils; Sloths
Glyptodonts, 12, 2628, 32, 34, 40, 42, 44, 63,
7985, 8889, 9192, 94, 9899
Gonads, 152, 158, 317
GPS radiocollar, 26162, 264
Grazers/grazing, 41, 52, 9495, 9798. See also
Diet; Feeding
Great American Biotic Interchange (GABI),
2, 49, 57
Ground sloths. See Fossils; Sloths
Growth rates: anteaters, 267
Guaran, 24445, 24850
Guatemala, 251, 253
Habitat: destruction/degradation of, 217, 225,
254; fragmentation of, 267; use of (see
Behavior)
Hair, 5253, 109, 113, 131, 133, 142, 165, 198, 200,
249, 276, 314
Hairless appearance, 99
Hearing: sloths, 132
Heart rate, 128, 13638
Hemochorial. See Placenta
Hemopoietic, 12123
Hemphillian, 43, 4849
Herbivores/herbivory, 2, 2627, 51, 53, 58, 80,
88, 94, 9799, 109, 260. See also Behavior;
Diet; Feeding
Heritability, 12728
Heterozygosity/heterozygous, 18485, 18990,
194
367
Hibernation, 317
Histopathology, 107, 114, 11617
Hofbauer cells, 16768, 170
Home range, 141, 264, 277, 283; exclusivity,
289; overlap, 141, 26163, 266, 28889; size,
237, 26163, 27475, 28889, 3012, 308,
312. See also Behavior
Hormones, 105, 140, 15259
Huaorani, 25051
Huayquerian, 43, 4647, 49, 7071, 7376
Human exploitation of xenarthrans, 21731,
24454, 29598, 3045
Human exposure to leprosy, 118
Human genome, 18184, 18687, 18990,
19293
Hunter self-monitoring, 24547, 307, 312
Hunting, 3, 5, 54, 110, 217, 21920, 22325,
22728, 24454, 262, 268, 295, 305, 307,
312, 318; subsistence, 220, 222, 22425, 228,
245, 24749, 250, 253, 268, 295
Husbandry, 5, 112, 23235, 23738, 243
Hyoid complex, 98, 134
Hypoxia/hypoxic, 121, 125, 127
Hypselodont, 2, 26, 89, 97
Hypsodonty, 5859, 95, 97
Latency period of paracoccidioidomycosis,

1056, 109
Laventan, 44, 49, 70, 73, 76
Leishmaniasis, 104
Leprosy, 25, 11019, 127, 184, 192, 283
Leucine, 53
Library(ies). See Genome/genomic
Likelihood-based tests, 1315, 17, 209
Limb bone strength, 92
Limb proportions, 92, 94
Line transects, 252, 311
Linkage/linked. See Genome/genomic
Linoleic acid, 53
Liver, 1078, 114, 13839
Llanos, 222, 259, 26164
Locomotion, 40, 53, 5759, 68, 70, 72, 89,
9194, 9899; cardiovascular effects of, in
sloths, 137. See also Behavior
Lujanian, 40, 43, 4950, 57, 59, 6263, 71, 73,
75, 77, 9899
Lungs: anatomy of, in sloths, 13738; infection
in, 105, 1078, 113; physiology (see Respiration)
Luteinizing hormone (LH), 152, 155, 159
Lymphocyte cord, 122
Immunological response, 104, 11314, 192

Implantation, 2, 5, 21, 151, 15457, 161, 16566,
168, 17376, 17880, 29192. See also Delayed implantation
Incisiform, 41
Incubation period of leprosy, 11314
Indicator of athletic ability, 93
Indigenous peoples, 222, 224, 24454, 29698,
307
Insect defenses against anteaters, 260
Insular dwarfism. See Dwarfism
Interdigitant cells, 122
International Union for the Conservation of
Nature (IUCN), 6, 173, 21617, 231, 248;
Red List, 216, 248
Interspecific differences: in anteater behavior, 263; in burrow design, 292; in sloth
behavior, 275; in spleen anatomy,
12122
Intraspecific fights. See Behavior
Irvingtonian, 27, 40, 43, 51, 71, 73, 75
Isoso, 220, 24549, 25254, 3079, 31112
Macrophages, 12223, 167, 16971

Management: conservation, 6, 210, 24546,
25254, 287, 307; in captivity, 23243, 295.
See also Population
Mangrove forest, 219, 274
Manus, 18, 34, 41, 57, 6870, 7576, 87, 9294
Marker(s). See Genetic(s)
Masticatory apparatus, 58, 80, 89, 91, 9495, 98
Masticatory musculature, 94, 134
Maternal effects, 12728
Mating, 112, 14042, 165, 242, 267, 290, 317; in
fungi, 109; system, 209, 277, 290. See also
Behavior; Breeding/mating season; Reproduction/reproductive
Maturation, 14142; sexual, 140, 15254, 267,
317. See also Development
Maximum likelihood. See Likelihood-based
tests
Maya, 251
Medical problems in captive xenarthrans,
11213, 23341
Medicinal use of xenarthrans, 24950, 253
Megatheres, 30, 4042, 44, 46, 48, 50, 80
Megatheriid, 3233, 57, 91
Metabolism/metabolic rate, 2, 40, 55, 91, 97,
99, 125, 12728, 13435, 139, 265, 267, 271,
287, 317
Metacentric/submetacentric. See Chromosomes
Mexico, 49, 51, 71, 75, 115, 118, 203, 219, 25153,
258
Microsatellites. See DNA; Genetic(s)
Microscopy: electron, 12224, 131, 14546,
161, 16465, 169, 171; optical, 145; phase
contrast, 145
Kaa-Iya National Park, 24547, 308

Karyotype, 197204; ancestral, 22, 197, 200
Kayap, 249
Kidneys, 34, 107, 113, 13839, 165, 286
Kin selection, 2078, 29192
Labyrinthine tissue, 16364, 16869
Lacandn, 25152
Lactation, 14142, 307
Lamellae, 16364, 16769
Land use, 228, 245, 305
368
Subject Index
Midpiece. See Sperm

Migration, 57, 6162, 110
Miocene, 20, 2831, 33, 40, 4649, 57, 62, 65,
6970, 72, 7478, 81, 9192, 94
Mitochondrial genes. See DNA; Genetic(s)
Molariforms, 41, 44, 46, 48, 58
Molecular: dating, 1820, 71, 73; evolution, 14,
2022; markers, 108, 190, 20510; phylogenetics/systematics, 1, 6, 1122, 2628, 30,
32, 81, 105, 184, 202; strain typing, 107, 109,
11819. See also DNA; Genetic(s)
Monosomy, 200201
Montehermosan, 43, 48, 59, 6163, 71, 73, 76
Morpho-geometry, 89, 99
Morphology: anteater, 4, 6469; armadillo,
128, 154, 158, 282, 31415; eco, 89; fossil
sloth, 4042, 46, 5759; functional, 8889,
94; fungal, 109; glyptodont, 8183; xenarthran, 6, 1214, 1618, 21, 2436, 8898. See
also Anatomy; Placenta; Sperm; Spleen
Morphometrical analysis, 121, 31415
Moths, 131
Motor activity, 13436, 138
Mouse genome, 181, 184, 187, 189, 19192
Muscle(s)/musculature, 3, 6566, 69, 92, 94,
9798, 123, 130, 13335, 137, 165
Musum National dHistoire Naturelle, 87
Mustersan, 27, 42, 73
Muzzle shape, 35, 95, 9798
Myeloid, 123
Mylodont, 43, 48, 52; mylodontid, 30, 9192,
94; mylodontine, 32
Myocardial activity, 13536
Myrmecophagy, 1314, 6465, 74, 88, 94,
25860, 28486
Nasal: anatomy, 3435, 65; cartilages, 97
Natural history, 21728, 283, 29599, 31318
Nematodes, 52, 104, 113
Neotropical forests, 74, 264, 270, 27374, 276,
280
Neotropics, 21617, 21920, 22728, 239, 245,
248, 25354, 27578
Nests: anteater, 240, 266; armadillo, 110, 236,
292, 298, 303, 318; insect, 3, 64, 66, 243,
25960, 262, 296, 298; sloth, 241
Neuromuscular physiology, 135
Niche differentiation, 63, 95, 98, 305
Nine-banded armadillo, 1, 5, 20, 22, 104,
10719, 12629, 15162, 165, 17380, 18393,
196, 20610, 236, 24749, 25253, 282, 287,
291, 297
Nocturnal activity, 134, 216, 219, 22627, 238,
258, 26366, 270, 27779, 296, 302, 310
Nodules, 114, 122, 170
North America, 23, 27, 2931, 40, 4344, 46,
4854, 5657, 59, 80, 88, 95, 9798, 110, 118,
185, 22729, 23233, 235, 282, 284, 287, 298
Northern tamandua. See Anteaters
Nothrotheres, 30, 33, 4042, 46, 48, 52, 54
Nothrotheriid, 30, 33, 94

Nuclear: genes, 1215, 1719, 2627, 184, 206
10; organizing region, 198; volume, 148.
See also DNA; Gene(s)/genic; Genetic(s);
Genome/genomic
Nutrients, 128, 131, 137, 25960, 271
Nutrition, 134, 180, 23435, 23841, 260, 271
Occlusal surface area, 4142, 48, 95, 97
Olecranon, 66, 68, 76, 92
Oligocene, 40, 42, 4446, 48, 57, 69
Omnivores, 88, 94, 112, 286, 304, 317. See also
Behavior; Diet; Feeding
Oocyte, 144, 15152
Ortholog/orthologous, 184, 18687, 19192
Osteodentine, 2, 89
Ovariectomy, 155, 157
Ovaries, 113, 15055, 157, 161, 169, 173
Ovulation, 140, 151, 15354, 157, 173, 17980,
236
Owen, R., 80, 87
Oxygen utilization, 138, 154, 179
Pairing behavior, 29091. See also Behavior;
Mating; Reproduction/reproductive
Paleobiogeography/paleogeography, 40,
4251, 5763, 7071, 7578
Paleobiology/paleontology, 2599
Paleoburrows. See Burrows
Paleocene, 2, 1819, 72, 80, 88, 91
Paleoecology, 40, 5153, 9899
Pampas, 50, 59, 62, 300305
Pampatheres, 2, 2628, 32, 8082, 85, 88, 91,
94, 98
Panama, 16, 4849, 57, 105, 130, 19798, 200,
216, 21819, 242, 250, 253, 259, 26364, 266,
27071, 277
Pancreas, 139
Pantanal, 25865
Paracoccidioidomycosis (PCM), 1059
Paragravigrade, 40, 42
Paraguay, 61, 115, 210, 220, 22325, 227, 246,
248, 250, 253, 259, 29499, 3012
Parasites, 52, 1045, 113, 123, 239
Parasitic phase of fungal infection, 104
Pathogens, 10319
Pathologies, 23435, 23839. See also
Disease(s)
Pedolateral, 41, 92
Peltephiline, 82, 94
Periarterial sheath, 122
Pericentric inversion, 197, 201
Peru, 6, 48, 54, 5663, 9495, 184, 228, 249,
25153
Pes, 34, 4142, 57, 6970, 7677, 87, 92
Phylogenetic(s): analyses, 1133, 5759, 7174,
8285, 87, 105, 14849; hypotheses, 6,
1218, 2021, 32, 71, 81, 8385; networks,
1718; relationships (see Phylogeny);
uncertainty, 19
Phylogenomics, 20
Phylogeny: of anteaters, 16, 1820, 28, 30,
7174; of armadillos, 1620, 2729, 14849;
of glyptodonts, 8285; of placental mammals, 1215, 2627, 183; of sloths, 16, 1820,
3032, 5759; of xenarthra, 6, 14, 2122
Phylogeography, 6, 22, 105, 109, 206, 20910,
220, 230
Physiology, 6; of anteaters, 16165; of armadillos, 11314, 12229, 15159, 17380, 31517;
of sloths, 13042, 16571
Pichi, 91, 22728, 31318
Placenta: fetal/maternal barrier, 163, 167,
16970; hemochorial, 16364, 169, 173, 176,
267; septa, 16668; villi, 16165, 16870,
17580
Placental mammals, 1, 1215, 20, 22, 2627, 65,
88, 183, 187, 287
Placental tree rooting, 1315, 20, 22, 2627
Plant-eating. See Herbivores/herbivory
Plasmid(s), 185, 18788, 194
Plasticity of organs, 120
Pleistocene, 2, 19, 25, 27, 2933, 40, 4655, 57,
59, 6162, 71, 75, 77, 80, 88, 9192, 9495,
9799, 219, 228
Pliocene, 19, 28, 33, 40, 4749, 57, 59, 6162,
71, 76, 91, 94
Podocarp forest, 42
Pododermatitis, 113
Polyembryony, 2, 5, 21, 173, 17980, 2068,
230, 283, 29192
Polygyny, 209, 277, 290
Polymerase chain reaction (PCR), 1078, 114,
118, 18485, 19192
Population: abundance, 29899, 308, 31011;
bottleneck, 206, 210, 268; density, 99,
11718, 141, 242, 246, 252, 262, 27677,
283, 28889, 308, 311; density estimates,
25253, 276, 308; differences, 11617, 198,
210, 21820, 259, 261, 27273, 275, 27778,
28485, 28889, 311, 31415; genetics (see
Genetic[s]); management, 196, 23032, 245,
25254, 305, 312 (see also Conservation)
Postcranial skeleton, 27, 32, 48, 6669, 81, 85
Predation, 109, 125, 208, 266, 287, 298
Predators: defenses against, 127, 230, 236, 282,
287, 29193; of fungi, 104; of xenarthrans,
131, 262, 266, 276, 296; xenarthrans as, 94,
260, 287
Pregnancy, 141, 155, 16162, 16566, 17778,
200, 235, 239, 242, 289, 308. See also Gestation
Prevalence of leprosy, 11518. See also Leprosy
Prey preferences, 97, 241, 259, 266, 27174,
28487, 317. See also Behavior; Diet
Primer(s), 18485, 19193
Primitive streak, 17677, 180
Progesterone, 116, 140, 15258, 236
Proline, 53
Propagation. See Leprosy
Subject Index
Prostate gland, 158
Protein(s), 12, 105, 134, 140, 142, 182, 189,
19192, 235, 241, 271
Protein electrophoresis, 206, 209
Puerto Rico, 45, 48, 54
Quadrupedal stance, 3, 40, 59, 63, 92
Quadruplets, 152, 173, 2067, 247, 291. See also
Polyembryony
Quichua, 251
Radiotelemetry, 26164, 278, 308, 31112
Ramn, J., 8788
Rancholabrean, 40, 43, 51, 98
Ranging patterns, 253, 27475, 287. See also
Behavior; Home range
Real Gabinete de Historia Natural de Madrid, 87
Red List. See International Union for the
Conservation of Nature (IUCN)
Red pulp. See Spleen
Reproduction/reproductive, 6, 109, 14042,
144, 157, 160, 165, 173, 196, 23536, 239, 242,
246, 26668, 283, 29092, 3078, 317; cycle,
5, 15159, 23536; success, 5, 209, 233, 236,
239, 242, 29091; system, 140, 144, 154,
15859; tract, 5, 21, 144, 14950. See also
Behavior; Breeding/mating season; Genital
tract; Mating
Respiration, 127, 13839
Rest sites, 26365
Rete mirabile, 137
Reticuloendothelial tissue, 11314
Road kill, 116, 223, 262, 268, 286, 29596,
299
Robertsonian fusion, 197, 201
Saint-Hilaire, E., 87
Santacrucian, 3031, 40, 44, 4649, 57, 65,
6970, 7274, 7778, 92
Saprobic phase, 1045, 1079
Scaffold(s), 18687, 189, 191, 194
Scale models, 91
Scelidothere, 4142, 46, 48, 50, 54
Schiff reaction, 122
Scutes, 3, 18, 34, 8083, 85, 314
Seasonal variation: in activity, 26465, 302;
in diets, 141, 235, 25860, 28487, 304; in
hormone secretion, 153, 159; in movements,
275
Semi-arboreal, 53, 72, 92
Septa. See Placenta
Sequence(s)/sequencing. See Genome/genomic
Sequence evolution, 1320. See also Phylogenetics; Phylogenomics
Serology, 107, 115, 117
Serra da Canastra National Park, 259, 26163,
26768
Sex chromosomes. See Chromosomes
Sex determination, 19698, 275
Sex ratios, 27677

Sexual dimorphism, 99, 196, 209, 275, 277,
290, 314
Sexual reproduction, 144; in fungi, 109
Shotgun genome sequencing, 18485, 18788,
191, 194. See also Genome/genomic
Shuar, 251
Silky anteater. See Anteaters
Single nucleotide polymorphisms (SNPs), 118,
19394
Siona-Secoya, 248, 252
Sirion, 24950
Skulls, 2, 14, 26, 29, 31, 33, 35, 41, 45, 4752, 58,
6061, 65, 67, 7071, 7477, 8182, 9598
Sloths: aquatic, 40, 48, 53, 9495; giant, 87, 92,
99; ground, 12, 16, 26, 30, 42, 46, 53, 56
63, 82, 87, 89, 9199; three-toed, 16, 19, 31,
13042, 165, 16771, 192, 199200, 21819,
27073, 275, 278, 280; two-toed, 16, 1920,
22, 25, 31, 13035, 14042, 16568, 183,
200201, 217, 241, 27071, 27879; urban,
198, 276. See also Anatomy; Behavior; Diet;
Ecology; Evolution; Fossils; Physiology
Social behavior. See Behavior
Soil ingestion, 97, 286
South America, 12, 5, 12, 20, 25, 2832, 40,
4246, 4850, 5254, 5657, 59, 6163, 65,
70, 7274, 80, 8789, 91, 95, 9799, 105,
10910, 130, 185, 203, 2067, 210, 21517,
22122, 229, 235, 238, 258, 270, 286, 290,
294, 301, 307
Southern tamandua. See Anteaters
Species accounts of extant xenarthrans: all
species, 21728; in Argentina, 3015; in
Paraguay, 29598
Sperm: acrosome, 144, 146, 148; dimensions,
14446, 14950; evolution, 14344, 14950;
head, 14448, 150; midpiece, 144; morphology, 14445, 14950; morphometry,
145; nucleus, 14445, 148; shape, 14446,
14850; size, 14446, 148; tail, 144, 146,
14849
Spermatogenesis, 141, 158
Spermatozoa/spermatozoon, 1619, 14445,
14751, 158, 317
Spleen: ellipsoidal sheath, 122; red pulp,
12224; storage, 121, 125; transitional zone,
12223; white pulp, 12123
Spool-and-thread estimates of home range,
253, 308, 312
Sporotrichosis, 110
Stable isotope, 52, 55
Stearic acid, 53
Sternebrae, 2, 3334, 238
Steroids, 112, 140, 161
Stomach, 3, 35, 134, 241, 271; contents, 251,
272, 28487, 304, 3079, 317
Strength indicator, 92, 94
Submetacentric. See Chromosomes
Subsistence hunting. See Hunting
369
Subspecies: armadillos, 197, 210; silky anteaters, 220; sloths, 21719; tamanduas, 220
Supraordinal relationships, 2627, 32, 183
Susceptibility to leprosy, 5, 113, 117, 119, 19293
Sustainable use, 245, 252, 254
Sylvan leprosy, 11415, 118
Syncytium, 161, 164, 16769, 171
Synsacrum, 2, 69
Syntenic/synteny, 18687, 194
Systematics, 12, 27, 32, 40, 202
Tacana, 247, 24950
Tail, 2, 34, 59, 69, 88, 92, 94, 26566, 290, 315;
injuries to, 113, 234; prehensile, 35, 6970,
72, 263. See also Sperm
Tardigrades, 2, 40, 42, 88, 90, 92, 9697
Taxonomy, 25, 2728, 32, 40, 42, 4647, 83,
85, 105, 204, 295; alpha-level, 33, 49, 55, 71;
parataxonomy, 3
Teeth, 23, 6, 26, 3336, 4042, 5859, 8081,
83, 89, 95, 97, 134, 142, 235, 238, 258, 270,
282, 314. See also Dentition
Telescoping of masticatory apparatus, 80, 94
Telocentric. See Chromosomes
Telomere(s), 198
Termitaria, 74
Termite(s), 3, 6466, 72, 220, 238, 25860,
263, 266, 28487, 304; nests, 64, 233, 239,
243, 25960, 262, 267, 282, 296, 298
Terrestrial, 2, 48, 53, 6870, 72, 74, 8788, 92,
28687
Tertiary, 2, 12, 18, 20, 8889
Testes, 13, 16, 34, 113, 15859, 165, 290, 317
Testosterone, 141, 15253, 15859
Thermal conductance, 53, 127, 266
Thermoregulation, 26568, 287, 29293, 302,
317. See also Metabolism
Thyroid, 139
Ticuna, 252
Tinguirirican, 42, 44, 73
Toba, 249
Tolerance, 288, 29091. See also Behavior
Topologies: alternative, 1315
Torpor, 314, 317
Trabeculae, 121, 16165, 167
Trabecular system, 12122
Track plots, 308, 31112
Transitional zone. See Spleen
Translocation: chromosomal, 19697, 200201
Traviportal, 92
Tree diversity, 27274
Trophoblast, 16165, 16769, 171, 17475,
17778, 180; cytotrophoblast, 161, 16465,
16869, 17579; plasmoditrophoblast, 161;
syncytiotrophoblast, 167, 17579
Tsimane, 247, 24950
Tunnels, 94
Twinning/twins, 21, 12728, 173, 180, 207, 291.
See also Polyembryony
Tyrosine, 53
370
Subject Index
Umbilical cord, 152, 16162, 16467, 16970

Uquian, 43, 49, 71, 73
Uterus, 13, 21, 151, 157, 161, 16364, 16870,
17374, 17680; bicornuate, 160, 180; simplex, 154, 16062, 166, 169, 171, 173, 17980.
See also Fundus
Villi. See Placenta

Virulence, 104, 107
Virus, 104
Viscera, 1078
Vision, 133
Vocalizations, 132, 276, 296, 318
Venezuela, 6163, 222, 228, 230, 25253, 259,

26164, 28385
Ventricular fibrillation, 136
Vertebral articulations, 2, 12, 26, 33
Waimiri, 252
White pulp. See Spleen
Wich, 249
Wildlife trade, 222
Xavante, 249, 252

Yanesha, 252
Yanomam, 249, 252
Yekwana, 252
Yolk sac, 161, 16465, 168, 171, 174, 17678,
180
Zoological gardens, 23243

Number Tu Biology

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Number Tu Biology

Uploaded by

Copyright:

Available Formats

The Biology

Edited by Sergio F. Vizcano and W. J. Loughry

university press of florida

The Biology of the Xenarthra

University Press of Florida

The Biology of the Xenarthra

Edited by Sergio F. Vizcano and W. J. Loughry

University Press of Florida

Copyright 2008 by Sergio F. Vizcano and W. J. Loughry

To Bill and Janet Loughry for then,

To Eric and Negra, my hardy roots,

3. Morphology-based investigations of the

11. The spleen of the armadillo: Lessons of organ

12. The use of armadillo clones from the genus

13. The physiology of two- and three-toed sloths 130

14. Sperm evolution in dasypodids 143

Timothy J. Gaudin and H. Gregory McDonald

15. Reproductive biology of the nine-banded

Part 2. Fossil Xenarthra

4. Fossil history of sloths 39

5. Paleogeographic distribution and anatomical

6. Skeletal anatomy and the fossil history of the

7. The evolution of armored xenarthrans and a

8. Form, function, and paleobiology in

Part 3. Living Xenarthra: Physiology and Genetics

10. Leprosy 111

16. Reproductive parameters and placentation in

17. Placentation in armadillos, with emphasis on

18. Sequencing the armadillo genome 181

19. Chromosomal studies in the Xenarthra 196

20. Genes and demes: Population genetic analyses of

Part 4. Living Xenarthra: Conservation

22. Maintenance of Xenarthra in captivity 232

groups in Latin America, and implications for

28. Ecology and conservation of three species of

Andrew J. Noss, Rosa Leny Cullar, and Erika Cullar

Agustn M. Abba and Marcelo H. Cassini

Part 5. Living Xenarthra: Ecology and Behavior

29. Biology and ecology of armadillos in the Bolivian

24. Anteater behavior and ecology 257

25. Sloth ecology: An overview of field studies 269

26. Behavioral ecology of armadillos 281

27. Xenarthrans of the Paraguayan Chaco 294

30. The natural history of the pichi, Zaedyus pichiy, in

11.3. The transitional zone of Chaetophractus

2.1. Remaining uncertainties in armadillo molecular

15.8. Plasma progesterone levels of adult female

17.10. Area of attachment 178

22.9. Juvenile Bradypus variegatus feeding on fresh,

28.2. Diagrams of the burrows of the three species of

23.2. Hunting of Xenarthra and Other Principal

2.1. Where Do Xenarthrans Fit among Placentals? 15

be followed by shorter chapters covering specific topics

The beginnings of this book occurred during our brief

in this book. Consequently, the reader should be alerted

Evolution of the Xenarthra

There are, at this writing, 31 living species of xenarthrans:

S. F. Vizcano and W. J. Loughry

ern-day armadillos and noted for the armor covering

brae, from which the name of the order is derived, (2)

Xenarthran biology: Past, present, and future

tended somewhat in a gorilla-like fashion, are certainly

reduced diversity of extant xenarthrans relative to fossil

S. F. Vizcano and W. J. Loughry