Content uploaded by Eric G Ekdale
Author content
All content in this area was uploaded by Eric G Ekdale on Mar 11, 2015
Content may be subject to copyright.
Copyright
by
Eric Gregory Ekdale
2009
The Dissertation Committee for Eric Gregory Ekdale Certifies that this is the
approved version of the following dissertation:
VARIATION WITHIN THE BONY LABYRINTH OF MAMMALS
Committee:
Timothy Rowe, Supervisor
Christopher J. Bell
James T. Sprinkle
Matthew W. Colbert
Zhe-Xi Luo
VARIATION WITHIN THE BONY LABYRINTH OF MAMMALS
by
Eric Gregory Ekdale, B.A.; M.S.
Dissertation
Presented to the Faculty of the Graduate School of
The University of Texas at Austin
in Partial Fulfillment
of the Requirements
for the Degree of
Doctor of Philosophy
The University of Texas at Austin
December 2009
Dedication
To my father, Tony, and sister, Joan, for their unyielding support and love.
Epigraph
“It is one of the most beautiful objects with which the anatomist is at all likely to meet…”
-Albert A. Gray (1907, p. 39), upon observing the inner ear of the yellow-faced baboon
vi
Acknowledgements
I give the most sincere thanks to my advisor, Tim Rowe. Besides fulfilling the
role of supervisor, Tim has been a mentor, and I look forward to calling him a life-long
colleague. Chris Bell, who has been a second advisor to me in many ways, challenged
me to think critically at every step of the way, which is a skill that I will carry with me
throughout the rest of my career. Matt Colbert introduced me to the exciting world of
computed tomography, and Jim Sprinkle, with whom I worked as a teaching assistant
across several semesters, expanded my knowledge of paleontology beyond the
vertebrate lineage. Zhe-Xi Luo added significant insight to my research through
conversations concerning the mammalian ear and early mammal evolution, and he
taught me the power of imagery in my dissertation.
The staff of The University of Texas High-Resolution X-ray CT facility
(UTCT), an NSF-supported multi-user facility funded by NSF EAR-0345710,
participated in hours of data collection and image processing, particularly Matt Colbert,
Rich Ketcham, Jessie Maisano, Amy Balanoff, Alison Mote, and Rachel Racicot. All
CT data utilized in the dissertation were collected at UTCT, except the data for
Trichechus manatus, which was scanned for Tim Hullar at Washington University in
St. Louis, MO. Funding for the scanning of specimens was provided by the Jackson
vii
School of Geosciences at The University of Texas at Austin, the Texas Academy of
Science, and the Paleontological Society.
Access to preexisting CT datasets were provided by Maria Chapala, Ashley
Gosselin-Ildari, Tim Hullar, Chris Kirk, Ted Macrini, Mike Novacek, Sentiel Romel,
Tim Rowe, Erik Sieffert, Nancy Simmons, and “Digital Morphology”
(www.digimorph.org). Additional specimens that were CT scanned were provided by
several individuals and institutions, including David Archibald and the members of the
Uzbekistan/ Russian/ British/ American/ Canadian joint paleontological expedition,
Kyzylkum Desert, Uzbekistan; Mike Novacek of the American Museum of Natural
History, New York, NY; Tim Rowe and Lyn Murray of the Vertebrate Paleontology
Laboratory, Texas Natural Science Center, Austin, TX; Jeff Saunders and Chris Widga
of the Illinois State Museum, Springfield, IL; the Southwest Foundation for Biomedical
Research in San Antonio, TX.
In addition to those mentioned above, there are numerous people who must be
acknowledged. Surely, there are many names missing from the list. Christian George
and Ryan Ewing instilled sanity when it was needed most. Jen Olori, Kerin Claeson,
and Heather Ahrens put up with more than they could have ever imagined. Ted
Macrini, Gabe Bever, and Chris Jass fostered my early academic ontogeny at UT. Other
individuals that positively affected my education include Joey Carlin, Dave Dufeau,
Sebastian Egberts, Murat Maga, Jeri Rodgers, Dennis Ruez, Nick Smith, Nina Triche,
Jon Wagner, and Patrick Wheatley.
Lastly, my undying gratitude goes to my family, Tony and Joan. Their eternal
love and support is the single reason I was able to wake up every morning, and to
complete this major life accomplishment.
viii
Variation within the bony labyrinth of mammals
Publication No._____________
Eric Gregory Ekdale, Ph.D.
The University of Texas at Austin, 2009
Supervisor: Timothy Rowe
The morphological diversity of the external and internal surfaces of the petrosal
bone, which contains the structures of the inner ear, across a broad range of therian
mammals is documented, and patterns of variation across taxa are identified. One
pattern of variation is the result of ontogenetic changes in the ear region, as described
for the external surface morphology of a sample of isolated petrosal bones referred to
Proboscidea from Pleistocene deposits in central Texas. The morphology of the
aquaeductus Fallopii for passage of the greater petrosal branch of the facial nerve
supports an ontogenetic explanation for some variation within the proboscidean sample,
and a sequence of ossification surrounding the aquaeductus Fallopii is hypothesized.
Further ontogenetic patterns are investigated using digital endocasts of the bony
labyrinth (preserved on the internal surfaces of the petrosal) constructed from CT data
across a growth series of the opossum Monodelphis domestica. Strong correlation
between skull length and age is found, but from 27 days after birth onward, there is no
correlation with age among most dimensions of the inner ear. Adult dimensions of
several of the inner ear structures are achieved before the inner ear is functional in M.
ix
domestica. Morphological variation within the inner ear of several eutherian mammals
from the Cretaceous of Asia, including zhelestids from the Bissekty Formation of
Uzbekistan, is described. The variation within the fossil sample is compared to that
observed within extant species of placental mammals, and it is determined that the
amount of variation within the Bissekty zhelestid population is within the range of that
measured for extant species. Additional evolutionary and physiological patterns
preserved within the walls of the bony labyrinth are identified through a high level
anatomical comparison of the inner ear cavities across Placentalia as a whole. In
particular, features of the inner ear support monophyly of Cetacea, Carnivora,
Primatomorpha, and caviomorph Rodentia. The volumetric percentage of the vestibular
apparatus (vestibule plus semicircular canals) of aquatic mammals is smaller than that
calculated for terrestrial relatives of comparable body size. Thus, aspects of the bony
labyrinth are both phylogenetically and physiologically informative.
x
Table of Contents
List of Tables ........................................................................................................ xii!
List of Figures...................................................................................................... xiv!
CHAPTER 1: INTRODUCTION TO THE EAR OF MAMMALS........................1!
GROSS ANATOMY OF THE MAMMALIAN EAR ...................................1!
BIOLOGICAL SIGNIFICANCE OF THE EAR ...........................................4!
DISSERTATION OVERVIEW .....................................................................7!
CHAPTER 2: MORPHOLOGY AND VARIATION IN THE PETROSAL
OF EXTINCT ELEPHANTOIDEA (PROBOSCIDEA) FROM
CENTRAL TEXAS ......................................................................................10!
ABSTRACT..................................................................................................10!
INTRODUCTION ........................................................................................10!
MATERIALS AND METHODS..................................................................12!
ANATOMICAL OBSERVATIONS OF FRIESENHAHN CAVE
PETROSALS .......................................................................................18!
ANATOMICAL OBSERVATIONS OF MAMMUT PETROSALS FROM
BONEY SPRING.................................................................................30!
DISCUSSION...............................................................................................31!
CONCLUSIONS ..........................................................................................45!
CHAPTER 3: POSTNATAL ONTOGENETIC VARIATION IN THE BONY
LABYRINTH OF MONODELPHIS DOMESTICA (MAMMALIA:
MARSUPIALIA)..........................................................................................47!
ABSTRACT..................................................................................................47!
INTRODUCTION ........................................................................................48!
MATERIALS AND METHODS..................................................................52!
RESULTS .....................................................................................................61!
DISCUSSION...............................................................................................70!
CONCLUSION.............................................................................................76!
xi
CHAPTER 4: THE BONY LABYRINTH OF ZHELESTIDS (MAMMALIA:
EUTHERIA) AND OTHER MESOZOIC MAMMALS..............................78!
ABSTRACT..................................................................................................78!
INTRODUCTION ........................................................................................79!
MATERIALS AND METHODS..................................................................83!
BONY LABYRINTH OF ZHELESTIDS ....................................................92!
COMPARISON WITH CRETACEOUS EUTHERIANS .........................101!
MORPHOLOGICAL VARIATION...........................................................106!
DISCUSSION.............................................................................................110!
CONCLUSIONS ........................................................................................118!
CHAPTER 5: THE BONY LABYRINTH OF PLACENTAL MAMMALS .....120!
ABSTRACT................................................................................................120!
INTRODUCTION ......................................................................................121!
MATERIALS AND METHODS................................................................125!
RESULTS –ANATOMICAL COMPARISONS........................................140!
RESULTS – DIMENSION COMPARISONS ...........................................386!
DISCUSSION.............................................................................................394!
CONCLUSIONS ........................................................................................407!
Appendix 1...........................................................................................................409!
Appendix 2...........................................................................................................411!
References............................................................................................................412!
Vita .....................................................................................................................439!
xii
List of Tables
TABLE 2.1. Variation observed among elephantoid petrosals from Friesenhahn Cave
(Bexar County, Texas).............................................................................................. 21!
TABLE 2.2. Semicircular canal radii of curvature (in mm) for extant elephants and
elephantoid from Friesenhahn Cave ......................................................................... 36!
TABLE 3.1. CT scanning parameters for specimens of Monodelphis domestica ............ 54!
TABLE 3.2. Angular measurements of the bony labyrinth across an ontogenetic series
of Monodelphis domestica ........................................................................................ 64!
TABLE 3.3. Linear measurements of the semicircular canals across an ontogenetic
series of Monodelphis domestica .............................................................................. 65!
TABLE 3.4. Linear measurements of the cochlea and other morphological structures
across an ontogenetic series of Monodelphis domestica........................................... 66!
TABLE 3.5. Volumes of compartments within the bony labyrinth across an ontogenetic
series of Monodelphis domestica .............................................................................. 68!
TABLE 3.6. Volume percentages and ratios calculated for the bony labyrinth across an
ontogenetic series of Monodelphis domestica .......................................................... 69!
TABLE 4.1. Volume of the cochlea, vestibule (including semicircular canals and
ampullae), and entire bony labyrinth for zhelestids and other selected Cretaceous
eutherians listed in Appendix 1................................................................................. 95!
TABLE 4.2. Dimensions and orientations of the cochlea of zhelestids and other
Cretaceous eutherians ............................................................................................... 96!
TABLE 4.3. Dimensions of internal structures of the cochlea at each quarter turn of
coiled canal for zhelestids and other Cretaceous eutherians..................................... 98!
TABLE 4.4. Orientations of the semicircular canals for zhelestids and selected
Cretaceous eutherians ............................................................................................. 102!
TABLE 4.5. Orientations of the semicircular canals for zhelestids and selected
Cretaceous eutherians ............................................................................................. 103!
TABLE 4.6. Coefficients of variation (CV) for measurements of labyrinth dimensions
of Cretaceous eutherians (including zhelestids) and selected extant taxa listed in
Appendix 1.............................................................................................................. 108!
TABLE 4.7. Ratios between dimensions of semicircular canals.................................... 114!
TABLE 5.1. Taxa examined and scanning parameters................................................... 128!
TABLE 5.2. Body mass, skull length, and dimensions of the entire bony labyrinth of
placentals................................................................................................................. 146!
TABLE 5.3. Dimensions and orientations of the cochlea of placentals ......................... 148!
TABLE 5.4. Dimensions of vestibular elements and orientations of semicircular
canals....................................................................................................................... 150!
TABLE 5.5. Linear dimensions of the semicircular canals ............................................ 152!
TABLE 5.6. Deviations and aspect ratios of the semicircular canals............................. 154!
TABLE 5.7. Coefficients of correlation (r) calculates for dimensions over body mass. 387!
TABLE 5.8. Coefficients of correlation (r) calculated for dimensions of the cochlea... 392!
xiii
TABLE 5.9. Coefficients of correlaton (r) calculated for dimensions of the
semicircular canals.................................................................................................. 393!
TABLE 5.10. Linear deviations of the semicircular canals of Monodelphis domestica 396!
TABLE 5.11. Ratios of Semicircular Canal Arc Radius of Curvature over Average
Body Mass and Slender Canal Length.................................................................... 398!
xiv
List of Figures
FIGURE 1.1. Gross anatomy of the inner ear..................................................................... 3!
FIGURE 2.1. Three-dimensional CT reconstrutions and labeled line drawings of
petrosal of elephantoid from Friesenhahn Cave (TMM 933-950)............................ 15!
FIGURE 2.2. Photograph and line drawing if stapes of elephantoid from Friesenhahn
Cave (TMM 933-950)............................................................................................... 20!
FIGURE 2.3. Variation within aquaeductus Fallopii of elephantoids from Friesenhahn
Cave, petrosals in tympanic view ............................................................................. 24!
FIGURE 2.4. Foramen for stapedius muscle in elephantoids from Friesnehahn Cave,
petrosals in medial view............................................................................................ 26!
FIGURE 2.5. Digital endocast and labeled line drawing of bony labyrinth of inner ear
of elephantoid from Friesenhahn Cave (TMM 933-950) ......................................... 27!
FIGURE 2.6. Diagram of petrosal of fetal elephant (redrawn from Eales, 1926)............ 42!
FIGURE 3.1. Graphical depiction of angular measurements between semicircular
canals and cochlea..................................................................................................... 57!
FIGURE 3.2. Cross-section through midline of the cochlea displaying the
measurements for height and width of the spiral of the cochlea............................... 59!
FIGURE 3.3. Comparison of endocasts of the bony labyrinths of Monodelphis
domestica across ontogenetic ages in ventral, lateral, and posterior views .............. 63!
FIGURE 3.4. CT slices through the cochlea of Monodelphis domestica ......................... 71!
FIGURE 3.5. Comparison of the shape of the semicircular canals of Monodelphis
domestica .................................................................................................................. 75!
FIGURE 4.1. Diagrams of measurements of cochlear dimensions .................................. 85!
FIGURE 4.2. Diagrams of measurements of semicircular canal dimensions................... 90!
FIGURE 4.3. Endocast of left zhelestid bony labyrinth (URBAC 03-39) ....................... 94!
FIGURE 4.4. Graphic reconstructions of internal structures of the cochlea of zhelestid
specimens following methods of Guild (1921), Schuknecht (1953), and Wever et
al. (1971 a, b) ............................................................................................................ 99!
FIGURE 4.5. Bony labyrinths of Cretaceous eutherians listed in Appendix 1 in lateral
view......................................................................................................................... 104!
FIGURE 5.1. Petrosal of Dasypus novemcinctus (TMM M-1885) within which sits
endocast of bony labyrinth...................................................................................... 123!
FIGURE 5.2. Cladogram of Theria including taxa considered ...................................... 126!
FIGURE 5.3. Measurement methods employed............................................................. 131!
FIGURE 5.4. Bony labyrinth of Didelphis virginiana ................................................... 141!
FIGURE 5.5. Original CT slices through ear region of Didelphis virginiana................ 143!
FIGURE 5.6. Bony labyrinth of Kulbeckia kulbecke ..................................................... 160!
FIGURE 5.7. CT slices through ear region of Kulbeckia kulbecke ................................ 162!
FIGURE 5.8. Bony labyrinth of Chrysochloris sp. ........................................................ 170!
FIGURE 5.9. CT slices through ear region of Chrysochloris sp.................................... 172!
FIGURE 5.10. Bony labyrinth of Hemicentetes semispinosum...................................... 174!
xv
FIGURE 5.11. CT slices through ear region of Hemicentetes semispinosum ................ 176!
FIGURE 5.12. Bony labyrinth of Macroscelides proboscideus ..................................... 183!
FIGURE 5.13. CT slices through ear region of Macroscelides proboscideus................ 185!
FIGURE 5.14. Bony labyrinth of Orycteropus afer ....................................................... 192!
FIGURE 5.15. CT slices through ear region of Orycteropus afer.................................. 194!
FIGURE 5.16. Bony labyrinth of Procavia capensis ..................................................... 199!
FIGURE 5.17. CT slices through ear region of Procavia capensis................................ 202!
FIGURE 5.18. Bony labyrinth of Trichechus manatus .................................................. 207!
FIGURE 5.19. CT slices through ear region of Trichechus manatus............................. 210!
FIGURE 5.20. Bony labyrinth of the fossil elephantoid proboscidean .......................... 215!
FIGURE 5.21. CT slices through ear region of the fossil elephantoid proboscidean..... 218!
FIGURE 5.22. Bony labyrinth of Dasypus novemcinctus .............................................. 222!
FIGURE 5.23. CT slices through ear region of Dasypus novemcinctus......................... 224!
FIGURE 5.24. Bony labyrinth of Bathygenys reevesi.................................................... 235!
FIGURE 5.25. CT slices through ear region of Bathygenys reevesi .............................. 237!
FIGURE 5.26. Bony labyrinth of Sus scrofa .................................................................. 239!
FIGURE 5.27. CT slices through ear region of Sus scrofa............................................. 241!
FIGURE 5.28. Bony labyrinth of fossil Balaenopteridae............................................... 249!
FIGURE 5.29. CT slices through ear region of fossil Balaenopteridae.......................... 251!
FIGURE 5.30. Bony labyrinth of Tursiops truncatus .................................................... 253!
FIGURE 5.31. CT slices through ear region of Tursiops truncatus ............................... 255!
FIGURE 5.32. Bony labyrinth of Equus caballus .......................................................... 262!
FIGURE 5.33. CT slices through ear region of Equus caballus..................................... 264!
FIGURE 5.34. Bony labyrinth of Canis familiaris......................................................... 271!
FIGURE 5.35. CT slices through ear region of Canis familiaris ................................... 273!
FIGURE 5.36. Bony labyrinth of Eumetopias jubatus................................................... 275!
FIGURE 5.37. CT slices through ear region of Eumetopias jubatus.............................. 277!
FIGURE 5.38. Bony labyrinth of Felis catus ................................................................. 279!
FIGURE 5.39. CT slices through ear region of Felis catus............................................ 281!
FIGURE 5.40. Bony labyrinth of Manis tricuspis.......................................................... 290!
FIGURE 5.41. CT slices through ear region of Manis tricuspis .................................... 292!
FIGURE 5.42. Bony labyrinth of Pteropus lyelli ........................................................... 297!
FIGURE 5.43. CT slices through ear region of Pteropus lyelli...................................... 299!
FIGURE 5.44. Bony labyrinth of Nycteris grandis ........................................................ 304!
FIGURE 5.45. CT slices through ear region of Nycteris grandis................................... 306!
FIGURE 5.46. Bony labyrinth of Rhinolophus ferrumequinum..................................... 308!
FIGURE 5.47. CT slices through ear region of Rhinolophus ferrumequinum ............... 310!
FIGURE 5.48. Bony labyrinth of Tadarida brasiliensis ................................................ 312!
FIGURE 5.49. CT slices through ear region of Tadarida brasiliensis........................... 314!
FIGURE 5.50. Bony labyrinth of Atelerix albiventris.................................................... 322!
FIGURE 5.51. CT slices through ear region of Atelerix albiventris .............................. 324!
FIGURE 5.52. Bony labyrinth of Sorex monticolus ....................................................... 326!
FIGURE 5.53. CT slices through ear region of Sorex monticolus.................................. 328!
xvi
FIGURE 5.54. Bony labyrinth of Mus musculus............................................................ 337!
FIGURE 5.55. CT slices through ear region of Mus musculus ...................................... 339!
FIGURE 5.56. Bony labyrinth of Cavia porcellus ......................................................... 341!
FIGURE 5.57. CT slices through ear region of Cavia porcellus.................................... 343!
FIGURE 5.58. Bony labyrinth of Lepus californicus..................................................... 349!
FIGURE 5.59. CT slices through ear region of Lepus californicus................................ 351!
FIGURE 5.60. Bony labyrinth of Sylvilagus floridanus................................................. 353!
FIGURE 5.61. CT slices through ear region of Sylvilagus floridanus ........................... 355!
FIGURE 5.62. Bony labyrinth of Macaca mulatta......................................................... 361!
FIGURE 5.63. CT slices through ear region of Macaca mulatta ................................... 363!
FIGURE 5.64. Bony labyrinth of Homo sapiens............................................................ 365!
FIGURE 5.65. CT slices through ear region of Homo sapiens ...................................... 367!
FIGURE 5.66. Bony labyrinth of Cynocephalus volans................................................. 373!
FIGURE 5.67. CT slices through ear region of Cynocephalus volans ........................... 375!
FIGURE 5.68. Bony labyrinth of Tupaia glis ................................................................ 379!
FIGURE 5.69. CT slices through ear region of Tupaia glis ........................................... 381!
FIGURE 5.70. Bivariate plots of labyrinth dimensions over body mass........................ 389!
1
CHAPTER 1: INTRODUCTION TO THE EAR OF MAMMALS
The otic (ear) capsule is part of the special sensory system of the nervous system
of vertebrates, and one of three major sensory capsules of the head, along with the optic
(for vision) and olfactory (for smell) capsules. The function of the ear is two-fold –
hearing and balance. The organs of hearing and balance within the ear of vertebrates are
tiny structures. In humans, they are contained in cavities totaling only around 165 mm
3
in
volume. Despite their small size, the organs of the ear are powerful. It is amazing that
such minute structures can cause a myriad of problems from tinnitus to motion sickness
to a general lack of balance. These reasons may explain why human beings are interested
in the physiology and morphology of the ear region.
GROSS ANATOMY OF THE MAMMALIAN EAR
The generalized ear of mammals is partitioned into three sections – the outer,
middle, and inner ears. The outer ear consists of the external pinna, which is prominent in
most mammals, and it leads to the external auditory meatus and associated canal. The
boundary between the external and middle ear is the tympanic membrane. The middle ear
(tympanic) cavity is bounded dorsally by cranial elements, particularly the petrosal bone,
and ventrally by an auditory bulla, which typically is ossified in extant mammals (see van
der Klauuw, 1931; Novacek, 1977). The middle ear ossicles (malleus, incus, and stapes)
are contained within the tympanic cavity, and they form a chain extending between the
tympanic membrane and petrosal. The stapes articulates with a window in the petrosal
called the fenestra vestibuli, which serves as one of two points of communication
between the middle and inner ear cavities. The other opening is the fenestra cochleae,
which is covered by a secondary tympanic membrane that accommodates expansion of
2
the inner ear space during changes in pressure. The petrosal itself separates the majority
of the tympanic cavity from the cranial cavity.
The inner ear consists of a set of interconnected spaces within the petrosal known
as the bony labyrinth (Figure 1.1), which contains a series of soft-tissue sacs and ducts,
known as the membranous labyrinth. The membranous labyrinth is separated into an
inferior division that includes the cochlear duct (containing the spiral organ of hearing)
and saccule of the vestibule (containing receptors sensitive to linear motion), and a
superior division that includes the utricle of the vestibule, the anterior, lateral, and
posterior semicircular ducts and ampullae, and the common crus between the anterior and
posterior ducts (all of which are involved with detecting rotational movement of the
head). The osseous semicircular canals and cochlea of the bony labyrinth mirror the
shape of the membranous ducts within, although the bony canals may not accurately
reflect the size of the ducts (Curthoys et al., 1977b).
The bony cochlear canal is divided into the scala tympani that communicates with
the fenestra cochleae, and the scala vestibuli that terminates at the fenestra vestibuli. The
division is formed by a bony primary spiral lamina that curves along the modiolus
(central bony pillar around which the cochlea coils) on the axial wall of the cochlea
(Figure 1.1C-D). A secondary lamina often mirrors the primary lamina for a short
distance on the opposing (radial) wall of the cochlea. The two laminae are connected by
the basilar membrane (the laminae do not contact each other directly), upon which the
spiral organ of hearing sits. The basilar membrane defines the tympanic wall of the
membranous cochlear duct (also known as the scala media). The vestibular membrane
crosses the width of the scala vestibuli to complete the cochlear duct at its vestibular
edge. A small opening known as the helicotrema is situated at the apex of the cochlea,
and it serves as a connection between the scalae tympani and vestibuli. The cochlear duct
3
FIGURE 1.1. Gross anatomy of the inner ear. A, bony labyrinth in anterior view (dorsal
towards top, medial towards left), where dashed line represents a slice through the
cochlea presented in C of this figure; B, bony labyrinth in lateral view (dorsal towards
top, anterior towards left); C, cross-section through entire cochlea (apex towards top) as
indicated by dashed line in A; D, cross-section through cochlear canal and cochlear duct.
Abbreviations: aa, anterior ampulla; ac, anterior semicircular canal; av, aqueduct of the
vestibule (passage of endolymphatic duct); bm, basilar membrane; cc, canaliculus
cochleae (passage of perilymphatic duct); cd, cochlear duct; cn, canal for cranial nerve
VIII, within modiolus (axis); co, cochlea; cr, common crus; er, elliptical recess; fc,
fenestra cochleae; fv, fenestra vestibuli; la, lateral ampulla; lc, lateral semicircular canal;
mo, modiolus; pa, posterior ampulla; pc, posterior semicircular canal; pl, primary bony
lamina; pt, petrosal bone; sl, secondary bony lamina; sr, spherical recess; tc, scala
tympani of the cochlea; vc, scala vestibuli of the cochlea; vm, vestibular membrane.
A B
C D
co
fv
fc
cc
fv
co
vc
vc
vc
vc
vc
tc
tc
tc
tc
sl
sl
sl
sr
sr
er
er
aa
aa
la
la
av
av
ac
ac
lc
lc
cr
cr
pc
pc
pt
co
mo
mo
co
cn
cd
vm
pl
bm
4
is filled with endolymph, and the surrounding space, which includes both the scala
tympani and scala vestibuli, is filled with perilymph. The perilymphatic duct exits the
inner ear near the fenestra cochleae through a bony passage known as the canaliculus
cochleae.
The bony vestibule is divided into the spherical recess inferiorly and the elliptical
recess superiorly. The recesses correspond loosely to the saccule (spherical recess) and
utricle plus semicircular ducts (elliptical recess), but the shapes of the membranous sacs
are preserved minimally within the bony vestibule. The saccule, utricle, and semicircular
ducts are filled with endolymph (exchange occurs between the cochlea and saccule at
their junction), and perilymph fills the remainder of the space. Varying amounts of
perilymph surround the semicircular ducts in different species (Gray, 1906, 1907, 1908),
and the endolymphatic duct exits the inner ear from the medial edge of the labyrinths,
passing through the bone and opening into the cranial cavity.
BIOLOGICAL SIGNIFICANCE OF THE EAR
Audition was the first function of the ear of humans to be recognized by early
anatomists. The role of the ear in hearing was deduced from injuries to the temporal
region of the head in ancient Egypt, as recorded in the Edwin Smith Surgical Papyrus
(Hawkins, 2004), but the specific role that the cochlea itself played in the sense of
hearing was not recognized until the mid 17th century AD, as reported by physician
Thomas Willis (Hawkins, 2004). Sensations of orientation were attributed to the
semicircular canals and vestibular system in the late 19th century (Dercum, 1879), and
functional similarities of the canals and the lateral line system in fishes and amphibians
were recognized at this time as well (Lee, 1898). Further advancements were made from
5
the middle of the 19
th
through early 20th centuries in comparative anatomy of broad
samples of vertebrate species through dissections and histological sectioning (Retzius,
1884), corrosion casting (Hyrtl, 1845), and extraction of the intact membranous labyrinth
from the surrounding bone (Gray, 1903, 1905, 1907), all of which added significant
contributions to our understanding of the structure and function of the inner ear.
From a functional standpoint alone, the auditory and vestibular structures are
significant agents in vertebrate biology. Hearing certainly played an important role in
early tetrapod evolution, where newly terrestrial animals moved from detecting water-
borne to air-borne sound waves (Manley, 1972; Clack, 2002). In the case of early
mammals, which were hypothesized to be nocturnal (Kielan-Jaworowska et al., 2004), a
reliance on non-visual senses to navigate the Mesozoic world would have been necessary.
The hearing capabilities within Mammalia are varied, and mammals can hear in a great
range of frequencies. Proboscideans are sensitive to very low frequencies (Payne et al.,
1986; Poole et al., 1988), and cetaceans as a group can hear in both subsonic and
ultrasonic ranges (Ketten, 1997). Microchiropteran bats use ultrasonic echolocation
during prey capture (Simmons et al., 1979), and even tenrecs (Gould, 1965) and soricid
lipotyphlans (Tomasi, 1979) are known to vocalize using ultrasonic noises.
Balance, and closely related equilibrium, was an important sense in vertebrate
history as well. When tetrapods left an aquatic environment for land, they faced unique
stability difficulties (Alexander, 2002), and the skeletons of tetrapods were well adapted
to the new terrestrial lifestyle (Shubin et al., 2004; Boisvert, 2005). In fact, the vestibular
end organs are similar in structure to, and likely derived from, the lateral line system of
early gnathostomes (Lee, 1898). Balance is an integral aspect of vertebrate locomotion,
and a broad range of mobility and agility is observed within Mammalia, from sure-footed
6
gazelles fleeing a sprinting cheetah, to sluggish koalas resting in eucalyptus trees (see
Spoor et al., 2007).
However, the physiological capability of the inner ear is not the only significance
possessed by the otic region. The potential that the inner ear region could be used to infer
the evolutionary relationships of animals was not lost on late 19th and early 20th century
otologists, particularly Albert Gray, who stated that “…the labyrinth must, of course,
have some value when considering these relationships, just as have the teeth, the
skeleton, and other parts of the body” (Gray, 1907, p. 2-3). This observation led to
several systematic surveys of the external surface of the petrosal and middle ear cavities
across a range of mammal taxa (Van Kampen, 1904; van der Klaauw, 1931; MacPhee,
1981), but broad phylogenetic surveys of the bony labyrinth are lacking.
Beyond the functional and phylogenetic significance of the ear region in extant
taxa, the petrosal bone serves as a resource in understanding the phylogenetic
relationships among extinct mammals. The petrosals are more robust and dense than
other cranial elements, and the otic region is the sensory system that is best preserved in
the fossil record, especially for mammals and their fossil relatives (e.g., Archibald, 1979;
Quiroga, 1979; Miao, 1988). Therefore, the majority of sensory physiology information
of fossil vertebrates revolves around the ear region, and the morphology of the inner ear
is used to make functional (Spoor et al., 1994; Clack, 2002; Witmer et al., 2003; Alonso
et al., 2004; Clarke, 2005) and phylogenetic (Hunt, 1987; Luo and Gingerich, 1999;
Ekdale et al., 2004) interpretations of fossil taxa.
Given the functional and phylogenetic significance of the ear region, it is not
surprising that the otic region is one of the most studied anatomical complexes. However,
one aspect of the ear region that often is neglected involves gross morphological variation
of the system (exceptions include the work of Caix and Outrequin, 1979; Muren et al.,
7
1986; Dimopolous and Muren, 1990). Observed variation can be the result of many
factors. Perhaps the most obvious influence is physiology. Differently shaped
semicircular canals may indicate different lifestyle habits (Georgi and Sipla, 2008). But
other phenomena also may play a role, such as ontogeny, which has had a light treatment
in published literature (Ryals et al., 1984; Jeffery and Spoor, 2004; Sánchez-Villagra and
Schmelzle, 2007).
A final factor that has received little attention is the phylogenetic significance of
the bony labyrinth, which can be presumed given that the external surface of the petrosal
is informative. Potential phylogenetically useful features within the cochlea and vestibule
were previously identified (Meng and Fox, 1995; Spoor and Zonneveld, 1998), but few
broad taxonomic surveys have been completed with the sole intent of determining
phylogenetic relationships based on morphology of the inner ear. The following chapters
of this dissertation present the results of my investigations into the sources of variation
observed within the inner ear cavities of mammals, ultimately placing the anatomy of the
bony labyrinth into a phylogenetic context.
DISSERTATION OVERVIEW
Chapter 2 is focused on a description of an unparalleled sample of isolated
petrosal bones referred to extinct elephantoid proboscideans from a Pleistocene cave
deposit in central Texas. The inner ear of extinct elephantids is described for the first
time, and variation observed on the external surface of the petrosal across the sample is
discussed, namely completion of the floor of the bony aquaeductus Fallopii for branches
of the facial nerve. The variation observed within the aqueductus Fallopii can be
8
explained by ontogenetic changes in the ear region, and an ossification sequence for the
bone surrounding the facial nerve through the petrosal is identified in this fossil sample.
An investigation into morphological changes in the walls of the inner ear forming
the bony labyrinth over maturation of the otic region in the extant marsupial Monodelphis
domestica is provided in Chapter 3. A few dimensions of the semicircular canals are
correlated to the maturity of an individual, but the majority of features examined are
independent of age. The results of Chapter 3, along with published ontogenetic sequences
of the otic region of humans (Jeffrey and Spoor, 2004) and rabbits (Hoyte, 1961) signify
that researchers utilizing the anatomy of the inner ear probably do not need to be
concerned with the maturity of an individual when using the bony labyrinth in
phylogenetic or physiological studies (as long as the walls of the bony labyrinth are fully
ossified).
Chapter 4 is a comparison of the bony labyrinths of fossil eutherian mammals
from the Cretaceous, particularly zhelestids from Uzbekistan. Phylogenetic analyses
place a monophyletic clade of zhelestid taxa from the Bissekty Formation either inside or
outside crown Placentalia, and the position of the Bissekty zhelestid grouping on the
mammalian tree plays a role in the timing of the origin of placentals. The morphology of
the bony labyrinth of the Bissekty zhelestids is described in Chapter 4, contributing an
essential body of knowledge to what is known about these taxonomically important
mammals. In addition, variation within the fossil sample is identified and quantified. The
degree of variation observed in the zhelestid sample is consistent with that observed in
single extant mammal species. This result indicates that certain features thought to be
important in phylogenetic and functional interpretations utilizing the ear region do not
vary much within a species or among closely related taxa.
9
The final section, Chapter 5, includes a broad scale comparison of the bony
labyrinths of placental mammals, where variation is observed and described across the
sample. Correlations between dimensions of the inner ear are identified, as well as
potential relationships between the morphology of the bony labyrinth and function. For
example, the ratio of the radius of the arc formed by a semicircular canal to the body
mass of the individual is significantly smaller in aquatic mammals than it is in their
terrestrial relatives. In particular, the proportion that the vestibule and associated
semicircular canals contributes to the overall volume of the bony labyrinth is distinctly
smaller in Cetacea than that calculated for any other mammal. Although a reduced
vestibular system in cetaceans may correlate to a fully aquatic lifestyle, the feature is
reconstructed as a synapomorphy for Cetacea.
A secondary common crus between the lateral and posterior semicircular canals is
ancestral for Theria, but lost in Placentalia (although a secondary crus is present in a
small number of taxa). The cochlea of most crown placentals coils to a greater degree
than that measured for Cretaceous eutherians (including the zhelestids from the Bissekty
Formation of Uzbekistan), as is exemplified by the four turns observed in the cochlea of
the guinea pig, Cavia porcellus (as opposed to the single coil observed in Zalambdalestes
from the Cretaceous of Mongolia). In fact, the conical shape of the highly-coiled cochlea
of the guinea pig is an attribute that only is observed within the inner ears of caviomorph
rodents.
10
CHAPTER 2: MORPHOLOGY AND VARIATION IN THE
PETROSAL OF EXTINCT ELEPHANTOIDEA (PROBOSCIDEA)
FROM CENTRAL TEXAS
ABSTRACT
An unparalleled sample of isolated petrosal bones assigned to elephantoid
Proboscidea was recovered from Pleistocene deposits in Friesenhahn Cave, Bexar
County, TX. Morphology of the middle and inner ear of the elephantoids is described and
variation within the sample is identified. The variations observed include the stapedial
ratio, completeness of the aquaeductus Fallopii, and connection of the crista
interfenestralis to the tympanohyal on the posterior aspect of the petrosal to form a
foramen for passage of the stapedius muscle. The observed morphological differences
may be the result of the taphonomic history of the specimens or else taxonomic
differences as two species of proboscidean have been identified in the deposits of
Friesenhahn Cave, but there are no unambiguous characters to distinguish between them.
However, the morphology of the aquaeductus Fallopii supports an ontogenetic
explanation for some variation, and a sequence of ossification surrounding the
aquaeductus Fallopii, from the anterior end of the canal to the posterior, is hypothesized.
A broad range of stapedial ratios is calculated across the sample, and this might have
dramatic effect on how these characters apply to mammalian systematics.
INTRODUCTION
The otic region of mammals as preserved on and within the petrosal bone garners
a great deal of interest in vertebrate morphology and physiology. The ear is one of the
special sensory organs of the nervous system, and the structures of the inner ear serve two
11
functions: hearing and balance. The morphology of the bony walls of the middle ear
cavity, which is preserved on the external surface of the petrosal, is phylogenetically
informative for many mammal groups (e.g., Cifelli, 1982; Wible and Novacek, 1988;
Hunt, 1989; Geisler and Luo, 1996), including elephants and their closest relatives
(Fischer and Tassy, 1993). Furthermore, the petrosal bones are dense and robust relative
to the rest of the skull, and they are readily preserved in the fossil record. Petrosal bones
contribute significantly to our knowledge of Mesozoic mammal faunas, and the auditory
region of early mammals and their relatives are among the best-preserved and most
studied skeletal elements (Olson, 1944; Archibald, 1979; Quiroga, 1979; Miao, 1988).
Because of this, the ear region provides paleontologists with many clues to the biology of
early mammals and their relatives (Graybeal et al., 1989; Meng and Wyss, 1995).
Although the petrosal contribution to the middle ear cavity is well-studied and is
used in a myriad of phylogenetic studies, thorough discussions of intraspecific
morphological variation in the system within a species are lacking in paleontological
literature (some variation discussed in Ekdale et al., 2004). Studies focusing upon
variation within the ear region largely are restricted to biomedical studies of the
physiology of the inner ear (e.g., Caix and Outrequin, 1979; Muren et al., 1986; Clark
and Smith, 1993). Ontogeny, or the development of an individual from conception to
ultimate death, is a significant source of variation in chordate anatomy (e.g., Wiens et al.,
2005), and the ontogenetic development of the ear region has been described for several
mammals (McCrady, 1938; Jeffrey and Spoor, 2004). Although ontogenetic variation can
cause problems in phylogenetic analyses (Brinkman, 1988; Tykoski, 2005), ontogeny
rarely is considered when scoring characters for phylogenetic analysis (see Brochu, 1996;
Colbert and Rowe, 2008).
12
The purpose of the present study is two-fold. The first goal is to provide a
thorough description of the ear region of the elephantoid proboscideans, which is lacking
in the scientific literature, and to compare the ear of elephantoids to that of other
proboscideans. Second, the sample provides a unique opportunity to study variation
within the middle ear region of a population of extant proboscideans. In fact, very few
studies have discussed variation within the middle ear region of any mammal species.
The comparative morphology of mammoth and mastodon petrosals is discussed, and a
hypothesized ontogenetic sequence is identified in the sample through an investigation of
the variation expressed by the bones. The present discussion is the first account of an
ontogenetic sequence identified for any fossil petrosal sample.
MATERIALS AND METHODS
Formation and history of Friesenhahn Cave
The examined petrosals were collected from Friesenhahn Cave in Central Texas
(Graham, 1976). The cave, which is located north of San Antonio, Texas (Bexar County)
on the eastern edge of the Edward’s Plateau, is one of at least 37 caves on the plateau
preserving Pleistocene faunas (Lundelius and Collins, 1999). The limestones of the
Edward’s Plateau are conducive to cave formation, and the caves are important from a
paleontological perspective (Graham, 1976). Caves in general provide a sheltered
environment that facilitates fossilization of vertebrate bones (George et al., 2007), and
Friesenhahn Cave is the most studied cave preserving Pleistocene faunas on the Edward’s
Plateau, and the first cave of its kind in Central Texas to undergo a thorough excavation
(see Evans, 1961).
13
The modern entrance of Friesenhahn Cave is a vertical shaft that formed around
10,000 years ago. An earlier entrance that was open 17,000-20,000 years ago, allowed for
the rapid accumulation of vertebrate remains, including proboscideans (Graham, 1976).
The largest bodied vertebrates recovered from the cave are extinct elephantoid
proboscideans that are identifiable as Mammut americanum and Mammuthus colombi
based on dentition (Graham, 1976). The Elephantoidea (group containing the ancestor of
Mammuthus and Mammut and all of its descendents; Lambert and Shoshani, 1998) are
represented in the cave by cranial (including isolated petrosals), dental, and postcranial
material, with the majority of material assigned to Mammuthus by Graham (1976). An
ontogenetic sequence of individuals represented by teeth, following the methods of Laws
(1966), was recognized by Graham (1976). Given the paucity of adult proboscidean
specimens in the cave (the oldest individual was estimated at four years old upon death;
Graham, 1976), it is hypothesized that the carcasses of juvenile mammoths likely were
dragged to the cave by carnivoran predators (Graham, 1976). That hypothesis is
supported by large carnivoran tooth marks (attributed to Homotherium serum; Meade,
1961) observed on the postcrania of large herbivores found in the cave.
Petrosal specimens and identification
Sixty-five isolated petrosal bones assigned to Mammuthus by Graham (1976) are
among the proboscidean specimens collected from Friesenhahn Cave. All of the petrosals
are housed at the Vertebrate Paleontology Laboratory of the Texas Natural Science
Center (TMM) in Austin, TX. The sample includes 37 right petrosals and 28 left
petrosals, indicating a minimum number of 37 individuals. In all, the collection of
petrosals provides an unprecedented opportunity to study the morphology of the ear
14
region of extinct proboscideans, upon which only a few studies have been focused (see
Court, 1992b).
No criteria were given for the initial referral of the petrosals to Proboscidea (see
identifications by Graham, 1976), although the decision to do so likely was based on the
size of the bones. The petrosals are large, and because mammoths and mastodons are the
largest bodied mammals recovered from the cave, it is reasonable to assume that the
petrosals are in fact from proboscideans. Further examination of the specimens confirms
the initial assessment through the identification of an apomorphic feature on the
petrosals. The fenestra cochleae is confluent with the canaliculus cochleae to form a
secondarily undivided perilymphatic foramen in all of the specimens examined here
(Figure 2.1). Among Quaternary mammals (both extant and extinct), this feature is only
observed in proboscideans and sirenians (Fischer, 1990; Court, 1994). Because sirenian
material has never been recovered from any Pleistocene cave deposit on the Edward’s
Plateau (owing to the complete lack of Pleistocene marine deposits on the plateau), the
referral of the petrosals to Proboscidea is accepted.
Within Proboscidea, the petrosals were referred to Mammuthus by Graham
(1976). Again, no criteria were provided, but the referral likely was made based on the
overabundance of Mammuthus material versus Mammut. Around 97% of proboscidean
teeth from the cave (500 out of 513) were assigned to Mammuthus by Graham, but it is
possible that one or more of the petrosals represent Mammut. In fact, by simple
probability, two petrosals out of the sample should be Mammut. Based on the 13 teeth
that Graham assigned to Mammut, there is a minimum number of three mastodon
individuals, and a maximum number of 13.
15
FIGURE 2.1. Three-dimensional CT reconstructions and labeled line drawings of
petrosal of elephantoid from Friesenhahn Cave (TMM 933-950). A-B, cerebellar view;
C-D, medial view; E-F, tympanic view. Abbreviations: af – floor of aqueductus Fallopii;
ant – anterior; av – aquaeductus vestibuli; ci – crista interfenestralis; dor – dorsal; ew –
epitympanic wing; ff – foramen faciale; fm – fossa musculus minor; fv – fenestra
vestibuli; hf – hiaus Fallopii; im – internal auditory meatus; ip – sulcus for inferior
petrosal sinus; lat – lateral; med – medial; pf – perilymphatic foramen; pr –
promontorium; sf – sulcus facialis; sp – sulcus for perilymphatic duct; th – tympanohyal.
A B
C D
E F
av
ip
ip
im
im
1 cm
1 cm
th
lat
ant
th
pf
pf
sp
pr
pr
fv
ew
ci
ci
fm
hf
af
ff
sf
dor
ant
med
ant
16
Unfortunately, there is no published information that discusses the anatomical
differences between mammoth and mastodon petrosals, if any differences exist. An
attempt is made here by comparing a sample of definitive Mammut petrosals collected
from Pleistocene deposits from Boney Spring (BS) in the Western Ozark Highland,
Missouri (housed at Illinois State Museum in Springfield, IL). Only proboscideans
referred to Mammut americanum have been recovered from Boney Spring, so the
taxonomic identity of this petrosal sample is more certain than that collected from
Friesenhahn Cave. As is observed for Friesenhahn Cave, a growth series of
proboscideans is identifiable in the fossil sample from Boney Spring (Saunders, 1977).
Although the examination of the Friesenhahn Cave and Boney Spring samples is far from
thorough, the comparison is a first attempt at investigating the ear regions of mammoths
and mastodons, which may allow a finer identification of the petrosals from Friesenhahn
Cave or other localities from which both mammoths and mastodons are known.
Computed tomography and measurement methods
The morphology of the bony labyrinth of the inner ear contained within the
petrosal is described for the first time for Elephantoidea. The bony labyrinth of the inner
ear is contained within cavities entirely surrounded by the petrosal bone itself. In order to
observe the internal auditory structures, the surrounding bone must be removed. This is
either accomplished through physical destruction of the bone, or it can be accomplished
digitally through high-resolution X-ray computed tomography (CT), as was performed
here. The only other data from the inner ear of extinct proboscideans comes from two
Eocene taxa, Moeritherium and Numidotherium (Court, 1992b). Internal auditory
structures of an elephantoid from Friesenhahn Cave were imaged using high resolution
X-ray computed tomography (CT) at the University of Texas High-Resolution CT facility
17
(UTCT). A single petrosal (TMM 933-950) was CT scanned, and a digital endocast of the
bony labyrinth of the inner ear was constructed in Amira 3.1 © computer software. The
field of reconstruction for the CT scan was 53 mm, and each CT slice has a resolution of
1024 by 1024 pixels. The interpixel spacing (distance between pixels) is 0.052 mm, and
the interslice spacing (between slices) is 0.13 mm. A total of 275 slices were collected
through the petrosal. The digital endocast was compared to published reports of the inner
ear of other proboscideans.
Anatomical terminology follows MacPhee (1981) for external petrosal anatomy,
and Sisson and Grossman (1938) and Evans (1993) for internal anatomy. Methodologies
for measurements of the inner ear follow Spoor and Zonneveld (1995) mostly, although
two specific measurements, the stapedial ratio and coiling of the cochlea, follow Segall
(1970) and a modified method of Geisler and Luo (1996) respectively. The stapedial ratio
is an index of the shape of the footplate of the stapes that is calculated as the greatest
length of the footplate divided by the greatest width perpendicular to the length. In the
absence of the stapes, dimensions of the fenestra vestibuli were used as a proxy here
(following Segall, 1970).
The method employed to obtain the number of turns completed by the cochlea
follows the method utilized by Geisler and Luo (1996), and is comparable to that
conceived by West (1985). Both methods orient the cochlea so the field of view points
down the axis of rotation. A line is drawn from the center of the axis of rotation to a
secondary landmark near the basal end of the cochlea, and the number of times that the
cochlea crosses this line is counted as one half turn. An additional angle of rotation is
added to this value if the cochlea passes beyond a half turn. The methods differ in their
choice of a secondary landmark. One landmark that often is used is the point of inflection
between the cochlea and vestibule (West, 1985). Although that method commonly is used
18
(e.g., Meng and Fox, 1995), the landmark at the inflection is arbitrary and difficult to
locate. A different landmark is located at the proximal end of the basilar membrane
(Geisler and Luo, 1996). The organ of hearing, which serves as the functional unit of the
cochlea, rests on the basilar membrane in life. I used the proximal end of the basilar
membrane as a landmark because it is more biologically appropriate (given the functional
importance of the basilar membrane) and less ambiguous than the former (utilized by
West, 1985).
ANATOMICAL OBSERVATIONS OF FRIESENHAHN CAVE PETROSALS
All of the petrosals from Friesenhahn Cave examined are isolated without any
evidence of a tympanic bulla, nor any other cranial associations. Because of this, the
orientations described here are only approximate. In general, the cerebellar (intercranial)
surface of the petrosal faces dorsomedially and the tympanic surface is oriented ventrally
when the petrosal is articulated with the rest of the skull (MacIntyre, 1972).
Cerebellar surface of petrosal
The cerebellar surface is slightly convex in all of the specimens, giving the
petrosal an almost tubular appearance in this view (Figure 2.1A). There is no evidence of
a subarcuate fossa, which holds a parafloccular lobe of the cerebellum in life in most
mammals. Rather, the petrosal is inflated in the area where the fossa is located in other
taxa. The opening of the internal auditory meatus is oriented towards the anterior apex of
the petrosal rather than dorsomedially as is the case for most non-proboscidean
mammals, and the opening is parallel to the long axis of the bone. The sulcus for the
inferior petrosal sinus extends along the medial edge of the petrosal in an anterior-
posterior direction (Figure 2.1A-2.1B). The fissure-like opening for the aquaeductus
19
vestibuli is positioned at the posterior end of the petrosal, posterolateral and dorsal to the
posterior terminus of the sulcus for the inferior petrosal sinus.
Tympanic surface of petrosal
The anterior aspect of the tympanic surface is known as the promontorium, which
houses the cochlea internally (Figure 2.1C). The bulbous promontorium is irregularly
shaped with a distinct depression anteromedially. A broad epitympanic wing extends
anteriorly from the promontorium. No sulci for the internal carotid or stapedial arteries
are observed on the promontorium of any of the petrosals from Friesenhahn Cave.
The perilymphatic foramen (joining of the fenestra cochleae and canaliculus cochleae) is
an irregular, pear-shaped opening on the posteromedial edge of the promontorium. The
medial aspect of this opening leads to a depression that likely accommodated the
perilymphatic duct in life.
The fenestra vestibuli on the posterolateral edge of the promontorium is round to
oval in shape. Two specimens (TMM 933-950 and 933-951) preserve the footplate of the
stapes was preserved for. The footplate of TMM 933-951 is preserved in articulation
within the fenestra vestibuli, but the structure had come loose in TMM 933-950 and
fallen through the fenestra and into the inner ear cavities. The footplate of TMM 933-950
was carefully removed with forceps and examined (Figure 2.2). The stapedial ratio of the
stapes itself is slightly less than the ratio calculated for the fenestra vestibuli on the same
specimen (1.7 versus 1.8). The range of stapedial ratios among the full petrosal sample is
1.4-2.1 (Table 1), with an average of 1.8 and a standard deviation of 0.1.
The footplate of the stapes of TMM 933-950 is concave outwards towards the
crura, which are broken, and only their attachments to the footplate remain. Because there
are bases for two crura, a stapedial foramen would have been present in a complete
20
FIGURE 2.2. Photograph and line drawing if stapes of elephantoid from Friesenhahn
Cave (TMM 933-950). A-B, lateral view, shaded area reconstructed in B; C-D, map
view, shaded area reconstructed in D. Abbreviations: as – anterior crus of stapes; fs –
footplate of stapes; ps – posterior crus of stapes.
A
B
C
D
ps
ps
as
as
fs
fs
fa
fa
1 mm
21
TABLE 2.1. Variation observed among elephantoid petrosals from Friesenhahn Cave
(Bexar County, Texas).
Specimen No.
Aquaeductus Fallopii
Stapedius Muscle Foramen
Stapedial Ratio
TMM 933-12
partial
-
1.7
TMM 933-69
-
-
1.6
TMM 933-166
partial
incomplete
1.6
TMM 933-414
-
complete
1.9
TMM 933-415
-
-
2.0
TMM 933-416
-
complete
1.8
TMM 933-507
partial
-
1.8
TMM 933-508
partial
-
1.9
TMM 933-548
partial
-
1.9
TMM 933-746
-
-
1.6
TMM 933-747
-
-
1.7
TMM 933-950
complete
-
1.8
TMM 933-951
complete
-
1.9
TMM 933-953
-
-
1.8
TMM 933-954
-
-
1.7
TMM 933-955
complete
complete
1.6
TMM 933-956
-
-
1.7
TMM 933-957
complete
-
2.1
TMM 933-959
-
-
1.7
TMM 933-1032
complete
incomplete
1.9
TMM 933-1033
-
-
2.0
TMM 933-1034
-
-
1.6
TMM 933-1035
partial
-
1.9
TMM 933-1036
complete
-
1.7
TMM 933-1078
-
-
1.9
TMM 933-1317
partial
-
1.7
TMM 933-1389
-
-
1.9
TMM 933-1489
partial
complete
1.9
TMM 933-1533
complete
-
1.7
TMM 933-1677
-
-
1.6
TMM 933-1678
-
-
1.7
TMM 933-1679
-
-
1.7
TMM 933-2032
-
-
1.8
TMM 933-2043
-
-
1.7
TMM 933-2044
-
-
1.9
TMM 933-2223
-
-
1.6
TMM 933-2248
-
-
1.6
TMM 933-2249
complete
-
1.4
22
TABLE 2.1. Continued.
Specimen No.
Aquaeductus Fallopii
Stapedius Muscle Foramen
Stapedial Ratio
TMM 933-2323
partial
complete
1.8
TMM 933-2414
partial
-
1.7
TMM 933-2415
-
-
1.8
TMM 933-2634
complete
complete
1.6
TMM 933-2635
-
-
1.7
TMM 933-2637
-
-
1.6
TMM 933-2638
-
-
1.8
TMM 933-2918
-
-
1.5
TMM 933-2920
-
-
1.8
TMM 933-2940
partial
-
1.9
TMM 933-3467
-
-
2.0
TMM 933-3468
-
-
1.7
TMM 933-3838
-
-
1.7
TMM 933-5732
-
-
2.0
TMM 933-5724
-
-
1.6
TMM 933-5733
-
-
1.7
TMM 933-5735
-
complete
1.6
TMM 933-5866
-
-
2.0
Average
-
-
1.8
Standard
Deviation
-
-
0.1
23
stapes. The base of one crus is more robust than the other, and the overall shape of the
footplate is not a perfect ellipse, but rather is egg-shaped, although it is more or less
symmetrical along its long axis. When compared to the shape of the fenestra vestibuli, the
narrow end of the footplate is the anteriormost end, with the robust crus directed
posteriorly.
The aquaeductus Fallopii, a canal for transmission of branches of the facial nerve
(cranial nerve VII), is positioned lateral to the promontorium. The ventral surface of the
aquaeductus bears a distinct fissure where the medial and lateral walls of the floor meet
in several specimens (Figs. 1.3, 3; Table 2.1). The ventral floor of the canal is incomplete
with a gap between the medial and lateral aspects extending the entire length of the canal
in most of the specimens (Figure 2.3A). The canal is partially complete in several
specimens, having a noticeable gap between the posterior aspect of the medial and lateral
contributions of the canal, but with the anterior end fused (Figure 2.3B). Many of these
specimens are damaged near the aquaeductus Fallopii, but there is no observable damage
in several of the specimens. Furthermore, the medial sheet underlying the aquaeductus
Fallopii of TMM 933-548 bears a longitudinal groove along its lateral edge that likely
was a facet for the lateral sheet, which is not preserved. The floor of the aquaeductus
cochleae is complete in the remaining specimens (Figure 2.3C).
The sulcus facialis for the hyomandibular branch of the facial nerve extends
posteriorly from the foramen faciale (Figure 2.1C). This groove curves along the
posterolateral border of the promontorium. A complete stylomastoid foramen for passage
of the facial nerve is not present in any of the specimens. A sigmoidal sulcus is observed
on the ventral surface of the crista interfenestralis, which forms a robust process between
24
FIGURE 2.3. Variation within aquaeductus Fallopii of elephantoids from Friesenhahn
Cave, petrosals in tympanic view. A-B, floor of canal incomplete (likely result of
damage), both hiatus Fallopii and foramen faciale incomplete (TMM 933-1389); C-D,
floor of canal partially complete, hiatus Fallopii complete, foramen faciale
incomplete(TMM 933-166); E-F, floor of canal, hiatus Fallopii, and foramen faciale
complete (TMM 933-2043). Abbreviations: af – aquaeductus Fallopii; ci – crista
interfenestralis; ff – foramen faciale; fv – fenestra vestibuli; hf – hiatus Fallopii; pf –
perilymphatic foramen; pr – promontorium; th - tympanohyal.
A C E
B D F
pr
pr
pr
1 cm1 cm
af
af
hf
hf
ff
ff
af
fv
fv
fv
th
th
th
ci
ci
ci
pf
pf
pf
25
the foramina cochleae and vestibuli. This groove served as a facet for the tympanic bone
that would form an auditory bulla in an articulated skull. The crista interfenestralis and
tympanic bulla define the stylomastoid foramen in extant elephants (Eales, 1926), but
because no bullae are preserved for the Friesenhahn petrosals, the stylomastoid foramen
is incomplete in all specimens examined here.
Posteromedial to the crista interfenestralis is the fossa musculus minor for the
stapedius muscle. A sulcus for the muscle extends between the fossa and the fenestra
vestibuli. The crista interfenestralis on several of the petrosals contacts a process on the
posterior aspect of the tympanic surface of the petrosal known as the tympanohyal. The
connection between the ventral process of the crista interfenestralis and the tympanohyal
is incomplete in most specimens (Figure 2.4A; Table 2.1), but when the crista
interfenestralis and tympanohyal join, they form a foramen (Figure 2.4B; Table 2.1). In
specimens such as TMM 933-2323, the sulcus for the stapedius muscle leads to the
foramen, suggesting that the stapedius muscle passed through the foramen in life. The
posterior region of the petrosal is damaged in the incomplete specimens; however, two
exceptions are TMM 933-166 and 933-1032 wherein the crista interfenestralis and
tympanohyal nearly touch, but do not contact one another. There is no apparent damage
to account for the lack of contact in those specimens.
Inner ear
The bony labyrinth within the petrosal consists of the cochlea anteriorly and the
vestibule with its semicircular canals posteriorly (Figure 2.5). The cochlea completes a
little over two complete turns (765°) and is planispiral in shape. The diameter of the basal
turn of the cochlea (13.7 mm) is more than twice as large as the height of the cochlear
26
FIGURE 2.4. Foramen for stapedius muscle in elephantoids from Friesnehahn Cave,
petrosals in medial view. A, crista interfenestralis and tympanohyal not connected,
foramen for stapedius muscle incomplete (TMM 933-1932); B, crista interfenestralis and
tympanohyal connected, foramen for stapedius complete (TMM 933-416). Abbreviations:
ant – anterior; ci – crista interfenestralis; ew – epitympanic wing; fmm – fossa musculus
minor; pf – perilymphatic foramen; pr – promontorium; sp – sulcus for perilymphatic
duct; th – tympanohyal; ven – ventral.
A
B
pr
ew
ew
pr
ci
th
fmm
pf
sp
pf
sp
ant
ven
fmm
1 cm
th
ci
27
FIGURE 2.5. Digital endocast and labeled line drawing of bony labyrinth of inner ear of
elephantoid from Friesenhahn Cave (TMM 933-950). A-B, lateral view; C-D, tympanic
view; E-F, ventral view. Abbreviations: aa – anterior ampulla; ac – anterior semicircular
canal; av – aquaeductus vestibuli; cr – crus commune; co – cochlea; ed – endolymphatic
duct; er – elliptical recess; fv – fenestra vestibuli; la – lateral ampulla; lc – lateral
semicircular canal; ps – out pocket for perilymphatic sac; pa – posterior ampulla; pf –
perilymphatic foramen; pc – posterior semicircular canal; sr – recessus sphericus.
co
ps
pf
la
pa
pc
av
ant ant lat
dorlatdor
cr
ac
lc
A
C
E
fv
aa
er
ps
pf
pc
lc
ac
fv
la
sr
pa
co
pf
co
ps
ed
av
5 mm5 mm
pc
ac
cr
er
pa
lc
aa
B
D
F
28
spiral (6.6 mm). The cochlea as a whole contributes only 30% of the total volume of the
inner ear cavities (351 mm
3
out of 1145 mm
3
total). A curved out-pocketing for the
perilymphatic sac is anteriorly adjacent to the fenestra cochleae. There is little to no
development of a secondary bony labyrinth.
The medial edge of the fenestra vestibuli is the boundary between the cochlea and
vestibule, with the fenestra vestibuli completely in the vestibule. The endocast of the
vestibule has a sulcus that is laterally adjacent to the opening of the fenestra vestibuli.
This sulcus represents a bony ridge on the outer wall of the vestibule, and defines the
division between the spherical recess anteriorly, which houses the membranous saccule,
and the larger elliptical recess for the membranous utricle posteriorly. The spherical
recess communicates with the stapes via the fenestra vestibuli, and the recess is confluent
with the cochlea anteriorly and the elliptical recess posteriorly. A groove on the dorsal
wall of the spherical recess extends posteriorly. This groove, which is represented by a
rounded ridge on the endocast, accommodates the endolymphatic duct in life. The groove
extends towards the aquaeductus vestibuli, which opens ultimately onto the cerebellar
surface of the petrosal.
The elliptical recess receives a total of six openings in addition to the opening into
the spherical recess. Three of the openings are aligned along the border between the two
recesses. They are, from medial to lateral, the posterior ampulla, medial limb of the
lateral semicircular canal, and lateral ampulla. The opening for the anterior ampulla is
laterally adjacent to the lateral ampulla, and the opening for the short and squat common
crus is situated at the opposite end of the vestibule from the other four openings. The
sixth and final orifice is for the aquaeductus vestibuli, dorsal to the common crus and at
the terminus of the groove for the endolymphatic duct that begins in the spherical recess.
29
The lateral semicircular canal is the most planar of the three canals, and it opens
into the vestibule, separate from the posterior canal and ampulla. The posterior and
anterior semicircular canals join to form the crus commune. Among the three
semicircular canals, the posterior is the largest and the lateral is the smallest in terms of
arc radius of curvature (posterior = 5.51 mm, anterior = 4.99 mm, lateral = 2.67 mm, with
an average semicircular canal radius = 4.39 mm; measured following the method outlined
by Spoor and Zonneveld, 1995) and volume (posterior = 51.7 mm
3
, anterior = 48.9 mm
3
,
lateral = 22.4 mm
3
; measured in Amira ® software). However, the anterior canal is the
longest of the three (anterior = 24.6 mm, posterior = 24.3 mm, lateral = 12.5 mm;
ampullae not included, but length of the crus commune included for both anterior and
posterior lengths).
The ratio of the length of a semicircular canal to the radius of the respective canal
arc is thought to correlate to the rotational movement sensitivity of the head (Boyer and
Georgi, 2007). The ratios of the length of the semicircular canals to the radius of
curvature are similar for the anterior and lateral canals (anterior = 4.93, lateral = 4.70),
but the ratio is lower for the posterior canal (4.41). The planes of the semicircular canals
are not at right angles to one another. The plane of the anterior semicircular canal forms
acute angles with the planes of the lateral and posterior canals (74° and 66° respectively),
but the planes of the posterior and lateral canals form a slightly obtuse angle (92°). None
of the canals are perfectly planar, although the lateral canal fits onto a single plane better
than either of the other two canals.
30
ANATOMICAL OBSERVATIONS OF MAMMUT PETROSALS FROM BONEY
SPRING
All of the petrosals examined from Boney Spring are isolated without any
evidence of a tympanic bulla, as is observed in the sample from Friesenhahn Cave. The
morphology of the definitive Mammut sample from Missouri does not differ significantly
from that observed in the Friesenhahn elephantoid sample. For example, the subarcuate
fossa is absent from the cerebellar surface of Mammut, and a secondarily undivided
perilymphatic foramen is present on the tympanic surface of the petrosal. Out of the five
petrosals examined, only one specimen (144cBS71) preserves an intact aquaeductus
Fallopii. The ventral floor of the passage exhibits the partial condition in this specimen,
where the floor is fused anteriorly, but is open at the posterior end near the fenestra
vestibuli. Completion of the aqueductus could not be determined for the other four
specimens (right and left petrosals of 56BS71 and right and left petrosals of 361BS71).
The tympanohyal is damaged in every specimen; thus the completion of the
stapedius muscle foramen could not be determined. The sigmoid sulcus observed on the
ventral surface of the crista interfenestralis of the petrosals from Friesenhahn Cave is
observed on the petrosals from Boney Spring, excepting the left petrosal of 361BS71. A
groove for the perilymphatic duct extends medially from the undivided perilymphatic
foramen in the Boney Spring petrosals, as is observed for the sample collected from
Friesenhahn Cave. No stapedial footplates are preserved for the Boney Spring petrosals,
but the stapedial ratio was estimated for three of the five petrosals examined (average is
1.9 with a range of 1.9-2.0).
31
DISCUSSION
Comparison with other Proboscidea
Apart from Moeritherium and Numidotherium (Court and Jaeger, 1991; Court,
1992b, 1994), there is little information concerning the ear region of fossil proboscideans.
The petrosal is fused solidly to the skull in extant proboscideans as well as in
Mammuthus and Mammut specimens examined at the University of Texas at Austin. The
bony bulla covering the tympanic cavity, along with the massive size of proboscidean
skulls, makes the otic region of proboscideans difficult to study. Nonetheless, the isolated
petrosals of Elephantoidea individuals preserved in Friesnehahn Cave allow comparisons
among other proboscidean taxa.
The external petrosal surface of the elephantoids from Friesenhahn Cave agrees in
many respects with the ear regions of other non-elephantoid proboscideans. The most
obvious similarity is the confluence of the fenestra cochleae and canaliculus cochleae into
a common perilymphatic foramen. This structure also is observed in extant elephants, and
in fossils as early as the Eocene (Court, 1994), although separate in the Eocene
proboscidean Numidotherium (Court, 1992b, 1994). An undivided perilymphatic foramen
was hypothesized to be an adaptation for low frequency hearing by Court (1994). It was
demonstrated previously that extant elephants are specialized for low frequency hearing
(Payne et al., 1986; Poole et al., 1988). Given that an undivided perilymphatic foramen is
present in the Friesenhahn elephantoids, one might hypothesize that extinct elephantoids
were capable of hearing in low frequencies. However, a more thorough study of the bony
labyrinths of extinct proboscideans and the closest extant relatives of elephants (Sirenia;
Bininda-Emonds et al., 2007) is necessary to assess the auditory capabilities of
Elephantoidea as a whole in a sufficient manner.
32
Although a groove for the perilymphatic duct is medial to the perilymphatic
foramen in Mammut from Boney Spring and the elephantoids from Friesenhahn Cave,
such a sulcus is absent in the extant elephant Loxodonta (Fischer and Tassy, 1993) and
the extinct proboscidean Moeritherium (Court, 1994). The sulcus for the perilymphatic
duct is present in the extant sirenian Trichechus (Fischer and Tassy, 1993), but the groove
is considered to have derived independently in both taxa (for further discussion, see Court
and Jaeger, 1991; Court, 1994).
Another feature shared by the elephantoids from Friesenhahn Cave and other
proboscideans, namely Mammut from Boney Spring, Moeritherium (Court, 1994) and
Elephas (TMM M-6445), is the absence of the subarcuate fossa on the cerebellar surface
of the petrosal. A well-developed subarcuate fossa is associated with coordination of
head and eye movement, and even agility, if only indirectly (Jeffry and Spoor, 2006).
Extant elephants have low agility (‘medium slow’ as coded by Spoor et al., 2007), which
agrees with the absence of the subarcuate fossa. The subarcuate fossa was lost in
Proboscidea by the Eocene epoch (as exemplified by Moeritherium; Court, 1994)
suggesting that fossil proboscideans were not agile either.
Only the footplate of the stapes is preserved completely for the elephantoid
specimens from Friesenhahn Cave, but enough remains to compare to the stapes of extant
elephants. The broken pieces of the crura on the stapes indicate that the stapes was
perforated with a foramen, a condition seen in most mammals that possess a proximal
stapedial artery (Novacek and Wyss, 1986a; Wible, 1987). The stapes of the modern
elephant is perforated (Blair, 1710-1712b) as in the elephantoid proboscideans from
Friesenhahn Cave, suggesting that the proximal stapedial artery was present in these taxa.
However, a proximal stapedial artery is not described for any living proboscidean (at
least such a vessel was not mentioned by Blair, 1710-1712a, b, 1717-1719, Watson,
33
1874, or Eales, 1926). Furthermore, a sulcus for the stapedial artery is absent on the
promontorium of the Friesenhahn Cave and Boney Spring proboscideans. There is an
apparent contradiction since the stapedial foramen supports reconstruction of a proximal
stapedial artery through the ear region of Elephantoidea, but absence of a sulcus for the
stapedial artery medial to the fenestra vestibuli indicates an absence of the vessel.
There are two hypotheses for the reconstruction of the stapedial artery system
within the ear region of Mammuthus. First, the stapedial foramen developed in the
complete absence of a proximal stapedial artery. This accounts for absence of the sulcus
for the artery, as well as the lack of identification of the vessel in published accounts of
the elephant middle ear. Such a condition is observed in Cetacea where the proximal
stapedial artery is absent (Wible, 1987), but the stapes possesses a small foramen (the
‘massive-microperforate’ condition of Novacek and Wyss, 1986a).
A second hypothesis is that a proximal stapedial artery was present, penetrating
the stapes, but it did not leave a sulcus on the promontorium, and the vessel has remained
unreported. For the second scenario to be accepted, the proximal stapedial artery must
occur in at least one other species of mammal without leaving a trace on the petrosal,
which is observed in some lipotyphlans and bats (Wible, 1987). The stapes of either
Mammuthus or Mammut (it is unclear for which species the stapedial footplate is
preserved) likely was the typical bicrurate form, as it is in Elephas, and the lack of a
proximal stapedial artery in published descriptions of the ear region of elephants would
be the result of the difficulty in studying the otic system of Proboscidea. In fact,
descriptions of the otic vasculature in proboscideans are superficial. There is a great need
for a thorough investigation of the circulation pattern across the elephant basicranium.
Until such a study is completed, the presence or absence of a proximal stapedial artery for
Elephantoidea remains unresolved.
34
Several studies that were focused on the ear region of proboscideans discuss the
morphology of the bony labyrinth of the inner ear (Blair, 1717-1719; Hyrtl, 1845; Court,
1992b). Although not a description of otic morphology per se, the supplemental
information provided by Spoor et al. (2007) included dimensions of inner ear structures
for the extant elephant species, Loxodonta africana and Elephas maximus. Those
published measurements form a basis for a comparative discussion of the inner ear of
extinct and extant proboscideans.
The radii of the semicircular canals of the elephantoid from Friesenhahn Cave are
closer to those reported for extant Loxodonta than for Elephas (Table 2; Spoor et al.,
2007). A positive correlation between the radius of curvature of the anterior canal and
afferent sensitivity of that canal was found by Yang and Hullar (2007). Their results
indicated that the sensitivity of the anterior canal was higher than that of the smaller
lateral canal within an individual, as well as between the canals of different mammal
species. This result caused them to conclude that canals with larger radii are more
sensitive than those with smaller radii. However, mammals with large body sizes tend to
have large canal arc radii (see Spoor et al., 2007). If the specimen for which the bony
labyrinth was studied is Mammuthus, then the data for proboscideans reported here
confirms this, because Mammuthus and Loxodonta have both larger body masses
(Christiansen, 2004) and larger canal radii than Elephas.
The two coils of the cochlea of the elephantoid from Friesenhahn Cave is similar
to what is observed in Elephas (Hyrtl, 1845). However, the cochlea of an unidentified
extant elephant completed over three coils, as reconstructed by Blair (1717-1719). The
number of coils completed by the cochlea can vary within a species (see Chapter 3), but
there are no reports of variations greater that 360° (see also Chapters 3-4, for further
discussion of inner ear variation). Although the morphology of the petrosals potentially
35
represents taxonomic variation, these differences more likely are artefactual. Hyrtl
examined casts of the inner ear cavities, whereas Blair filed the bone to expose the
structures. Casting of the bony labyrinth (e.g., Hyrtl, 1945) would provide a more precise
reconstruction of the cochlea, but the possibility that the cochlea of extant elephants
varies to such an extent has yet to be confirmed.
As with the elephantoid from Friesenhahn Cave, the cochlea of extant Elephas
maximus (personal observation of CT data of TMM M-6445, Elephas maximus at
http://digimorph.org) and the Eocene fossil Numidotherium koholense (Court, 1992b) are
planispiral. Both Numidotherium and TMM 933-950 share similar ratios between the
height of the cochlea and maximum width of the basal turn (0.52 and 0.48). In contrast,
the same ratio is much higher in the Eocene proboscidean Moeritherium (0.79; Court,
1992b). This index of the shape of the cochlea might reflect functional aspects of the ear
region. Gosselin-Ildari (2006) found a positive correlation between the index and both
low frequency limits and best frequency achieved by the cochlea. If Gosselin-Ildari’s
correlations are correct, then Moeritherium would have had both a higher low frequency
limit, as well as a higher best frequency, than either Numidotherium or the elephantoid
from Friesenhahn Cave.
There is little to no development of a secondary lamina in the Friesenhahn Cave
elephantoid, unlike the well-developed lamina that extends throughout the basal turn in
Numidotherium (Court, 1992b, fig. 2). The secondary lamina supports the basilar
membrane, which in turn supports the spiral organ of Corti for hearing. The width of the
basilar membrane can be determined by measuring the gap between the secondary lamina
on the outer wall of the bony labyrinth, and the primary lamina that juts into the cochlear
cavity from the modiolus (central axis of the cochlea). The size of the gap between
laminae is significant, because the width of the basilar membrane is correlated to auditory
36
TABLE 2.2. Semicircular canal radii of curvature (in mm) for extant elephants and
elephantoid from Friesenhahn Cave.
Radius
Elephas
Loxodonta
Elephantoid
Radius Average
Anterior
4.0
5.2
5.0
4.7
Lateral
2.8
3.6
2.7
3.0
Posterior
4.1
5.5
5.5
5.0
Species
Average
3.6
4.8
4.4
4.3
37
function (Fleischer, 1976). For example, the basilar membrane of microchiropteran bats
that echolocate is narrower through the basal turn of the cochlea (Ramprashad et al.,
1979) than in mammals sensitive to low frequencies, such as elephants (Payne et al.,
1986; Poole et al., 1988). The absence of the secondary lamina and presence of the
secondarily undivided perilymphatic foramen supports the sensitivity to low frequency
vibrations by the proboscideans from Friesenhahn Cave.
Variation within the ear region
Although the external surfaces of the petrosals are similar across the Friesenhahn
Cave sample, there are minor anatomical variations. This variation includes connection of
the crista interfenestralis to the tympanohyal to form a foramen for the stapedius muscle,
completion of the ventral floor of the aquaeductus Fallopii, and the shape of the fenestra
vestibuli (stapedial ratio). The variations may be the result of one or several factors. The
taphonomic history of the specimens both before and after burial is one possibility.
Phylogenetic variation also is possible, because remains of two proboscidean species
were recovered from the cave. Lastly, ontogenetic variations given that a growth series of
individuals was identified based on teeth, and individual variation (in the absence of
phylogeny and ontogeny) are other potential sources. Each type of variation is discussed
below in the context of the morphological features that vary within the Friesenhahn Cave
petrosal sample.
Taphonomy is an important concept in paleontology, and data that are lost as a
result of post-mortem degradation of fossil specimens can be both extreme and frustrating
(e.g., Lawrence, 1968). All of the petrosals in the sample from Friesenhahn Cave show
taphonomic damage to some extent. Because the petrosal is so solidly fused to the rest of
38
the cranium in mammoths, the skulls of the young elephants found in the cave would
have to be greatly fragmented in order to isolate the petrosals. In fact, no complete skulls
of Mammuthus or Mammut were recovered from the cave.
The lack of complete skulls might be a result of predation following the
hypothesis that Friesenhahn was a carnivoran den (Graham, 1976). Although there is no
direct evidence of predation on the petrosal bones themselves, tooth marks were observed
on many of the long bones in the cave, and the cranial elements may have fragmented as
individuals of Homotherium gnawed on skulls. A second more likely hypothesis for the
fragmentation of skulls is damage via compaction of sediments after the elephant remains
were buried. The skulls of mammoths, mastodons, and extant elephants are quite delicate,
with a high degree of pneumaticity. But as stated earlier, the petrosals are dense and lend
themselves well to preservation. It is credible that only petrosals and jaw fragments with
teeth would be preserved, as this is a phenomenon that is observed in many mammal
faunas (Archibald, 1979).
Variation among the stapedial ratios cannot be explained by taphonomic
destruction since the measurement was calculated only for those specimens with
complete fenestrae vestibuli. Taphonomy does account for the majority of the variation
observed for the foramen for the stapedius muscle, with noticeable damage is observed in
most cases where the crista interfenestralis does not meet the tympanohyal on the
posterior aspect of the petrosal. However, the region is undamaged in TMM 933-166 and
933-1032, where the foramen is not closed. Thus, taphonomy cannot account for the
observed variation.
The ventral floor of the aquaeductus Fallopii is broken in most of the specimens,
but a few specimens, including TMM 933-548, appear to be undamaged in the vicinity of
the canal. The longitudinal groove that spans the length of the medial portion of the floor
39
of the canal likely is a facet for a missing lateral edge that would have completed the
canal. The floor of the aquaeductus is thin, but finished edges on the sheets are easily
discernable from broken edges.
Phylogenetic differences between the gross anatomy of molars of Mammuthus
and Mammut are well documented, and remains of both Mammuthus and Mammut were
found in Friesenhahn Cave. Although the vast majority of identifiable specimens (teeth)
were referred to Mammuthus by Graham (1976), one or more petrosals might represent
Mammut, and the variation observed might be taxonomic. In fact, the partial condition of
the aquaeductus Fallopii is the only feature that varies in the Friesenhahn Cave sample
that also is observed in the sample of Mammut petrosals from Boney Spring. However,
the condition is only observed in one specimen. No other specimen from Boney Spring
preserves this region of the petrosal intact.
Nonetheless, the partial condition may be a distinguishing feature of the petrosal
of Mammut. The partial condition is observed within 57% (12 out of 21 specimens
preserving the region intact) of the elephantoid petrosals from Friesenhahn Cave. If the
partial condition is unique to Mammut (at the very least when comparing Mammut and
Mammuthus), then at least 12 out of the 65 petrosals (18%), but as many as 37 petrosals
(57%) recovered from Friesenhahn Cave represent Mammut. Such a figure does not agree
with the relative abundance based on numbers of molars, where the vast majority of
molars (500 out of 513, or 97%) are identifiable as Mammuthus. A more likely scenario
is that the variation between partial and complete aquaeducti will be uncovered in a larger
sampling of Mammut petrosals, and that mammoth and mastodon fossils cannot be
distinguished based on petrosal characters alone. However, in order to fully test this
hypothesis, a large sample of definitively Mammuthus and definitively Mammut (greater
than the sample from Boney Spring that was examined here) is needed. Until such a study
40
is accomplished, the least inclusive taxon that can be identified for any Friesenhahn Cave
petrosal sample is Elephantoidea, yet this may change as additional proboscidean taxa are
included in anatomical comparisons.
The completion of the stapedial muscle foramen could not be determined for
Mammut from Boney Spring, so whether the variation identified in the Friesenhahn Cave
sample is taxonomic in nature is inconclusive. The average stapedial ratio is slightly
larger for Mammut than the Friesenhahn Cave sample (1.9 versus 1.8), but the range falls
within that calculated for the Friesenhahn Cave elephantoids. This would be expected if
there are multiple mastodons in the Friesenhahn fossil sample, but there is no clear
indication if there is a difference in stapedial ratios between Mammut and Mammuthus.
Until stapedial ratios are calculated for a large sampling of definitive mastodons and
mammoths, the taxonomic nature of the stapedial ration among elephantoids remains
inconclusive.
Ontogenetic variation may explain some variation observed in the sample,
including that for the aquaeductus Fallopii. A growth series of mammoths is recognized
in Friesenhahn Cave based on dental and postcranial material (Graham, 1976), and it is
reasonable to expect the petrosals to have come from individuals of different maturities.
Furthermore, all the petrosals likely are from juveniles because no adult mammoth teeth
were recovered from the cave.
In order to ascertain whether or not the observed variation is related to the
maturity of an individual, there must be some way to determine the relative maturities of
the petrosals. Evidence suggests that dimensions of the inner ear of mammals do not
change significantly once the bony walls are ossified (Hoyte, 1961; Ekdale, 2005), but
such is not the case with the external surface of the petrosal, which expands externally via
accretionary growth of bone (Chapters 3-4; Hoyte, 1961). However, there is no
41
correlation between overall size of the petrosals in the sample and any of the variation
observed. Overall size of petrosals is affected by breakage, especially in the sample from
Friesenhahn Cave, so using size would not be a reliable indicator of maturity.
The best candidate for ontogenetic variation is the aquaeductus Fallopii, provided
that the variation is not the result of taxonomy (which is unlikely; see above). The
embryological development of the aquaeductus is undocumented for proboscideans, but a
published account of the skull of a fetal African elephant (Eales, 1926) sheds light onto
the ossification of the elephant otic region. A similar groove extending the length of the
ventral floor of the aquaeductus Fallopii was observed in the fetal skull (Eales, 1926, pl.
9; Fig. 2.6). The aquaeductus was not fully ossified in the fetus, and the groove formed a
seam between the bony petrosal medially and a cartilaginous wing that extended to the
squamosal laterally. Further evidence from the ontogeny of the ear region of the gray
short-tailed opossum, Monodelphis domestica, where the lateral aspect of the petrosal is
the last part of the ear region to ossify (Clark and Smith, 1993), agrees with Eales’
observation in Loxodonta. If the lateral portion of the ventral floor of the canal of a
juvenile mammoth was not ossified completely upon death, an incomplete (Figure 2.3A)
or partially complete aquaeductus (Figure 2.3B) would be preserved in the fossil record.
If one was to assume that completion of the ventral floor of the aqueductus
Fallopii is the result of variations in maturity, then the petrosals from Friesenhahn Cave
can be divided into two maturity stages based on aquaeductus development – an
immature canal versus a mature canal. The incomplete state illustrated in Figure 2.3A is
the result of damage to the bone. However, the specimens that possess a partially
complete aquaeductus Fallopii (Figure 2.3B) represent an immature stage. In all of these
partially complete specimens, the aquaeductus is enclosed anteriorly, but not posteriorly.
Assuming the specimens represent immature stages, then ossification of the ventral
42
FIGURE 2.6. Diagram of petrosal of fetal elephant (redrawn from Eales, 1926).
Abbreviations: af – aquaeductus Fallopii; ci – crista interfenestralis; ff – foramen faciale;
fv – fenestra vestibuli; hf – hiatus Fallopii; lw – lateral wing (shaded), cartilaginous in
fetal elephant; pf – perilymphatic foramen; pr – promontorium.
pr
hf
lw
ff
af
fv
ci
pf
43
aspect of the floor of the aquaeductus Fallopii begins at the anterior edge of the canal and
continues posteriorly. As stated above, 57% of the petrosals within the sample have the
partial, or immature state (see Table 2.1).
Variation in the stapedius muscle foramen might have an ontogenetic basis as
well. One hypothesis is that the petrosals in which the foramen is not complete (Figure
2.4A) are less mature than individuals with a complete foramen (Figure 2.4B). There is
no correlation between a complete connection of the crista interfenestralis and
tympanohyal and a complete aquaeductus Fallopii because the stapedius muscle foramen
is complete regardless of completion of the aquaeductus Fallopii. This would suggest that
the crista interfenestralis and tympanohyal ossify together earlier in ontogeny than the
medial and lateral portions of the aquaeductus Fallopii. Interestingly, the stapedius
muscle foramen is incomplete in TMM 933-166 and 933-1032, however the aquaeductus
Fallopii is partially complete in TMM 933-166, while the floor of the canal is complete in
TMM 933-1032, causing ambiguity. The closure of the foramen for the stapedius muscle
could be ontogenetic in nature, but it would have to ossify out of sequence with the
aquaeductus Fallopii. Ontogenetic sequence polymorphism may be a fairly common
phenomenon (see Colbert and Rowe, 2008).
The broad range in stapedial ratios is not correlated with size of the petrosals, nor
with the ontogenetic sequence based on the maturity of the aquaeductus Fallopii. That is,
the ratios of immature petrosals are neither more nor less elliptical than the ratios of the
bones with mature canals. In fact, both the minimum (1.4) and maximum (2.1) ratios
were calculated for specimens with complete canals.
The variation in the stapedius muscle foramen and the stapedial ratio cannot be
explained by taphonomy or ontogeny, and variation in both structures is inconclusive for
taxonomy. The two specimens in which the crista interfenestralis and tympanohyal do not
44
meet to form the foramen for the stapedius muscle may instead reflect individual
variation. Likewise, the variation observed in the stapedial ratio appears to be
intraspecific variation that is not explained by maturity.
Marsupials tend to have more circular stapedial footplates with ratios below 1.8
when compared to placentals, which tend to have ratios at or above 1.8 (Segall, 1970).
With a few exceptions, the pattern of the ratio has held for extinct and extant therians
(marsupials plus placentals). Because of this, the ratio is used to distinguish isolated
marsupial and placental petrosals (see Archibald, 1979; Wible et al., 2001; Ekdale et al.,
2004), and the ratio is used in phylogenetic analyses (Wible, 1990; Rougier et al., 1998;
Archibald et al., 2001; Ladevèze, 2007). However, those studies did not take variation of
the ratio into account. The rounded average stapedial ratio for the sample from
Friesenhahn Cave is 1.8, which agrees with the values for placental mammals that Segall
(1970) calculated. However, when the range observed within the entire Friesenhahn
sample is considered, the ratios span both sides of Segall’s cut-off (1.4-2.1; Table 1). In
fact, the proportion of specimens with ratios below 1.8 versus above 1.8 is 30:27, nearly
1:1 (note that the ratio was not calculated for every specimen, only for those that have the
complete fenestra).
The observed range of variation casts doubt on the validity of using the stapedial
ratio to distinguish between placental and marsupial taxa. If identification of the
elephantoid petrosals from Friesenhahn Cave were based solely on the stapedial ratio,
then half of the specimens would be identified as marsupials, and half as placentals.
Broad ranges of ratios have been reported for other taxa, including Monodelphis
domestica (Chapter 3) and the Cretaceous eutherian Kulbeckia kulbecke (Ekdale et al.,
2004). In fact, a much greater range is observed in Kulbeckia than in the elephantoid
petrosal sample from Friesenhahn Cave (which likely contains two taxa). How the range
45
in stapedial ratios of the Friesenhahn petrosals compares with ratios in modern elephant
populations is unknown.
The ratios reported here (as well as in Ekdale et al., 2004) were calculated using
the fenestra vestibuli as a proxy. Ideally the stapes would fit snugly into the fenestra
vestibuli in life. Stapes are unknown for K. kulbecke, but stapedial footplates were
recovered for Mammuthus (as described above). Perhaps the dimensions of the fenestra
vestibuli do not serve as an accurate proxy for dimensions of the stapedial footplate. The
footplate of the stapes is held into the fenestra vestibuli by an annular ligament, so the
footplate must be smaller than the fenestra. Indeed, the ratios differ when using the
fenestra vestibuli versus the actual footplate in TMM 933-950 (1.8 versus 1.7,
respectively), but the difference in these values is within one standard deviation of the
mean (0.1). Furthermore, the stapes of TMM 933-951 does, in fact, fit tightly in the
fenestra vestibuli. Blair (1717-1719) observed the same in the extant elephant.
Nonetheless, no thorough investigations comparing ratios between the footplate of the
stapes and the respective fenestra vestibuli have been published for any mammal taxon,
and such a study is beyond the scope of this paper.
CONCLUSIONS
The sample of elephantoid proboscidean petrosals collected form Friesenhahn
Cave is significant for several reasons. First, there is a poor representation of isolated
proboscidean petrosals in the fossil record. That such a large sample of proboscidean
petrosals exists from Friesenhahn Cave is a testament to the importance of that locality.
Furthermore, the petrosal is solidly fused to the skull in extant proboscideans, as well as
in Mammuthus and Mammut. Together with the presence of a bony bulla covering the
46
tympanic cavity in mature petrosals, as well as the massive size of elephant skulls, the
otic region of proboscideans is a difficult region to study. The Friesenhahn sample
provides an opportunity to study an important region of fossil proboscidean skulls that
otherwise is unavailable.
The further significance of the petrosal sample is that a unique growth series is
identified for the first time from a collection of fossil petrosals using data from the
aquaeductus Fallopii. If variations observed in other elements besides the petrosal
coincide with one of the two maturity stages identified from Friesenhahn Cave, then that
information can be used to construct an ontogenetic sequence between elements within an
individual, and expand our knowledge of the biology of fossil mammals. The possibility
that the variation is the result of multiple taxa in the sample exists, although this
phenomenon is not likely given the discrepancy between the frequencies of variable
states and taxonomic frequencies within molar samples.
Lastly, the sample from Friesenhahn Cave exhibits a broad range of stapedial
ratios, a measurement that often is used in phylogenetic analyses. Given that this
measurement is variable, not only in the elephantoids from Friesenhahn Cave and Boney
Spring, but also select Mesozoic taxa, it seems inappropriate for the stapedial ratio to be
used to distinguish metatherian and eutherian petrosals. In fact, the general systematic
utility of the ratio among all mammals is questionable. Likewise, the measurement should
not be used for phylogenetic analyses unless treated as a continuous character to account
for variation and large samples are examined and measured before character scoring in an
analysis.
47
CHAPTER 3: POSTNATAL ONTOGENETIC VARIATION IN THE
BONY LABYRINTH OF MONODELPHIS DOMESTICA
(MAMMALIA: MARSUPIALIA)
ABSTRACT
Ontogeny, or the development of an individual from conception to death, is a
major source of variation in vertebrate morphology. All anatomical systems are affected
by ontogeny, and knowledge of the ontogenetic history of these systems is important to
understand when formulating biological interpretations of evolutionary history and
physiology. The present study is focused on how variation affects the bony labyrinth
across a growth series of an extant mammal. Digital endocasts of the bony labyrinth were
constructed using CT data across an ontogenetic sequence of Monodelphis domestica, an
important experimental animal. Various aspects of the labyrinth were measured,
including angles between the semicircular canals, number of turns of the cochlea,
volumes of inner ear constituents, as well as linear dimensions of semicircular canals.
There is a strong correlation between skull length and age, but from 27 days after birth
onward, there is no correlation with age among most of the inner ear measurements.
Exceptions are the height of the arc of the lateral semicircular canal, the angular deviation
of the lateral canal from planarity, the length of the slender posterior semicircular canal,
and the length of the canaliculus cochleae. Adult dimensions of several of the inner ear
structures, such as the arcs of the semicircular canals, are achieved before the inner ear is
functional, and the non-ontogenetic variation in the bony labyrinth serves as an important
source for behavioral, physiological, and possibly phylogenetic information.
48
INTRODUCTION
Variation is a phenomenon that affects all morphological systems. Phenotypic
variation is the result of many factors, including phylogenetic history, gender, geography,
and ontogeny. Ontogeny, or the development of an individual from conception to death,
is a major source of variation that has garnered much attention in scientific literature.
Ontogenetic variation is a special problem for systematists, particularly paleontologists,
when only one specimen or a small sample of specimens of a taxon is known, which can
lead to vexing systematic problems (Brinkman, 1988; Tykoski, 2005; Wiens et al., 2005).
For example, two specimens that differ slightly in morphology might represent two
separate taxa, or they might represent different ontogenetic stages of a single species, or
the variation might indicate any of a series of other possibilities (see Bever, 2006).
Because of this, the effect of ontogenetic transformations on anatomy should be taken
into consideration when scoring states of morphological characters for phylogenetic
analyses.
All morphological systems are affected by ontogenetic variation at some level.
The fully formed bony labyrinth of the inner ear, as preserved within the internal cavities
of the petrosal bone of mammals, often is considered immune to ontogenetic change,
because the petrosal ossifies and matures early in development in humans (Jeffery and
Spoor, 2004). This is not to say that the mammalian bony labyrinth does not exhibit
variation throughout ontogeny. The auditory system undergoes profound and pronounced
ontogenetic transformations, but it quickly achieves mature morphologies that appear to
remain stable through the remainder of the life of an individual, if only for select
placentals including humans (Jeffery and Spoor, 2004) and rabbits (Hoyte, 1961). In
contrast, the majority of other systems, such as the long bones of the vertebrate limb,
experience a more prolonged ontogeny relative to the inner ear and teeth. Given the rapid
49
maturation of the mammalian ear region, as well as the prevalence of ear regions in the
fossil record (the petrosal is among the densest elements in the body), the otic region is
special and important to vertebrate paleontologists. In fact, petrosals are common
elements preserved in Mesozoic mammal faunas (Archibald, 1979).
If the fully ossified bony labyrinth does not change throughout postnatal
ontogeny, then the maturity of a specimen would not need to be known in order to make
biological interpretations of behavior or phylogeny, using the ossified inner ear. The
results of recent studies included information from the bony labyrinth that has been used
by researchers to investigate bipedalism in fossil hominids (Spoor et al., 1994, 1996),
evolutionary relationships among early humans (Hublin et al., 1996), aquatic adaptations
in cetaceans (Spoor et al., 2002), and behaviors in extinct dinosaurs (Rogers, 1998, 1999)
where the relative maturity of individuals was not ascertained.
The middle ear region of mammals, as preserved on the external surface of the
petrosal, is a widely used source of phylogenetic information. This region is useful for the
inference of evolutionary relationships within and among major groups of mammals,
including Carnivora (Bugge, 1978; Hunt, 1987, 1989), Cetacea (Geisler and Luo, 1996;
Luo and Gingerich, 1999), Chiroptera (Wible and Davis, 2000; Wible and Novacek,
1988), Primates (Harvati and Weaver, 2006; MacPhee, 1981; Wible and Martin, 1993),
and even some of the earliest mammals (Archibald, 1979; Ekdale et al., 2004; Ladevèze,
2004; MacIntyre, 1972; Meng and Fox, 1995; Wible, 1990; Wible et al., 2001). However,
mammalian systematists often are unable to evaluate the complex structures of the inner
ear, such as the cochlea, vestibule, and semicircular canals, because these structures are
surrounded completely by bone and are difficult to observe. Thus, the degree of
ontogenetic variation within the inner ear is unknown for most mammals.
50
Natural endocasts of the inner ear chambers are known (Court, 1992b; Kielan-
Jaworowska, 1984; Meng and Wyss, 1995), but preservation of such fossils in the rock
record is rare. In the absence of such exceptional preservation, removal of the
surrounding bone therefore is necessary in order to observe the bony labyrinth.
Historically, access to the inner ear cavities was granted through serial sectioning or
dissolution of bone (Gray, 1907, 1908; Luo and Marsh, 1996; Novacek, 1986; West,
1985). More recently, the advent of high-resolution X-ray computed tomography (CT)
has allowed observation of the internal cranial structures and description of the cranial
osteology of vertebrate taxa (Bever et al., 2005; Maisano et al., 2002; Rowe et al., 1995,
1997; Spoor and Zonneveld, 1995; Tykoski et al., 2002; Hullar and Williams, 2006).
Computed tomography was used here in order to investigate changes following
the onset of ossification of the bony labyrinth across an ontogenetic series of
Monodelphis domestica, which is a species of marsupial mammal that is important for
experimental biomedical research (Fadem et al., 1982; Macrini, 2004). Monodelphis
commonly is used in clinical studies (e.g., Eugenín and Nicholls, 1997; Kusewitt et al.,
1999; Wang et al., 2003; Halpern et al., 2005), and knowledge of the ear benefits future
studies of the auditory system of mammals.
The development of the inner ear of the closely related species Didelphis
virginiana is well documented (e.g., Larsell et al., 1935; McCrady, 1938), and both D.
virginiana and M. domestica share similar chronologies in skeletal maturation (although
both species exhibit different rates of growth; Rowe, 1996, 1997). Because of this,
information about the maturation of the inner ear of D. virginiana likely is similar in
relative timing to that of M. domestica. At birth, the inner ear of D. virginiana is clearly
recognizable (Larsell et al., 1935: fig. 3, stage 35), but it is still immature in form. The
anterior, lateral, and posterior semicircular canals are present as complete tubes, but they
51
are not as large as those of adults. The cochlea completes one half of a turn at birth, and
will eventually complete two and a quarter turns (Larsell et al., 1935).
Ossification of the bony labyrinth has not begun at birth, and the petrosal bone
itself is one of the last endochondral bones to begin ossification in M. domestica, around
postnatal day 12 (Clark and Smith, 1993). Ossification of the petrosal of M. domestica
begins in three ossification centers around the developing cochlea, and the bone
surrounding the cochlea is fully ossified by day 25. Expansion of ossification of the
cochlear cartilage and ossification of the postparietal into the region of the semicircular
canals triggers the onset of ossification of the canalicular cartilage (Clark and Smith,
1993). Ossification of the region of the semicircular canals is underway by day 20, and
the bony walls of the semicircular canals are the first to ossify, followed by the spaces
between them.
The petrosal of M. domestica is not fully ossified by day 30, but the only
unossified portions are part of the styloid process and the lateral wall of the subarcuate
fossa, which houses a petrosal lobule of the paraflocculus of the cerebellum. The styloid
process does not contribute to the morphology of the bony labyrinth within the petrosal
bone, but the subarcuate fossa is defined by the semicircular canals. The development of
the bony semicircular canals versus the subarcuate fossa is not entirely independent in
primates (Jeffery and Spoor, 2006), but the influence that subarcuate fossa ossification
has on the dimensions and orientation of the fully ossified semicircular canals is
unknown.
Beyond anatomy, the development of auditory and vestibular physiological
responses in M. domestica is well documented (Aitkin et al., 1997; Reimer, 1996). The
inner ear of M. domestica is not functional at birth, because the middle ear cavity is filled
with fluid until 26 days after birth, and the external auditory meatus does not open until
52
28-30 days postnatal (Aitkin et al., 1997). Pouch young of D. virginiana respond to
vestibular stimuli before acoustic reflexes are observed (Larsell et al., 1935), and the
same may be the case for M. domestica (given similarities in skeletal development
between the two species), although such a relationship between the physiological
responses has yet to be determined for M. domestica. Evoked auditory responses can be
detected in M. domestica at 28 days after birth, and adult hearing thresholds are achieved
by day 39 (Reimer, 1996).
The ontogeny of hearing in M. domestica follows a similar pattern observed in
placental mammals of similar size, allowing M. domestica to be used as a generalized
model for physiological investigations of the inner ear of therian mammals (Reimer,
1996). The morphology of the membranous labyrinth of D. virginiana (Larsell et al.,
1935) develops in a similar fashion to that of Mus musculus (Morsli et al., 1998), further
suggesting that the ontogenetic pattern of the inner ear of didelphids (to which both
Monodelphis and Didelphis belong) also may afford insights into the developmental
morphology of placentals.
MATERIALS AND METHODS
Eleven whole dried skulls of Monodelphis domestica representing different ages
(absolute numbers of postnatal days) were scanned at The University of Texas High-
Resolution X-ray CT facility (UTCT). All specimens were born and raised in captivity
under controlled conditions at the Southwest Foundation for Biomedical Research in San
Antonio, TX, where exact numbers of postnatal days were recorded. Once CT scanned,
the specimens were accessioned into the Vertebrate Paleontology Laboratory (Austin,
TX) recent mammal collection (TMM M). The specimens used for this study range in age
53
from 27 to 465 days after birth. Day 27 was chosen as the earliest age, because the bony
labyrinth itself likely is to be fully ossified by that time.
The specimens used are TMM M 7595 (day 27), 8261 (day 27), 8265 (day 27),
7536 (day 48), 8266 (day 56), 7539 (day 57), 7542 (day 75), 8267 (day 76), 7545 (day
90), 8268 (day 90), and 8273 (day 465). One additional individual (TMM M 7599), a
fully mature retired breeder from the colony, was used in this study, although the exact
age of the individual is unknown. That specimen (TMM M 7599) was not used to
ascertain age-related changes, but rather it was used for morphological comparisons with
the rest of the specimens in the sample. Scan data for many of these specimens are
publicly available at the website entitled ‘Digital Morphology: a National Science
Foundation Library at the University of Texas at Austin’
(http://digimorph.org/specimens/Monodelphis_domestica/whole/). Table 3.1 includes
scanning parameters used for each specimen.
The inner ear cavities were digitally segmented into constituent parts (cochlea;
vestibule; anterior, lateral, and posterior semicircular canals; anterior, lateral and
posterior ampullae; common crus) by isolating voxels on individual CT slices that define
the anatomical structure of interest. These segmented data can be rendered as a 3-
dimensional digital endocast of the segmented structure, which provides easy observation
of inner ear structures. Segmentation, visualization, and measurement of digital endocasts
was performed in the computer programs VG Studio Max 1.2
©
(Volume Graphics) and
Amira 3.1
©
. Anatomical terminology follows Evans (1993) and Spoor and Zonneveld
(1995).
Numerous angular and linear measurements, as well as indices describing various
aspects of the digital endocasts, were made in Amira software, and volumetric
measurements were made in both VG Studio Max and Amira. Measurement methods and
54
TABLE 3.1. CT scanning parameters for specimens of Monodelphis domestica. The
specific parameters employed during the scanning and post-scanning image processing
for each specimen is provided here. Definitions of parameters are as follows: Spec. # –
specimen number, catalogued as TMM M; # Slices – number of CT slices through the
inner ear collected in the coronal (original) slice plane; Interslice – interslice spacing, the
distance between consecutive slices (mm); Field Rec. – field of reconstruction,
dimensions of individual CT slice (mm); Interpixel – interpixel spacing, vertical and
horizontal dimensions of individual pixel measured by dividing field of reconstruction by
resolution (mm); Resolution – number of pixels in CT image, either 512 X 512 pixels, or
1024 X 1024 pixels.
Spec. #
# Slices
Interslice
Field Rec.
Interpixel
Resolution
7595
116
0.0371
10.8
0.0211
512
8261
69
0.0483
16.5
0.0161
1024
8265
108
0.0338
13.0
0.0127
1024
7536
81
0.0625
14.0
0.0273
512
8266
119
0.0355
17.0
0.0166
1024
7539
76
0.0676
14.9
0.0291
512
7542
96
0.0608
17.9
0.035
512
8267
91
0.048
22.0
0.0215
1024
7545
81
0.0682
19.0
0.0371
512
8268
109
0.048
22.0
0.0215
1024
8273
61
0.117
55.65
0.0543
1024
7599
151
0.09
23.0
0.0449
512
55
indices calculated largely follow those used by Spoor and Zonneveld (1995, 1998), and
Jeffery and Spoor (2004), many of which were used previously to formulate
interpretations of mammalian systematics and the physiology of the inner ear. A detailed
explanation of each measurement is provided in the following section.
Definitions and exploration of measurements
Angular Measurements
Most of the angles considered in this study are thought to be important to inner
ear physiology, such as planar relationships between semicircular canals (Calabrese and
Hullar, 2006), as well as the evolutionary relationships of mammals, such as coiling of
the cochlea (Ekdale et al., 2004; Graybeal et al., 1989; Rougier et al., 1998). The degree
of coiling completed by the cochlea is calculated to investigate if the number of cochlear
turns increases as the bony labyrinth matures after the onset of ossification. The coiling
of the cochlea often is used to separate therian mammals from non-therian mammals,
including monotremes and non-mammalian synapsids (Graybeal et al., 1989), and
cochlear coiling, in conjunction with the length of the cochlear canal, is related to audible
frequencies (West, 1985). To measure the number of turns of the coil, the cochlea is
viewed from above, perpendicular to the axis of rotation. A line is drawn from the
junction of the primary and secondary bony laminae at the base of the cochlea (following
Geisler and Luo, 1996) to the center of the axis of rotation. Each time that the cochlea
crosses this line is counted as one half turn. An additional value is added to this – an
angle measured from the line drawn through the axis of rotation to the apical tip of the
cochlear canal – in order to determine the total number of degrees completed by the
cochlea.
56
The deviations of semicircular canals away from the orthogonal planes of the
skull are calculated and reported in several physiological studies (e.g., Calabrese and
Hullar, 2006; Hullar and Williams, 2006). The orientation of canals might signify that a
vertebrate is more sensitive to certain movements of the head, such as pitch or roll
(Hullar and Williams, 2006). Orthogonal planes in the published studies are defined by
landmarks across a complete skull, so the orthogonality of the planes of canals cannot be
determined for isolated petrosals (which often is the case with fossils). Nonetheless, the
angles between the planes are measured here. If the angles between the canals changes
across ontogeny, then the orientation of the canals with respect to the orthogonal planes
of the skull changes also. Orientations of the canals are measured as the angle between
the planes of two canals, when both planes are perpendicular to the field of view (Figure
3.1a-c). The plane of a canal is fit to points at the center of the lumen at the midpoint of
the arc, as well as the aperture of the canal into the ampulla at one end, and vestibule at
the other.
The plane of the lateral semicircular canal is used to measure the angle between
the canal and the cochlea. It often is assumed that mammals hold their heads so that the
lateral semicircular canal is parallel to Earth horizontal (de Beer, 1947), but such is not
always the case (see Hullar, 2006, for a review). The angle is measured in the same
manner as the orientations of the three canals, where the plane of the cochlea intersects
three points equally spaced through the basal turn of the cochlear canal (Figure 3.1d).
Measurement of angular deviation of a canal from its plane was made following modified
protocols outlined by Calabrese and Hullar (2006) and Hullar and Williams (2006). The
dimension is calculated by determining the maximum linear deviation achieved by the
center of the lumen of the canal when the plane of the canal is oriented parallel to the
horizon (Figure 3.1e). Partial angles of deviation are determined trigonometrically using
57
FIGURE 3.1. Graphical depiction of angular measurements between semicircular canals
and cochlea. A, planes of posterior and lateral canals; B, planes of anterior and lateral
canals; C, planes of anterior and posterior canals; D, planes of basal turn of cochlea and
lateral canal; E, semicircular canal in profile and viewed with plane parallel to horizon.
Arrows indicate angle measured. Abbreviations: ac – anterior semicircular canal; am –
ampulla; ar – arc radius of curvature; co – cochlea; cp – plane of semicircular canal; lc –
lateral semicircular canal; ld – linear deviation of canal from plane; pc posterior
semicircular canal; sc – semicircular canal; vb – vestibule.
ac
pc
lc
lc
pc
A B
C
E
D
ac
lc
co
ar
sc
vb
am
cp
sc
lc
ld1
ld2
58
the linear deviations and the arc radius of curvature of the semicircular canal (described
below with other linear measurements). The total angular deviation, reported here, is the
sum of the two partial deviations.
Linear Measurements
The measurements made on digital endocasts are reported in millimeters, and they
include total length of the labyrinth, radius of the arcs of the semicircular canals,
diameters of the canal lumen, and lengths of the semicircular and cochlear canals, the
canaliculus cochleae (for the perilymphatic duct), and the aquaeductus vestibuli (for the
endolymphatic duct).
The total length of the bony labyrinth is a linear measure of size of the entire
series of inner ear cavities (following Jeffery and Spoor, 2004). The length is measured as
the greatest distance from the posterior-most point at the center of the lumen of the
posterior semicircular canal to the center of the lumen of the outer bend of the basal turn
of the cochlea.
Linear measurements of the size of the cochlear spiral follow the method
proposed by Gosselin-Ildari (2006). The width of the cochlea is measured as the greatest
distance from the ventral edge of the fenestra cochleae to the outer wall of the outer curve
of the basal turn of the cochlea (Figure 3.2). The height of the cochlea is measured from
the top of the spiral to the level of the dorsal edge of the fenestra cochleae, perpendicular
to the width and parallel to the plane of the basal turn of the cochlea.
The radius of curvature of the arc of the semicircular canals (“R” of Jones and
Spells 1963) is correlated to both sensitivity of the canal, as well as agility of the animal.
For animals of similar body size, larger canals are more sensitive (Hullar and Williams,
2006) and are indicative of a more maneuverable and agile animal (Spoor et al., 2007).
59
FIGURE 3.2. Cross-section through midline of the cochlea displaying the measurements
for height and width of the spiral of the cochlea. Abbreviations: ch – height of spiral of
cochlea; co – cochlear canal; cw – width of spiral of cochlea; fc – fenestra cochleae; cn –
canal for cochlear branch of cranial nerve VIII within the modiolus; rp – reference plane
for measuring height of spiral of cochlea.
ch
rp
cw
fc
cn
co
co
co
co
60
The arc radius of a canal is half the average of the height and width of the arc.
The height of the anterior semicircular canal is measured as the greatest distance from the
wall of the vestibule to the center of the lumen of the canal, perpendicular to the plane of
the lateral canal. The height of the posterior semicircular canal is measured parallel to the
plane of the lateral canal from the center of the lumen of the common crus to the center of
the lumen of the posterior limb of the canal. The height of the lateral canal is measured as
the greatest distance from the wall of the vestibule to the center of the lumen of the canal.
The widths for all of the canals are perpendicular to the respective heights, and measured
from the center of the lumen of opposing limbs.
Dynamic responses of the semicircular canals also are dependent on the radius of
the membranous semicircular duct within the bony canal (Hullar, 2006). There is not a
1:1 size correlation between the bony and membranous labyrinth (Curthoys et al., 1977a,
b; Igarashi, 1967), so the radius of the membranous duct cannot be determined in the
absence of soft tissues (such is the case with fossils). The diameter of the bony canals is a
common measurement taken, and diameters for M. domestica are reported here, but they
should not be used as proxies for the diameters of the membranous ducts.
The lengths of the slender semicircular canals, defined as the unampullated
portion of the canals, relates to sensitivity of rotations by the head (Boyer and Georgi,
2007). The canal lengths were measured using the SplineProbe tool in the Amira
software, wherein anchor points were placed at varying intervals at the center of the
lumen of the canal. The length of the cochlear canal was measured in the same manner,
with the starting point at the junction of the primary and secondary laminae.
61
Volumetric Measurements
Planar boundaries were maintained as much as possible between cavities and the
separation between the cochlea and vestibule was made at the medial border of the
fenestra vestibuli. The entirety of the fenestra itself was included within the vestibule.
The canaliculus cochleae and aquaeductus vestibuli were included in the volumes of the
cochlea and vestibule, respectively.
Indices and Ratios
Several indices and ratios were calculated from the linear measurements described
above. Partial volumes of specific segments were calculated by dividing the volume of
the segment (e.g., cochlea) by the total volume of the bony labyrinth. The majority of the
remaining indices describe the aspect ratios of the arcs of the semicircular canals (which
might signify agility in locomotion; see Hullar, 2006) and the cochlea, which is related to
auditory capabilities (Gosselin-Ildari, 2006). The aspect ratios are calculated as respective
height over width.
The stapedial ratio is a value that often is used to distinguish marsupials from
placentals in systematic studies. Segall (1970) defined the ratio as the greatest height
versus width of the footplate of the stapes, and he concluded that marsupials tended
towards a more circular fenestra, and that placentals tended towards a more elliptical
fenestra. Subsequently, the ratio has been used in phylogenetic systematic analyses
(Archibald et al., 2001; Rougier et al., 1998; Wible et al., 2007).
RESULTS
The basic shape of the bony labyrinth of Monodelphis domestica remains constant
over the ontogenetic series studied (Figure 3.3). The specific measurements described
62
above were plotted over number of postnatal days in order to investigate whether specific
dimensions are correlated with maturity. Correlations between the measurements were
calculated from these graphs, and any correlations equaling 0.70 or above are considered
to be strong. There is no clear correlation between most of the angular dimensions
measured and age (Table 3.2), although the total angular deviation of the lateral
semicircular canal (r=0.72) expresses a strong correlation with the number of postnatal
days. The oldest M. domestica specimens tend to have lateral semicircular canals that
deviated above the canal plane to a greater degree than younger individuals.
Another measurement that is correlated strongly with age is the height of the arc
of the lateral semicircular canal (r=0.80), although the width (r=0.25), length (r=0.55),
and arc radius of curvature (r=0.54) are not (Table 3.3). The correlation between the
length of the slender lateral semicircular canal and age is not considered strong under the
parameters of the current study (r=0.67), but it is close. The coefficient of correlation
likely will change with an increase in sampling of individuals.
A strong correlation is observed between age and the length of the posterior
semicircular canal (r=0.79). In short, the posterior canal increases in length through the
age sequence studied, even though the volume and radius of curvature of the canal do
not. The length of the anterior semicircular canal is not correlated with age.
The length of the canaliculus cochleae also shows a positive correlation to age,
(r=0.86; Table 3.4). This bony tube for the perilymphatic duct increases in length as the
bony labyrinth matures. The canaliculus cochleae is included in the segmentation of the
cochlea, and therefore its volume is included in the cochlear volume. Despite the strong
correlation of the length of the canaliculus cochleae, the overall volume of the cochlea is
correlated with age only weakly (perhaps as a result of grouping the volume of the
canaliculus cochleae with that of the cochlea).
63
FIGURE 3.3. Comparison of endocasts of the bony labyrinths of Monodelphis domestica
across ontogenetic ages in ventral (A-H), lateral (I-P), and posterior views (Q-X). A, I, Q
– line drawing of bony labyrinth major features identified; B, J, R – TMM M 8265 (day
27); C, K, S – TMM M 7536 (day 48); D, L, T – TMM M 8266 (day 56); E, M, U –
TMM M 7542 (day 75); F, N, V – TMM M 8267 (day 76); G, O, W – TMM M 8268
(day 90); H, P, X –TMM M 8273 (day 465). Anterior is towards bottom in A-P, lateral
towards right in A-H, and Q-X, dorsal towards right in I-P, and ventral towards bottom
in Q-X. Abbreviations: aa – anterior ampulla; ac – anterior semicircular canal; av
aquaeductus vestibuli; cc – canaliculus cochleae; cr – common crus; co – cochlea; fc –
fenestra cochleae; fv – fenestra vestibuli; la – lateral ampulla; lc – lateral semicircular
canal; pa – posterior ampulla; pc – posterior semicircular canal; vb – vestibule.
A
Diagram Day 27 Day 48 Day 56 Diagram Day 27
Day 75 Day 76
1 mm
aa
ac
la
lc
pa
vb
co
co
co
vb
aa
ac
la
lc
pc
pa
cr
av
cc
fv
fc
cc
pa
pc
av
cr
lc
la
aa
ac
av
cr
pc
1 mm
1 mm
Day 90 Day 465 Day 48 Day 56
Daigram Day 27 Day 48 Day 56 Day 75 Day 76
Day 75 Day 76 Day 90 Day 465 Day 90 Day 465
B C D
E F G H
I J K L
M N O P
Q R
S T
U V
W X
64
TABLE 3.2. Angular measurements of the bony labyrinth across an ontogenetic series of
Monodelphis domestica. Measurements expressed in degrees, except for number of turns.
Abbreviations: ac, anterior semicircular canal; co – cochlea; lc – lateral semicircular
canal; pc – posterior semicircular canal; r – coefficient of correlation, calculated from
measurement over number of days, strong correlations italicized.
Specimen
Plane Angles
Deviation
Cochlea
(# Days)
ac-lc
ac-pc
lc-pc
ac
lc
pc
co-lc
Coiling
Turns
TMM M-7597 (27)
88.5
88.0
86.5
4.60
5.30
8.30
33.7
682
1.90
TMM M-8261 (27)
77.9
88.0
91.9
6.90
0.00
6.00
37.9
606
1.70
TMM M-8265 (27)
84.4
103
95.5
15.2
2.50
6.80
34.3
604
1.7
TMM M-7536 (48)
88.9
89.1
80.0
25.1
3.00
8.20
62.5
658
1.8
TMM M-8266 (56)
87.1
100
86.6
12.1
7.00
3.80
47.9
665
1.9
TMM M-7539 (57)
80.9
90.0
83.9
0.00
0.00
7.80
55.9
671
1.9
TMM M-7542 (75)
87.6
95.9
98.7
20.1
3.80
8.70
57.7
658
1.8
TMM M-8267 (76)
88.9
90.1
85.9
10.9
6.10
11.8
30.2
621
1.7
TMM M-7545 (90)
91.6
100
91.4
14.5
9.20
5.50
52.9
685
1.9
TMM M-8268 (90)
96.2
102
91.0
17.3
5.80
3.20
23.0
638
1.8
TMM M-8273
(465)
73.7
83.0
87.0
15.3
12.2
3.80
65.6
650
1.8
r
0.52
0.19
0.10
0.03
0.72
0.39
0.46
0.07
0.07
65
TABLE 3.3. Linear measurements of the semicircular canals across an ontogenetic series
of Monodelphis domestica. Measurements expressed in millimeters. Abbreviations: ASC
– anterior semicircular canal; d – diameter of canal; h – height of canal; l – length of
canal; LSC – lateral semicircular canal; PSC – posterior semicircular canal; R – arc
radius; r – coefficient of correlation, calculated from measurement over number of days,
strong correlations italicized; Spec. – specimen number (TMM M); w – width of canal.
Spec.
ASC
LSC
PSC
(#
Days)
l
h
w
R
d
l
h
w
R
d
l
h
w
R
d
7597
(27)
3.99
1.55
1.9
0.86
0.20
5.52
1.12
1.17
0.57
0.27
3.16
1.42
1.41
0.71
0.26
8261
(27)
3.18
1.8
2.02
0.96
0.21
1.91
1.08
1.21
0.57
0.24
2.92
1.39
1.64
0.76
0.18
8265
(27)
4.23
1.96
1.94
0.97
0.29
2.45
0.94
1.22
0.54
0.25
3.42
1.27
1.65
0.73
0.22
7536
(48)
4.58
1.68
2.04
0.93
0.19
2.85
1.15
1.08
0.56
0.19
3.95
1.45
1.45
0.81
0.24
8266
(56)
4.39
1.95
2.02
0.99
0.19
2.87
1.21
1.47
0.67
0.19
3.79
1.40
1.40
0.76
0.19
7539
(57)
4.35
1.80
2.02
0.96
0.21
2.51
1.03
0.99
0.51
0.19
3.69
1.39
1.40
0.73
0.22
7542
(75)
4.38
1.76
1.99
0.94
0.23
2.62
1.14
1.25
0.60
0.22
3.66
1.54
1.54
0.76
0.25
8267
(76)
4.4
1.85
2.09
0.99
0.17
2.84
1.21
1.44
0.66
0.19
3.85
1.44
1.67
0.78
0.22
7545
(90)
4.62
1.85
1.95
0.95
0.17
2.88
1.21
1.34
0.64
0.25
3.98
1.48
1.48
0.77
0.27
8268
(90)
4.68
2.19
2.14
1.08
0.17
2.56
1.15
1.37
0.63
0.20
4.17
1.56
1.75
0.83
0.20
8273
(465)
4.95
2.02
2.16
1.04
0.23
3.08
1.41
1.3
0.70
0.20
4.62
1.53
1.53
0.82
0.24
r
0.55
0.40
0.62
0.51
0.15
0.55
0.80
0.25
0.54
0.30
0.73
0.50
0.34
0.52
0.15
66
TABLE 3.4. Linear measurements of the cochlea and other morphological structures
across an ontogenetic series of Monodelphis domestica. Measurements expressed in
millimeters. Abbreviations: Aq. Vest. – aquaeductus vestibuli; Coch. Can. – canaliculus
cochleae; h – height; l – length; Labyr. – bony labyrinth; r – coefficient of correlation,
calculated from measurement over number of days, strong correlations italicized; w –
width.
Specimen
Cochlea
Can. Coc.
Aq. Vest.
Labyr.
Skull
(# Days)
l
h
w
l
l
l
l
7597 (27)
5.08
1.4
2.21
0.29
0.83
3.91
18.4
8261 (27)
4.20
1.71
1.92
0.18
-
3.21
NA
8265 (27)
5.19
1.50
2.42
0.22
0.84
3.41
NA
7536 (48)
5.37
1.38
2.33
0.32
0.87
3.38
23.8
8266 (56)
5.37
1.39
2.42
0.57
0.84
3.42
26.3
7539 (57)
5.09
1.10
2.10
0.32
0.82
3.36
26.4
7542 (75)
4.83
1.53
2.49
0.62
1.12
3.28
29.3
8267 (76)
5.8
1.66
2.44
0.56
0.96
3.43
31.0
7545 (90)
5.22
1.36
2.34
0.46
0.85
3.43
31.0
8268 (90)
5.57
1.56
2.51
0.58
1.29
3.56
34.2
8273 (465)
4.44
1.45
2.14
1.02
-
3.57
40.5
r
0.37
0.02
0.17
0.86
0.52
0.60
0.71
67
No strong correlations were observed between age and partial or total volumes
within the bony labyrinth (Table 3.5). The highest coefficient of correlation was
calculated for volume of the lateral semicircular canal (r=0.67), which is close to, but
does not meet, the criterion for being considered a strong correlation. Likewise, none of
the ratios or indices expressed a strong correlation with age.
Additional correlations also were found among the inner ear dimensions
themselves. For example, there is a correlation between the length of the inner ear and the
radius of the arcs of the three semicircular canals (anterior, r=0.86; lateral, r=0.72;
posterior, r=0.73). All of these values increase as the length of the inner ear becomes
larger. The total length of the bony labyrinth also is correlated with the length of the skull
(r=0.74), which in turn is correlated with number of postnatal days (r=0.71;
measurements in Table 3.6). Although there is a strong link between the lengths of the
inner ear and skull, the inner ear length does not correlate with age (r=0.60), as is the case
with the skull.
The lengths of each canal also correlate with the respective radii of curvature
(anterior, r=0.72; lateral, r=0.74; posterior, r=0.87), so that canals with large arc radii of
curvature have long slender canal lengths. An additional strong correlation is observed
between the aspect ratio of the anterior semicircular canal and the radius of its arc
(r=0.76), but there is no observed correlation between the radius of the arc of any
semicircular canal and its respective volume. Nor is a single canal more voluminous in
every specimen. Either the anterior or posterior semicircular canal expressed the highest
volume (each canal with the highest value in 5 of the 11 specimens). Unlike the volume,
however, the anterior semicircular canal had the longest arc radius in every individual
examined (see Table 3.3).
68
TABLE 3.5. Volumes of compartments within the bony labyrinth across an ontogenetic
series of Monodelphis domestica. Measurements expressed in cubic millimeters.
Abbreviations: aa – anterior ampulla; ac – anterior; Crus – common crus; la – lateral
ampulla; lc – lateral; pa – posterior ampulla; pc – posterior; r – coefficient of correlation,
calculated from measurement over number of days; SC – semicircular canals.
Specimen
Cochlea
SC
Ampullae
Crus
Vestibule
Total
(# Days)
ac
lc
pc
aa
la
pa
7597 (27)
2.23
0.08
0.06
0.06
0.16
0.14
0.23
0.08
0.68
3.73
8261 (27)
2.42
0.07
0.06
0.08
0.16
0.13
0.25
0.09
0.82
4.07
8265 (27)
2.92
0.08
0.06
0.08
0.2
0.12
0.27
0.08
0.95
4.76
7536 (48)
2.46
0.08
0.05
0.07
0.13
0.19
0.11
0.11
0.75
3.94
8266 (56)
2.59
0.06
0.05
0.09
0.16
0.11
0.18
0.09
0.8
4.14
7539 (57)
1.98
0.06
0.07
0.05
0.16
0.12
0.26
0.1
0.72
3.5
7542 (75)
2.33
0.08
0.05
0.07
0.15
0.11
0.22
0.15
0.88
4.05
8267 (76)
2.4
0.08
0.06
0.08
0.2
0.14
0.28
0.1
0.83
4.18
7545 (90)
2.41
0.08
0.06
0.08
0.14
0.12
0.23
0.12
0.71
3.94
8268 (90)
2.51
0.07
0.04
0.07
0.19
0.14
0.25
0.1
0.82
4.19
8273 (465)
2.55
0.07
0.03
0.05
0.12
0.11
0.15
0.09
0.69
3.85
r
0.12
0.33
0.67
0.52
0.48
0.3
0.42
0.12
0.39
0.6
69
TABLE 3.6. Volume percentages and ratios calculated for the bony labyrinth across an
ontogenetic series of Monodelphis domestica. Abbreviations: aa – anterior ampulla; ac –
anterior semicircular canal; cr – common crus; co – cochlea; la – lateral ampulla; lc –
lateral semicircular canal; pa – posterior ampulla; pc – posterior semicircular canal; r –
coefficient of correlation, calculated from measurement over number of days; vb –
vestibule.
Specimen
% Volume
Aspect Ratios
STR
(# Days)
aa
ac
cr
co
la
lc
pa
pc
vb
ac
lc
pc
co
7597 (27)
4.35
2.09
2.23
59.90
3.78
1.53
6.23
1.53
1.83
81.50
96.20
101
63.40
1.5
8261 (27)
3.99
1.80
2.13
59.40
3.26
1.43
6.08
5.85
2.01
89.20
89.10
84.60
89.10
1.4
8265 (27)
4.23
1.71
1.67
61.30
2.52
1.27
5.69
1.73
1.99
101
77.50
77.10
62.00
1.7
7536 (48)
3.37
2.00
2.66
62.40
4.87
1.17
2.74
1.70
1.91
82.20
107
81.40
59.20
1.5
8266 (56)
3.86
1.53
2.28
62.60
2.77
1.17
4.40
2.10
1.93
96.40
82.50
84.80
57.40
1.7
7539 (57)
4.60
1.60
2.74
56.50
3.40
1.94
7.28
1.37
2.06
88.70
104
90.00
52.40
1.5
7542 (75)
3.73
1.90
3.76
57.70
2.79
1.11
5.54
1.71
2.18
88.60
91.10
101
61.50
1.5
8267 (76)
4.84
1.86
2.41
57.40
3.37
1.38
6.79
2.02
1.99
88.50
84.10
85.80
68.00
1.6
7545 (90)
3.50
1.93
3.00
61.30
3.00
1.52
5.79
1.90
1.81
94.70
90.10
92.40
58.10
1.5
8268 (90)
4.42
1.66
2.53
59.90
3.25
1.00
5.95
1.72
1.96
102
83.90
88.80
62.10
1.5
8273
(465)
3.22
1.69
2.21
66.10
2.91
0.86
3.90
1.20
1.79
93.70
108
88.70
67.80
1.3
r
0.49
0.23
0.06
0.64
0.21
0.53
0.37
0.59
0.45
0.16
0.48
0.05
0.07
0.61
70
The aspect ratios of the semicircular arcs are not correlated with age, and neither
is the aspect ratio of the cochlea (Table 3.6). There is not a strong relationship between
maturity and volume, and the percentages that the various chambers of the inner ear
contribute to the bony labyrinth do not follow a pattern of change across ontogeny.
DISCUSSION
It is not surprising that there is a strong correlation between the length of the skull
and age, nor should it be surprising that there is a correlation between the length of the
inner ear and size of certain components such as the radii of the arcs of the semicircular
canals. Yet the results reveal that there is little appreciable change in the bony labyrinth
once ossification has spread through the otic region. One exception appears to be some of
the dimensions of the lateral and posterior semicircular canals.
The length of the posterior canal increases after birth, which is not the case for
any other features of this canal. The correlation between age and posterior canal length
does not necessarily signify an ontogenetic lengthening of the membranous canal,
because there is not an exact relationship between size of the bony semicircular canal and
the membranous duct within it. At one extreme, the cross-sectional area of the
membranous posterior semicircular duct is only 7% of the cross-sectional area of the
bony canal (Curthoys et al., 1977b). Although there is a correlation between the length of
the posterior canal and maturity, the strength of the correlation likely will decrease as
more individuals are added to the sample. Because different aspects of the semicircular
canals are correlated with age, the three canals develop in different fashions. In fact,
ossification of the canals is not synchronous in humans (Jeffrey and Spoor, 2004), and the
71
FIGURE 3.4. CT slices through the cochlea of Monodelphis domestica. A, axial slice #45
of TMM M 7595 (day 27); B, axial slice #34 of TMM M 7536 (day 48); C, axial slice
#71 of TMM M 7539 (day 57); D, axial slice #40 of TMM M 7542 (day 75); E, axial
slice #35 of TMM M 7545 (day 90); f axial slice #70 of TMM M 7599 (adult); G,
generalized diagram through cochlea. White arrow points to tympanic surface of bony
promontorium. Abbreviations: co – cochlea; cn – canal for cochlear branch of cranial
nerve VIII; st – stapes within fenestra vestibuli; vb – vestibule.
Day 90 AdultDay 75
D E F
Diagram
co
vb
st
cn
G
Day 27
A B
Day 57Day 48
C
1 mm 1 mm
72
membranous canals of Didelphis are formed at different times, with the lateral canal the
last of the three to fully develop (Larsell et al., 1935).
Although there are few morphologic changes among structures inside of the
petrosal across the growth series, there is a correlation between age and features on the
external surface of the petrosal. For example, the facial nerve canal of elephantoid
proboscideans exhibits ontogenetic variation (Chapter 2), and the length of the
canaliculus cochleae, which connects the inner ear and the cranial cavity, increases over
age in M. domestica (r=0.86). The petrosal grows via accretion of bone on external
surface of the petrosal (Figure 3.4), and the canaliculus cochleae elongates in order to
connect the inner ear and endocranial cavities.
Although most of the measurements taken in this study did not show a correlation
with the maturity of the individual, ranges of variation were observed in every dimension.
Measurements with varying values are not significant phenomena given that variation is a
natural occurrence, but differences in certain measurements used in phylogenetic studies
were observed, which has significant implications for the assessment of the evolutionary
relationships of mammals.
One measurement that is thought to be phylogenetically informative in
mammalian systematics (Rougier et al., 1998; Segall, 1970) is the stapedial ratio.
Marsupials possess a stapedial ratio less than 1.8 (Segall, 1970), whereas the ratio for
placentals is above 1.8 (although some eutherians are an exception; see also Ekdale et al.,
2004, and Wible et al., 2001). Indeed, the average stapedial ratio for M. domestica (1.5)
falls within the range for marsupials, but there is a range of variation in the ratio (1.3-1.7;
Table 3.6), related in part to changes in the width of the fenestra vestibuli (which has a
measurement range spanning 0.20 mm versus the span of 0.06 mm for the height). The
stapedial ratio itself is not strongly correlated with age (r=0.61), nor does any M.
73
domestica individual possess a ratio in the placental range of Segall (1970), but the
variation that is observed suggests that the use of the measurement could be problematic.
It appears that the ratio varies in many taxa, not just M. domestica. For example, large
ratios are reported for two Cretaceous mammal taxa (Ekdale et al., 2004), as well as a
sampling of elephantoid petrosals collected from Pleistocene deposits (Chapter 2).
The coiling of the cochlea is another character that was used previously in
phylogenetic analyses (e.g., Archibald et al., 2001; Rougier et al., 1998). The cochlea is a
more-or-less straight tube in non-mammalian synapsids, and slightly curved (but not
coiled) in modern monotremes (Graybeal et al., 1989). No therian mammal (Theria is the
group containing the most recent common ancestor of marsupials and placentals, plus all
of the descendants of that ancestor) possesses a cochlea that is coiled less than 360
°
(Meng and Fox, 1995), except for a few potential exceptions known for extinct forms.
One of these exceptions is the Cretaceous eutherian Uchkudukodon nessovi, the cochlea
of which is coiled at least 270
o
, but certainly does not complete an entire revolution
(McKenna et al., 2000).
A cochlea coiled at or just under 360° might separate crown placental mammals
from their closest fossil eutherian relatives. The cochlea of zalambdalestids, a Cretaceous
group of mammals, which may fall outside of crown Placentalia (Wible et al., 2005,
2007), completes around one turn (Chapter 4; Kielan-Jaworowska, 1984). Uchkudukodon
nessovi likely is outside of Placentalia as well. Broad ranges in cochlear coiling has
tremendous implications for this hypothesis, especially because the variation of cochlear
coiling in M. domestica spans 81°, nearly a full quarter of a turn (at the moment, it is
unclear what the range of cochlear coiling is in other taxa).
The state of coiling in U. nessovi provides an example of the ambiguity to which
variation can lead. The cochlea of U. nessovi may not coil 360°, or perhaps there is a
74
range in this measurement (similar to that observed in M. domestica) that might span the
360° boundary. If the latter is true, the cochleae of some U. nessovi individuals might
complete a full coil while the cochleae of different individuals do not. Variation in the
otic system can affect the results of phylogenetic analyses (depending on the observed
state in the specimen used to score the data matrix). Although the maturity of the
individual should not affect scoring of the cochlear coiling, the character does vary
among individuals, and the variation should be considered when scoring matrices.
The relationships among the three semicircular canals in M. domestica agree with
observations made on other vertebrates. The degree to which the anterior semicircular
canal deviates from the plane of the canal is greater than the values for either the lateral
or posterior semicircular canals (Figure 3.5). That is to say, the anterior canal is less
planar than the other two, with an average total deviation of 12.9° versus 5.0° and 6.7°
(for the lateral and posterior semicircular canals, respectively), which is a pattern that
also is observed in laboratory mice (Calabrese and Hullar, 2006). Further, the radius of
the arc of the anterior semicircular canal is the greatest of the three arcs in M. domestica
and in various other mammals as well, including rodents, carnivorans, and primates
(Blanks et al., 1972, 1975; Calabrese and Hullar, 2006; Curthoys et al., 1975; Hullar,
2006; Muren et al., 1986; Spoor and Zonneveld, 1998). The results of many studies (e.g.,
Calabrese and Hullar, 2006; Hullar, 2006; Hullar and Williams, 2006; Spoor et al., 2007)
support the hypothesis that the radius of curvature of the semicircular canals is related to
locomotor agility. Because these features of the inner ear are not correlated with age, the
eventual behavior associated with labyrinthine anatomy emerges after the morphology is
established in M. domestica.
75
FIGURE 3.5. Comparison of the shape of the semicircular canals of Monodelphis
domestica. An endocast of a semicircular canal is on the left, and the chart on the right
includes lines drawn through the lumen of the three semicircular canals of various M.
domestica individuals. Side view is a view perpendicular to the plane of the semicircular
canal. Top view is a view parallel to the plane of the semicircular canal. The ampulla
associated with the respective semicircular canal is to the right in all views. Angular
deviations of specific canals reported in Table 3.2. Abbreviations: ac – anterior
semicircular canal; lc – lateral semicircular canal; pc – posterior semicircular canal.
TMM M
8265
TMM M
7536
TMM M
8266
TMM M
7542
TMM M
8267
TMM M
8268
TMM M
8273
Composite
ac
lc
pc
Side
Top
Side
Top
Side
Top
Side view
Top view
76
The variation and correlations reported here may be influenced by several
additional factors. For instance, males tend towards larger bodies than females in M.
domestica (Macrini, 2004), but the opossum sample examined here is inadequate for
determining sexual dimorphism in the inner ear of M. domestica. Another factor affecting
the observed range of variation is an artifact of specimen sampling. With the addition of
more individuals, the correlation of these measurements might decrease, or else new
correlations may be recovered. Future studies focused on the otic region of M. domestica
will further elucidate the causes of variation exhibited by the inner ear
CONCLUSION
The results reported above contribute to the first reconstruction of the bony
labyrinth of Monodelphis domestica, a biomedically important marsupial mammal, which
is consistent with the anatomy and variation observed in placental mammals (Hullar and
Williams, 2006; Spoor et al., 2007). The results also demonstrate that, first, taxonomists
and functional anatomists need not worry about the maturity of individuals when dealing
with most metrics of the inner ear of didelphid marsupials (although the implication
might be indicative of a broader taxonomic pattern as there is little postossification
change in the bony labyrinth of rabbits; Hoyte, 1961) as long as the bony labyrinth is
ossified. In several cases, the adult form precedes the onset of function of the system.
Second, there is noticeable intraspecific variation that is not related to ontogeny. The
variation may be the result of one or any combination of several factors, and observed
variation could influence greatly the results of physiological and phylogenetic studies.
Although most of the dimensions are immune to ontogenetic changes once the
walls of the labyrinth have ossified, the variation observed within the inner ear of M.
77
domestica is an important biological phenomenon that cannot be ignored. Systematic
biologists have made implicit assumptions that the inner ear is invariant throughout
ontogeny. Most measurements do not vary throughout ontogeny, but there is a broad
range of variation in inner ear morphololgy within a single species. Without further
investigation into the variation within a species, use of measurements, such as the
stapedial ratio and coiling of the cochlea, in phylogenetic analyses is ill-advised.
Nonetheless, the majority of shape and size variation observed within the bony
labyrinth of M. domestica indicate intraspecific variation, regardless of ontogenetic age,
as opposed to ontogenetic trends in morphologic change. Much like the adult teeth in
mammals, once the petrosal is ossified, its internal morphology does not change. The
morphology of the system is locked in place before the animal achieves mature hearing or
locomotor potential. Unlike teeth, however, the inner ear does not wear as it is being
used. Barring any major pathology, the bony labyrinth should retain the same
morphology throughout the life of the animal. Therefore, the petrosal of a juvenile is as
valuable in interpreting behavior, physiology, and possibly phylogeny as is the petrosal
from an elderly individual.
78
CHAPTER 4: THE BONY LABYRINTH OF ZHELESTIDS
(MAMMALIA: EUTHERIA) AND OTHER MESOZOIC MAMMALS
ABSTRACT
Zhelestids are a group of eutherian mammals from the Late Cretaceous. High
resolution X-ray computed tomography was used to image the bony labyrinths of
zhelestid petrosals and to construct digital endocasts of the inner ear. The endocasts were
used to measure aspects of the inner ear that are behaviorally and phylogenetically
significant, including cochlear coiling, the radii of curvature of semicircular canal arcs,
and canal planarity.
The morphology of the labyrinth of zhelestids agrees with that of other extinct
eutherians, including Kulbeckia kulbecke, Ukhaatherium nessovi, and Zalambdalestes
lechei. Features of the labyrinth of zhelestids include a cochlea with one and a half turns
and a secondary common crus. Although isolated petrosals likely represent multiple
species because several species of zhelestids are recognized based on dental characters,
the degree of variation within the considered zhelestid petrosal sample is comparable to
that observed in extant species. Planarities of individual semicircular canals were the
most variable measurements. Coiling of the cochlea and arc radii of the semicircular
canals do not vary significantly in the specimens examined. Presence of the secondary
common crus and a cochlea with one and a half turns or less are plesiomorphic features
of the ear of eutherians. The morphology of the bony labyrinth does not suggest any
phylogenetic affinities of zhelestids, but the anatomy described will be useful for future
studies of the evolution and physiology of the ear of mammals.
79
INTRODUCTION
The end of the Cretaceous witnessed massive extinctions of numerous animal
groups (Raup and Sepkoski, 1982; Sheehan and Fastovsky, 1992), yet several dominant
life forms on the planet today originated during the Cretaceous, or shortly after the
Cretaceous-Tertiary boundary. Other major organisms are thought to have crossed the
boundary into the Cenozoic, including angiosperms (Boulter et al., 1998), birds (Cracraft,
1986; Cooper and Penny, 1997; Clarke et al., 2005), and mammals (Hedges et al., 1996;
Kumar and Hedges, 1998; Bininda-Emonds et al., 2007).
One group of mammals that plays an important role in the origin of placentals is a
paraphyletic group of eutherians from the Late Cretaceous of Laurentia (Kielan-
Jaworowska et al., 2004) known as zhelestids. Although zhelestids have been recovered
from Europe (Pol et al., 1992; Gheerbrant and Astibia, 1994) and North America
(Lillegraven, 1976; Fox, 1989; Cifelli, 1990), they are best known from Asia, particularly
from the Bissekty Formation in the Kyzylkum Desert of Uzbekistan (Nessov, 1985, 1993;
Nessov et al., 1998). Zhelestids contribute a rich biota to the Bissekty local fauna, where
at least four, but as many as six species of zhelestids are recognized among other
eutherian taxa (Archibald and Averianov, 2005, 2007).
The group “Zhelestidae” (sensu Averianov and Archibald, 2005) is paraphyletic
(see Wible et al., 2007), although the named zhelestid taxa recovered from the Bissekty
Formation (Aspanlestes aptap, Parazhelestes mynbulakensis, P. robustus, and Zhelestes
temirkazyk; Kielan-Jaworowska et al., 2004; Archibald and Averianov, 2005, 2007) form
an unnamed clade to the exclusion of other eutherian taxa (second analysis of Archibald
et al., 2001; Wible et al., 2007). Because the petrosal bones referred to zhelestids were
collected from the Bissekty Formation, the fossil ear regions likely represent one or more
of the species within the Bissekty zhelestid group. When speaking of zhelestids in the
80
present chapter, I am referring only to those taxa from the Bissekty local fauna that form
a monophyletic group to the exclusion of all other taxa, some of which previously were
grouped within “Zhelestidae” (e.g., Eozhelestes).
Like most Mesozoic mammals, the Bissekty zhelestids are represented
predominately by teeth (Archibald and Averianov, 2005), but other skeletal elements,
including postcrania (Chester et al., 2007, 2008) and petrosal bones (Ekdale et al., 2004),
have been referred to zhelestids as a group, although not to individual species. The
petrosals themselves are of particular interest, given the importance of the ear region in
evolutionary studies of mammals (e.g., Cifelli, 1982; Hunt, 1987; Norris, 1994). The
petrosals were referred to zhelestids based on relative abundance, size, and anatomy (see
Ekdale et al., 2004), and although none of the fossil petrosals were assigned to individual
zhelestid species recognized in the Bissekty local fauna, they were used in conjunction
with teeth in phylogenetic analyses exploring the relationships among Cretaceous
eutherian taxa and members of extant placental clades (Archibald et al., 2001; Wible et
al., 2007).
The phylogenetic affinities of the Bissekty zhelestid clade to other Mesozoic
mammals are contentious. The teeth of many zhelestids appear derived (Nessov et al.,
1998; Archibald et al., 2001), but the petrosals referred to Bissekty Formation zhelestids
retain features that are considered ancestral for therians, such as the prootic canal (Ekdale
et al., 2004). Using primarily data from teeth, competing hypotheses either include all
Bissekty zhelestid taxa as stem members of archaic ungulate lineages within crown
Placentalia (Archibald, 1996; Nessov et al., 1998; Archibald et al., 2001), or else exclude
zhelestid taxa from Placentalia altogether (Wible et al., 2007). In the latter case,
zhelestids occupy basal positions on the eutherian phylogeny. A final hypothesis that has
81
not been proposed is an inclusion of some zhelestid taxa within Placentalia, such as the
Bissekty clade, with an exclusion of other non-Bissekty zhelestids.
Because information from dental material and the middle ear region of the
Bissekty zhelestids has yielded conflicting results, additional data may provide crucial
information concerning the relationship among zhelestids and other eutherians, both
extinct and extant. Toward an ultimate goal of resolving the positions of the Bissekty
zhelestids, I herein describe new details of the inner ear of these mammals. The external
morphology of the petrosals is well described (Ekdale et al., 2004), but a thorough
description of the internal cavities of the petrosal is lacking. The inner ear is difficult to
study, because the structures are contained in cavities that are completely surrounded by
bone. With the use of high resolution X-ray computed tomography (CT), the bony
labyrinth is easily observed. Computed tomographic data was incorporated in
investigations of several previous studies focused on the inner ear of a variety of extinct
mammals (Luo and Ketten, 1991; Luo and Eastman, 1995; Hublin et al., 1996; Hurum,
1998; Spoor et al., 2002), and the sample of zhelestid petrosals from the Bissekty
Formation provides a unique opportunity to study the auditory and vestibular anatomy of
these enigmatic mammals.
Despite the difficulty in observing the bony labyrinth, the anatomy of the inner
ear was used previously to make physiological interpretations about fossil taxa, including
pterosaurs (Witmer et al., 2003), dinosaurs (Alonso et al., 2004; Clarke, 2005),
multituberculates (Miao, 1988; Meng and Wyss, 1995), early therians (Meng and Fox,
1995), primates (Spoor et al., 1994, 1996), bats (Simmons et al., 2008), and cetaceans
(Fleischer, 1976; Geisler and Luo, 1996; Luo and Marsh, 1996; Spoor et al., 2002).
Furthermore, certain aspects of the bony labyrinth are phylogenetically informative for
82
early humans (Hublin et al., 1996) and higher taxonomic levels (e.g., the cochlea differs
between therian and nontherian mammals; Rowe, 1986, 1988; Luo and Ketten, 1991).
Although additional work is needed to clarify the affinities of zhelestids, the goal
of the present paper is not to place the Bissekty zhelestids on a phylogenetic tree, nor is it
an attempt to assign isolated petrosal specimens to specific zhelestid taxa diagnosed by
dental morphology. Rather, the goal is to describe a novel body of data for zhelestids and
other Cretaceous eutherian taxa. The described anatomy will prove beneficial to future
studies investigating the phylogenetic and physiological implications of the inner ear, and
the variation observed across the zhelestid sample initiates a discussion of variation and
variability within an anatomical system that often is assumed to be morphologically
conservative. Variation within the inner ear of mammals is discussed in clinical literature
(Caix and Outrequin, 1979; Dimopoulos and Muren, 1990), and the phenomenon of
morphological variation may affect phylogenetic hypotheses in significant ways.
Institutional Abbreviations
PSS-MAE, Collections of Joint Paleontological and Stratigraphic Section of the
Geological Institute, Mongolian Academy of Science, Ulaanbaatar – American Museum
of Natural History, New York; TMM M, Texas Natural Science Center recent mammal
collections, Vertebrate Paleontology Laboratory; URBAC, Uzbekistan/ Russian/ British/
American/ Canadian joint paleontological expedition, Kyzylkum Desert, Uzbekistan,
specimens in the Institute of Zoology, Tashkent; ZIN C., Systematic Collections,
Zoological Institute, Russian Academy of Sciences, Saint Petersburg, Russia.
83
MATERIALS AND METHODS
Computed Tomography Methods
Seven isolated petrosals assigned to zhelestids from the Late Cretaceous Bissekty
Formation of Uzbekistan were CT scanned at the University of Texas High Resolution X-
ray CT facility (UTCT). A complete list of specimens scanned for this study, along with
the parameters used during scanning and post-scanning image processing, is provided in
Appendix 1.
The bony labyrinths of the Bissekty zhelestids were digitally segmented into the
constituent parts of the inner ear (e.g., cochlea and vestibule) in order to calculate partial
volumes for the inner ear cavities. Segmentation is accomplished by isolating the voxels
on CT slices that define the anatomical structure of interest (such as the anterior
semicircular canal). Individual CT slices are ‘stacked’, and the extracted volume of the
selected voxels is visualized as a 3-dimensional (3D) endocast that can be studied and
measured. Segmentation and endocast extractions were performed in the computer
programs VG Studio Max 1.2
©
(Volume Graphics) and Amira 3.1
©
. The bony channels
for the cochlear and vestibular aquaeducts were included in the segments of the cochlea
and vestibule respectively. Canals for branches of cranial nerve VIII were not segmented,
nor were they included in any measurement. Planar boundaries between the components
were maintained as much as possible. The medial border of the fenestra vestibuli (oval
window) was used as the divider between the cochlea and vestibule, with the entirety of
the opening within the vestibule.
84
Measurement Methods
Several angular, linear, and volumetric measurements were made using the Amira
software. Most of these measurements are used to interpret physiology and evolutionary
relationships of a wide variety of mammals. Dimensions of the structures of the bony
labyrinth were measured on the endocasts, and the inner ear morphology of the Bissekty
zhelestids was compared to original observations and published information about other
Cretaceous eutherians, namely the Mesozoic eutherians Ukhaatherium nessovi, Kulbeckia
kulbecke, and Zalambdalestes lechei (see Appendix 1 for specimens and scanning
parameters). Two skulls of Zalambdalestes were examined, for which two petrosals each
are preserved (four labyrinths total). However, PSS-MAE 130 is damaged through the
left petrosal, and portions of the left labyrinth of that specimen are missing. The four
labyrinths of Zalambdalestes were treated independently, as though they were from
isolated elements like the zhelestid sample. This allows me to explore cranial asymmetry
between right and left bony labyrinths from the same skull.
Measurement methodologies follow those conceived by Fleischer (1976) and
Spoor and Zonneveld (1995), unless otherwise noted. Anatomical terminology of the
bony labyrinth follows Sisson and Grossman (1938) and Evans (1993), and orientation
terminology within the cochlea follows Fleischer (1976).
Measurements of the cochlea involve the degree of coiling completed by the
canal, as well as gross length of the canal and widths of internal cochlear structures
(Figure 4.1). The degree of coiling completed by the cochlea is calculated following two
separate methods, one modified from that proposed by West (1985), and the other a
method employed by Geisler and Luo (1996). The two methods are the same
operationally, although they utilize different landmarks (Figure 4.1A). Both methods
draw a line from the axis of rotation to a secondary landmark at the base of the cochlea
85
FIGURE 4.1. Diagrams of measurements of cochlear dimensions. A, cochlea in
vestibular view with straight lines used to count number of turns that intersect axis of
rotation (at intersection of lines) and secondary landmarks of West (1985) and Geisler
and Luo (1996); B, cochlea in profile to measure height and width of cochlear spiral. C,
cross section through cochlea with expanded view of cochlear canal. Abbreviations:
1985, dashed line intersecting land mark of West (1985); 1996, solid line intersecting
landmark of Geisler and Luo (1996); cf, foramina within cribriform plate for nerves
connecting cranial nerve VIII and spiral ganglion; cl, dashed line representing length of
cochlea, approximating length of basilar membrane; cn, canal for cochlear branch of
cranial nerve VIII; co, cochlea; fc, fenestra cochleae; ht, height of cochlear spiral; lg,
laminar gap between primary and secondary bony laminae; pl, primary bony lamina; pt,
petrosal bone; sg, canal for spiral ganglion; sl, secondary bony lamina; wt, width of
cochlear spiral.
A
C
B
wt
ht
lg
sl
sl
pl
sg
co
cn
cn
cf
fc
cl
fc
pt
19961985
86
near the fenestra cochleae when the cochlea is in vestibular (‘map’) view. The number of
times that the cochlea crosses this line is counted and multiplied by 180° to provide a
gross value of cochlear coiling. An additional value is added to this product – an angle
measured between the line drawn through the cochlea and a line drawn from the center of
the axis of rotation to the most apical point of the cochlear canal. This second angle falls
on a plane parallel to the field of view (perpendicular to the axis of the cochlea). The total
measurement of coiling of the cochlea is expressed both as number of degrees and total
number of turns (total number of degrees divided by 360°).
The secondary landmark used by West (1985) is the point of inflection of the
cochlea near the vestibule, which at times is very close to the fenestra cochleae. Most
subsequent researchers used West’s landmark to measure the coiling of the cochlea (e.g.,
Wible et al., 2001). However, the exact point of inflection is somewhat arbitrary,
especially if the inflection is a continuous curve, rather than a sharp angle. The landmark
used by Geisler and Luo (1996) is at the beginning of the laminar gap, or the space
between the bony primary and secondary laminae on the axial and radial walls of the
cochlear canal respectively (Figure 4.1A, C). The basilar membrane, upon which the
organ of hearing sits in life, occupies the laminar gap. Because the membrane and
associated organ of hearing is the functional unit of the cochlea, it is more appropriate
biologically to start measuring the coiling of the cochlea at the beginning of this
functional unit, rather than a more arbitrary inflection in the canal. Nevertheless, many
researchers followed the method proposed by West (1985), so measurements obtained
using that method are reported here for comparative purposes alongside measurements
following the method of Geisler and Luo. Only the measurements following Geisler and
Luo (1996) are discussed in the text.
87
The width of the cochlea is a linear measure of the size of the basal turn of the
cochlea, and the height of the cochlea is a measure of the overall height of the cochlear
spiral (following Gosselin-Ildari, 2006). Both measurements fall on the plane that
intersects all points along the axis of cochlear rotation and the center of the fenestra
cochleae. The width of the cochlea is measured as the greatest distance from the
vestibular edge (closest to the apex of the spiral) of the fenestra cochleae to the radial
wall on the opposite side of the basal turn of the cochlear canal (Figure 4.1B). This
measurement is parallel to the plane of the basal turn of the cochlea. The height of the
cochlea is measured as the greatest vertical distance (perpendicular to width) from the
level of the most tympanal edge of the fenestra cochleae to the vestibular-most wall of
the cochlear spiral (within the apical turn). The aspect ratio of the profile of the cochlea,
calculated as the height over the width, may correlate to auditory ability (Gosselin-Ildari,
2006).
The overall length of the cochlear canal extends from the beginning of the laminar
gap (landmark used for measuring number of turns in the cochlea following Geisler and
Luo, 1996) to the apical most point of the canal, as measured with the SplineProbe tool in
the Amira software (Figure 4.1A). The length of the canal approximates the length of the
basilar membrane, which West (1985) argued is related to audible frequencies.
Multiple measurements were made for structures within the cochlear canal
(following Fleischer, 1976). These include the width of primary and secondary laminae,
laminar gap between the laminae, and diameter of the spiral nerve ganglion canal within
the primary lamina (Figure 4.1). The measurements were taken on two perpendicular
planes, each of which bisects the cochlea across the axis of rotation, which is exemplified
by the section through the cochlea of URBAC 03-39 depicted in Figure 4.1C. One plane
intersects the point at which the basilar membrane begins (same landmark used for
88
measuring coiling of the cochlea following Geisler and Luo, 1996), as well as the axis of
rotation. The second plane is perpendicular to the first, and measurements were taken
from cross-sections at each quarter turn throughout the length of the cochlea on both of
the planes. The width of the laminar gap was measured as the distance between the
primary and secondary laminae as long as both structures were present, and between the
primary lamina and the radial wall of the canal when the secondary lamina was absent
(the secondary lamina does not extend the entire length of the cochlea).
The internal cochlear measurements were used for graphical reconstructions of
the widths of the internal cochlear structures, following a method conceived by Guild
(1921), and later modified by Schuknecht (1953) and Wever et al. (1971 a, b). Such
reconstructions previously were drawn for multiple species of fossil and extant cetaceans
(Luo and Eastman, 1995; Geisler and Luo, 1996; Luo and Marsh, 1996). The widths of
the internal structures (laminar gap in particular) are related to inner ear function, and the
type of diagrams described here allow accurate observation of the relative proportions of
the two bony cochlear laminae. The longitudinal and transverse diameters of the basal
turn are constrained in these ‘cochlear maps’. Because parts of the apical turns of the
cochlea (beyond the basal turn) obstruct the basal turn in vestibular view, the turns
beyond the basal turn are distorted in the illustrations to allow observation of the cochlea
in its entirety.
One additional measurement of the cochlea is the angle between the basal turn of
the cochlea and the lateral semicircular canal. The lateral canal is parallel to the horizon
when an animal is at rest as evidenced by X-radiographs of several mammals (de Beer,
1947). Although right and left lateral canals are not always symmetrical within the same
skull (Caix and Outrequin, 1979), the angle between the cochlea and lateral canal
provides a rough estimate of the angle of the cochlea to the horizon. The average plane of
89
the cochlea intersects points at the center of the lumen of the basal turn of the cochlear
canal, and the plane of the lateral canal fits to points at the center of the lumen at the
midpoint of the canal and the entry points of the canal into the vestibule.
Measurements of the vestibular apparatus, which includes the semicircular canals,
include angles between the planes of the canals, the angular deviation of each canal from
its average plane, the arc radius of curvature of each canal, and the length of the slender
canals (i.e., the course of the canal that does not include the ampulla, sensu Boyer and
Georgi, 2007; Figure 4.2). Angles between the planes of the three semicircular canals are
measured with both planes perpendicular to the field of view. The average plane of a
canal is fit to points at the center of the lumen at the midpoint of the arc of the canal, and
the entry points of the canal into its ampulla at one end and vestibule at the other. The
common crus is included with the paths of both the anterior and posterior semicircular
canals, and the posterior ampulla is included with the path of the lateral semicircular
canal when a secondary common crus is present.
The slender canal is defined as the course of the canal that does not include the
ampullated portion (sensu Boyer and Georgi, 2007). The lengths of the three slender
semicircular canals were measured in the Amira software using the SplineProbe tool in
the same fashion used to calculate the length of the cochlear canal, and the curved line
created by the tool passes through he center of the lumen of the canal (Figure 4.2A).
The radius of curvature of a semicircular canal (‘R’ of Jones and Spells, 1963)
describes the size of the arc completed by the canal, following Spoor and Zonneveld
(1995). The radius is calculated as half the average of the height and width of the canal
arc (Figure 4.2A-B). The height of the anterior canal is measured as the greatest distance
from the wall of the vestibule to the center of the lumen of the canal, perpendicular to the
plane of the lateral semicircular canal. The height of the posterior canal is measured as
90
FIGURE 4.2. Diagrams of measurements of semicircular canal dimensions. A,
semicircular canal in profile displaying height and width of arc and length of slender
canal; B, semicircular canal in profile displaying radius of arc; C, semicircular canal
viewed when plane is parallel to horizon (solid line behind canal), displaying linear
deviations of canal from its plane. Total angular deviation equals the arcsine of linear
deviation 1 over radius plus arcsine of linear deviation 2 over radius. Abbreviations: am,
ampulla; ar, arc radius of curvature (‘R’ of Jones and Spells, 1963); cp, plane of
semicircular canal; ht, height of semicircular canal arc; lc, length of slender
(unampullated) semicircular canal; ld, linear deviation of semicircular canal from its
plane; sc, semicircular canal; vb, vestibule; wt, width of semicircular canal arc.
A
C
B
ht
ar
sc
sc
lc
am
am
vb
wt
ld 1
ld 2
cp
lc
lc
91
the greatest distance from the center of the lumen of the posterior limb of the canal to the
center of the lumen of the common crus, parallel to the plane of the lateral semicircular
canal. The height of the lateral canal is measured as the greatest distance from the wall of
the vestibule to the center of the lumen of the canal. The widths for all canal arcs are
measured as the greatest distance between the center of the lumen in opposing limbs of
the canal, and perpendicular to the height of the respective canal.
The total angular deviation of a semicircular canal from its respective average
plane is calculated using trigonometry and two linear measurements of the canal (adapted
from Calabrese and Hullar, 2006, and Hullar and Williams, 2006). The linear measures of
the canal are the arc radius of curvature of the canal and the linear deviation of the canal
from the plane. The linear deviation is measured when the plane is perpendicular to the
field of view (Figure 4.2C). The maximum distance between the center of the lumen of
the canal and the average plane are measured on both sides of the plane. The angular
deviation of the canal on each side of the plane is calculated as the arcsine of the linear
deviation divided by the arc radius of curvature. The total angular deviation is the sum of
the angular deviations on both sides of the plane.
Variation Methods
Because multiple species of zhelestids are recognized based on teeth recovered
from the same deposits as the petrosals within the Bissekty Formation, any variation in
the measurements described above for the Bissekty zhelestids might reflect species-level
differences. In order to explore whether or not the variation observed within the zhelestid
sample is any greater or less than that observed within a single species, endocasts were
constructed for multiple individuals of the extant placentals Dasypus novemcinctus and
92
Tadarida brasiliensis, as well as the extant marsupial Monodelphis domestica (see
Appendix 1). Although Monodelphis is not a eutherian mammal, the ontogeny of the
inner ear within didelphid marsupials is similar to the ontogeny of the auditory region in
placental mammals (Larsell et al., 1935; Reimer, 1996; Morsli et al., 1998). Because of
this, the patterns observed in M. domestica can be compared to that of other therian
mammals.
The coefficient of variation (CV) was calculated in order to visualize and compare
amounts of variation among the various taxa listed in Appendix 1. The value is reported
as a percentage of the standard deviation with respect to the arithmetic mean. A CV is a
standardized value, and it can be compared between systems with different units, as well
as significantly different means. Because the coefficient is a ratio of standard deviation to
mean, high CV value indicates a high degree of variation in the measurement under
question. A CV of 100 indicates that the standard deviation is equal to the mean.
BONY LABYRINTH OF ZHELESTIDS
The inner ear of mammals contains a series of connected soft tissue ducts and sacs
that together embody the membranous labyrinth that is housed in the bony canals and
cavities of the bony labyrinth. The membranous labyrinth is divided into an inferior
division that includes the membranous cochlear duct and saccule of the vestibule, and a
superior division that includes the utricle of the vestibule, the three membranous
semicircular ducts (anterior, posterior, and lateral) with their respective ampullae, and the
common crus between the anterior and posterior semicircular ducts. The bony
semicircular canals and cochlea of the bony labyrinth approximate the shape of the
membranous ducts, although not necessarily the size (see Curthoys et al., 1977b). The
93
bony vestibule is divided into an inferior spherical recess and superior elliptical recess
(see Figure 4.3). The spherical and elliptical recesses loosely correspond to the saccule
and utricle respectively, but there is little of the shape and size of the membranous
vestibule preserved within the bony vestibular cavity (as opposed to divisions observed
within the bony labyrinths of non-mammalian amniotes; Georgi and Sipla, 2008).
Unfortunately, the membranous labyrinth is not preserved in the fossils, so all
information gleaned from the inner ear of extinct mammals, such as zhelestids, is derived
from the bony labyrinth by necessity. Volumes of the cochlea, vestibule, and bony
labyrinth as a whole is presented in Table 4.1.
Cochlear Canal
Dimensions of the bony cochlear canal (housing the membranous cochlear duct)
are presented in Table 4.2. The cochlea of the Bissekty zhelestids communicates with the
middle ear cavity via the fenestra cochleae, which is located near the basal end of the
bony cochlear canal (Figure 4.3). A curved groove within the bony cavity is situated
internal to the fenestra cochleae. The groove is expressed as an outpocketing on the
endocast, and it likely accommodated the perilymphatic sac in life, because it leads into
the bony channel for the perilymphatic duct known as the canaliculus cochleae. The
canaliculus opens onto the external surface of the petrosal dorsomedial to the fenestra
cochleae. Further discussion and orientation of the canaliculus cochleae was provided by
Ekdale et al. (2004).
The cochlea contributes 60% of the total volume of the inner ear cavity. Table 4.1
includes a list of volumes of the cochlea and vestibule for Bissekty zhelestids. The
cochlear canal itself is coiled and it invariably turns medioventrally from the fenestra
94
FIGURE 4.3. Endocast of left zhelestid bony labyrinth (URBAC 03-39). Orientations are
approximated because petrosals are isolated from the cranium. A-C, digital endocasts; D-
F, line drawings of labyrinth with labels; A and D, anterior view with dorsal towards top
and lateral towards right; B and E, dorsal view with posterior towards top and lateral
towards right; C and F, lateral view with dorsal towards top and posterior towards right.
Abbreviations: aa, anterior ampulla; ac, anterior semicircular canal; av, aquaeductus
vestibuli; cc, canaliculus cochleae; co, cochlea; cr, common crus; er, elliptical recess of
vestibule; fc, fenestra cochleae; fv, fenestra vestibuli; la, lateral ampulla; lc, lateral
semicircular canal; pa, posterior ampulla; pc, posterior semicircular canal; pl, primary
bony lamina; ps, outpocketing for perilymphatic sac; scr, secondary common crus; sl,
secondary bony lamina; sr, spherical recess of vestibule.
A B C
D E F
co
co
co
sl
sl
sl
pl
av
av
cr
cr
ac
1 mm 1 mm
ac
ac
lc
lc
lc
pc
pc
pc
aa
aa
aa
la
la
la
cr
cc
cc
ps
ps
cc
scr
fc
fv
fc
pa
pa
er
er
er
sr
sr
sr
95
TABLE 4.1. Volume of the cochlea, vestibule (including semicircular canals and
ampullae), and entire bony labyrinth for zhelestids and other selected Cretaceous
eutherians listed in Appendix 1. Only specimens with a complete cochlea and vestibular
apparatus are reported. Dashes indicate measurements unavailable on account of poor
preservation of the specimen. Values expressed in mm
3
.
Specimen
Cochlea
Vestibule
Total
Zhelestid
URBAC 99-73
5.59
-
-
URBAC 03-39
3.65
2.02
5.67
URBAC 04-233
3.75
3.14
6.89
Zhelestid Average
4.15
2.58
6.28
Kulbeckia kulbecke
URBAC 98-113
0.667
-
-
URBAC 00-16
3.51
-
-
URBAC 02-56
3.13
-
-
URBAC 04-36
3.07
2.30
5.37
Kulbeckia Average
2.59
2.3
5.37
Ukhaatherium nessovi
PSS-MAE 110
1.23
0.943
2.17
Zalambdalestes lechei
PSS-MAE 108 - left
2.31
3
5.59
PSS-MAE 108 - right
2.47
3.05
5.52
PSS-MAE 130 - left
3.44
3.65
7.09
PSS-MAE 130 - right
3.41
-
-
Zalambdalestes Average
2.91
3.23
6.07
96
TABLE 4.2. Dimensions and orientations of the cochlea of zhelestids and other
Cretaceous eutherians. Coiling follows West (1985; C-85) and Geyser and Luo (1996; C-
96), expressed in degrees. Number of turns calculated from coiling following Geisler and
Luo (1996). Length, height, and width expressed in millimeters. Angle refers to angle
between the cochlea and plane of the lateral semicircular canal, expressed in degrees.
Specimen
C-85
C-96
# Turns
Length
Height
Width
Angle
Zhelestid
URBAC 99-02
498
532
1.5
4.40
1.27
2.84
32.2
URBAC 99-41
505
531
1.5
4.57
1.15
2.98
34.4
URBAC 99-73
567
596
1.7
5.45
1.60
2.91
-
URBAC 03-39
511
545
1.5
5.17
1.37
3.01
31.8
URBAC 04-233
532
547
1.5
5.08
1.33
3.05
36.2
ZIN C. 85511
500
523
1.5
4.78
1.55
2.81
40.9
ZIN C. 85512
514
540
1.5
5.05
1.22
2.96
28.5
Zhelestid Average
518
545
1.5
4.93
1.36
2.94
34.0
Kulbeckia kulbecke
URBAC 98-113
385
390
1.1
2.94
0.770
1.84
16.4
URBAC 00-16
464
487
1.4
4.52
1.29
2.96
16.0
URBAC 02-56
429
460
1.3
4.53
1.34
2.86
5.54
URBAC 04-36
427
448
1.2
4.46
1.28
2.87
10.3
Kulbeckia Average
426
446
1.2
4.11
1.17
2.63
12.1
Ukhaatherium nessovi
PSS-MAE 110
371
380
1.1
2.77
0.830
2.40
6.63
Zalambdalestes lechei
PSS-MAE 108 - lt
360
365
1.0
3.35
0.920
2.84
10.1
PSS-MAE 108 - rt
360
382
1.1
3.27
0.960
2.94
11.4
PSS-MAE 130 - lt
340
350
0.97
3.49
1.19
2.81
19.2
PSS-MAE 130 - rt
360
376
1.0
3.47
1.18
3.17
13.1
Zalambdalestes Average
355
368
1.0
3.40
1.06
2.94
13.5
97
cochleae. The cochlea of the zhelestids completes nearly 1.5 turns on average (518º),
which is a slightly larger value than previous estimates made using damaged specimens
(Ekdale et al., 2004). The average length of the cochlea is 4.93 mm. The cochlear canal
as a whole is planispiral in profile, and the average aspect ratio of the cochlea (height
versus width) is 46.3. An osseous wall separates the basal and apical turns of the cochlear
spiral, as illustrated in Figure 4.1C. The cochlea is inflected near the confluence of the
cochlear canal and vestibule, lateral to the fenestra cochleae. The plane of the basal turn
of the cochlea is not parallel to the plane of the lateral semicircular canal in zhelestids,
but rotated ventrally 34°.
In life, the spiral organ of hearing rests upon a soft tissue structure known as the
basilar membrane, which is supported by a bony primary spiral lamina on the axial
surface of the cochlear curve, and either a secondary lamina or the wall of the canal on
the radial surface of the curve (Figure 4.1C). Within the primary lamina is a canal for the
ganglion of the spiral nerve. The CT scan slices reveal a cribriform plate between the
canal for the spiral ganglion and the canal for the auditory branch of cranial nerve VIII,
which is housed within the modiolus (bony axis) of the cochlea (Figure 4.1C). The
cribriform plate is perforated with numerous foramina for branches of the auditory nerve
leading into the spiral nerve ganglion.
The widths of the primary and secondary bony laminae, laminar gap, and spiral
ganglion canal at each quarter turn are listed in Table 4.3. The primary spiral lamina
extends along the inside of the coil for most of the length of the cochlea, as is illustrated
in the graphical reconstructions provided in Figure 4.4. The lamina is broadest at the base
of the cochlea and it decreases gradually before disappearing near the tip of the canal.
Likewise, the diameter of the canal for the spiral nerve ganglion, which is surrounded by
98
TABLE 4.3. Dimensions of internal structures of the cochlea at each quarter turn of
coiled canal for zhelestids and other Cretaceous eutherians. Measurements for
Zalambdalestes lechei only taken for the right labyrinth of PSS-MAE 108 and left
labyrinth of PSS-MAE 130 on account of poor preservation. Internal structures not
preserved in CT scans of Ukhaatherium nessovi. Dashes signify absence of the structure
in that turn of cochlea. Measurements expressed in millimeters.
Cochlea turns
1/4
2/4
3/4
4/4
5/4
Zhelestid
Primary lamina width
0.25
0.22
0.19
0.15
0.13
Secondary lamina width
0.11
0.06
0.04
-
-
Laminar gap width
0.45
0.46
0.48
0.51
0.54
Diameter of spiral ganglion canal
0.18
0.13
0.13
0.11
0.09
Kulbeckia kulbecke
Primary lamina width
0.32
0.29
0.28
0.28
-
Secondary lamina width
0.1
0.05
0.03
-
-
Laminar gap width
0.28
0.35
0.37
0.47
-
Diameter of spiral ganglion canal
0.19
0.19
0.19
-
-
Zalambdalestes lechei
Primary lamina width
0.40
0.35
0.26
-
-
Secondary lamina width
0.06
-
-
-
-
Laminar gap width
0.38
0.43
0.50
0.48
-
Diameter of spiral ganglion canal
0.21
0.16
0.13
-
-
99
FIGURE 4.4. Graphic reconstructions of internal structures of the cochlea of zhelestid
specimens following methods of Guild (1921), Schuknecht (1953), and Wever et al.
(1971 a, b). Internal structures are poorly preserved for URBAC 99-41, and the cochlea is
not reconstructed for that specimen. A, URBAC 99-02; B, URBAC 99-73; C, URBAC
03-39; D, URBAC 04-233; E, ZIN C. 85511; F, ZIN C. 85512; G, Average across
sample with labels. Orientations of B, C, E, and F reversed for comparison. Longitudinal
and transverse diameters of basal turn conserved, in each image, but apical turn distorted
to allow observation of complete canal. Abbreviations: fc, fenestra cochleae; fv, fenestra
vestibuli; lg, laminar gap (black band); pl, primary bony labyrinth (white band); sl,
secondary bony labyrinth (gray band).
sl
lg
pl
1 mm
fc
fv
G
A
C
E
B
D
F
100
the primary lamina, is largest near the base of the cochlea and tapers similarly before it
disappears near the apical terminus of the primary lamina (Table 4.3).
The secondary bony lamina is present in the zhelestid inner ear, but the osseous
ridge, which is expressed on the endocast as a groove along the outside surface of the
cochlea (Figure 4.3), does not extend much beyond 1/2 to 3/4 of the basal turn of the
cochlea (Table 4.3). The basilar membrane would attach to the radial wall of the canal
after the secondary lamina disappears within the last half or quarter of the basal turn to
the apex of the cochlea. As with the primary lamina, the secondary lamina is broadest at
the base of the cochlea and diminishes rapidly before it disappears.
The laminar gap is less than 0.5 mm at the base of the cochlea in all of the
zhelestids examined, and it widens towards the apex (as the primary and bony laminae
decrease in width; Figure 4.4). The widening of the laminar gap is punctuated within
individual specimens, rather than a continuous increase to the end of the canal. However,
when the laminar gap is averaged across the zhelestid sample, there is a continuous
increase in width (Table 4.3).
Vestibule and Semicircular Canals
The vestibule is partially divided in the Bissekty zhelestids into the small
spherical recess and the larger elliptical recess, where the spherical is continuous with the
cochlea and communicates with the middle ear cavity via the fenestra vestibuli (Figure
4.3). The elliptical recess is mediolaterally extended with two lateral and three medial
openings. The lateral openings are for the lateral and anterior ampullae, with the opening
for the lateral ampulla closer to the spherical recess and fenestra vestibuli. Those
openings are located in a separate depression within the elliptical recess (that depression
101
appears as an outpocketing of the vestibule in the endocast). The three openings that
penetrate the opposite wall of the vestibule are entrances into the posterior ampulla,
common crus, and the much smaller opening for the aquaeductus vestibuli posterodorsal
to the common crus. A secondary common crus that connects the lateral and posterior
semicircular canals to the posterior ampulla is observed in the zhelestid labyrinth, and
there is no separate opening into the vestibule for the lateral semicircular canal (Figure
4.3).
The planes of the semicircular canals form approximately right angles with each
other, and although none of the canals fit perfectly on a single plane, the average total
angular deviation of the canals from their planes is not great (Table 4.4). The radius of
curvature of the anterior semicircular canal is larger than the radii of the lateral and
posterior canals in all of the zhelestid labyrinths examined, and the radius of the lateral
canal arc was always the smallest (Table 4.5). The same pattern was observed for the
length of the slender canal.
COMPARISON WITH CRETACEOUS EUTHERIANS
Eutheria is an inclusive group that contains crown Placentalia, plus all taxa more
closely related to Placentalia than to crown Marsupialia. The overall morphology of the
inner ear of zhelestids from the Bissekty Formation is the same as that observed in extant
placental mammals, as well as Mesozoic eutherians that may or may not be members of
crown Placentalia, namely Kulbeckia kulbecke, Ukhaatherium nessovi, and
Zalambdalestes lechei (Figure 4.5). However, there are differences among the Cretaceous
taxa examined (Tables 4.1-4.55). For example, the cochleae of the zhelestids coil to a
102
TABLE 4.4. Orientations of the semicircular canals for zhelestids and selected
Cretaceous eutherians. Abbreviations: A, anterior semicircular canal; A-L, angle
between planes of anterior and lateral semicircular canals; A-P, angle between planes of
anterior and posterior semicircular canals; L, lateral semicircular canal; L-P, angle
between lateral and posterior semicircular canals; lt, left labyrinth; P, posterior
semicircular canal; rt, right labyrinth. Canal plane angles and angular deviations
expressed in degrees. Canal lengths and arc radii expressed in millimeters. Measurements
not reported for URBAC 98-113 (Kulbeckia kulbecke) because little more than the
cochlea is preserved. Only anterior canal complete in URBAC 99-73 (zhelestid).
Definitions of measurements in text.
Specimen
Angles
Deviations
A-L
A-P
L-P
A
L
P
Zhelestid
URBAC 99-02
88.9
102
94.3
9.44
5.95
15.3
URBAC 99-41
85.1
98.1
97.3
17.0
12.1
28.0
URBAC 99-73
-
-
-
-
-
-
URBAC 03-39
102
90.8
84.7
11.6
2.92
17.3
URBAC 04-233
86.0
94.3
88.9
17.5
2.88
8.90
ZIN C. 85511
88.7
107
97.1
-
11.0
-
ZIN C. 85512
82.2
88.5
96.1
8.75
6.43
6.66
Zhelestid Average
88.8
96.8
93.1
12.9
6.88
15.2
Kulbeckia kulbecke
URBAC 00-16
87.6
90.0
86.7
-
2.52
5.51
URBAC 02-56
65.6
77.5
90.1
-
3.06
-
URBAC 04-36
86.6
84.8
92.0
11.1
2.52
4.67
Kulbeckia Average
79.9
79.9
89.6
11.1
2.70
5.09
Ukhaatherium nessovi
PSS-MAE 110
88.8
105
88.4
8.22
6.21
9.92
Zalambdalestes lechei
PSS-MAE 108 (lt)
71.6
85.8
83.4
7.90
4.75
6.21
PSS-MAE 108 (rt)
82.1
91.9
82.8
6.85
4.88
6.20
PSS-MAE 130 (lt)
86.3
91.6
89.4
7.33
9.33
8.14
PSS-MAE 130 (rt)
84.0
91.5
86.6
1.23
-
-
Zalambdalestes Average
81.0
93.6
85.6
5.83
6.32
6.85
103
TABLE 4.5. Orientations of the semicircular canals for zhelestids and selected
Cretaceous eutherians. Abbreviations: A, anterior semicircular canal; L, lateral
semicircular canal; lt, left labyrinth; P, posterior semicircular canal; rt, right labyrinth.
Canal plane angles and angular deviations expressed in degrees. Canal lengths and arc
radii expressed in millimeters. Measurements not reported for URBAC 98-113
(Kulbeckia kulbecke) because little more than the cochlea is preserved. Only anterior
canal complete in URBAC 99-73 (zhelestid). Definitions of measurements in text.
Specimen
Arc Radii
Lengths
A
L
P
A
L
P
Zhelestid
URBAC 99-02
1.16
0.773
0.865
5.69
3.63
4.70
URBAC 99-41
1.09
0.758
0.888
5.25
3.31
4.83
URBAC 99-73
1.22
-
-
6.71
-
-
URBAC 03-39
1.19
0.785
0.865
5.87
3.38
4.53
URBAC 04-233
1.18
0.795
0.840
5.72
3.41
4.49
ZIN C. 85511
1.15
0.838
0.925
-
3.60
-
ZIN C. 85512
1.18
0.803
0.863
5.54
3.58
4.55
Zhelestid Average
1.17
0.792
0.864
5.80
3.49
4.62
Kulbeckia kulbecke
URBAC 00-16
-
0.910
0.936
-
4.07
4.63
URBAC 02-56
-
0.937
-
-
4.03
-
URBAC 04-36
1.19
0.910
0.983
5.70
3.72
4.47
Kulbeckia Average
1.19
0.919
0.960
5.70
3.94
4.55
Ukhaatherium nessovi
PSS-MAE 110
0.837
0.739
0.694
3.81
3.16
3.39
Zalambdalestes lechei
PSS-MAE 108 (lt)
1.45
1.21
1.20
7.13
5.33
5.75
PSS-MAE 108 (rt)
1.51
1.18
1.20
7.14
5.17
6.18
PSS-MAE 130 (lt)
1.49
1.23
1.20
6.53
5.09
5.61
PSS-MAE 130 (rt)
1.40
-
-
6.86
-
-
Zalambdalestes Average
1.21
1.20
1.21
6.92
5.20
5.85
104
FIGURE 4.5. Bony labyrinths of Cretaceous eutherians listed in Appendix 1 in lateral
view. A, line drawing of zhelestid (URBAC 03-39); B, Ukhaatherium nessovi (PSS-
MAE 110); C, Zalambdalestes lechei (PSS-MAE 108), right labyrinth reversed for
comparison; D, Kulbeckia kulbecke (URBAC 04-36), reversed for comparison. Blocky
appearance of PSS-MAE 110 and 108 indicative of smaller data size (512 vs. 1024) and
datasets with fewer slices through labyrinth (see Appendix 1). Abbreviations: aa,
anterior ampulla; ac, anterior semicircular canal; av, aquaeductus vestibuli; cc,
canaliculus cochleae; co, cochlea; cr, common crus; er, elliptical recess of vestibule; fc,
fenestra cochleae; fv, fenestra vestibuli; la, lateral ampulla; lc, lateral semicircular canal;
pa, posterior ampulla; pc, posterior semicircular canal; ps, groove for perilymphatic sac;
sl, secondary bony lamina; sr, spherical recess of vestibule.
co
sl
av
cr
ac
lc
1 mm
pc
aa
la
ps
cc
fv
fc
pa
er
sr
A B
C D
105
greater degree than what is observed in Ukhaatherium and Zalambdalestes, both of which
have cochleae that complete only a single coil following the measurement method of
Geisler and Luo (1996). The cochlea of Kulbeckia coils more than Ukhaatherium and
Zalambdalestes, but not as much as zhelestids. The degree of coiling observed within Z.
lechei agrees with the single complete turn reported for a natural endocast of the inner ear
from the Cretaceous of Mongolia (Kielan-Jaworowska, 1984). However, the left cochlea
of PSS-MAE 130 (Zalambdalestes) completes less than one turn, even though the right
ear of the same individual coils over 360°. The right ear of that particular specimen is
damaged having a fracture that traverses the vestibule in the CT data, but the cochlea
appears to be intact.
The extent of the bony secondary lamina in Kulbeckia is similar to that in
zhelestids, but the secondary lamina only is prolonged past the first quarter turn in
Zalambdalestes (Table 4.3). The internal cochlear structures were not preserved well
enough in Ukhaatherium to appear on the CT scans with a resolution sufficient to permit
reliable measurements. The primary lamina extends past the fifth quarter in at least one
Kulbeckia specimen (URBAC 02-56), which is the condition that is observed in most of
the Bissekty zhelestid labyrinths. The laminar gap between the primary and secondary
bony laminae is narrowest at the base of the cochlea in all of the taxa examined, and the
width of the membrane increases towards the apex. However, the gap increases at a faster
rate in Kulbeckia and Zalambdalestes than it does in zhelestids (see Table 4.3).
The cochleae of all of the Mesozoic eutherians are planispiral in profile, and the
smallest height-to-width ratio was observed in the left cochlea of PSS-MAE 108
(Zalambdalestes; calculated from measurements in Table 4.2), although the ratio in
Ukhaatherium (34.6) is similar to PSS-MAE 108 (32.4-32.7), and less than what is
observed in the other Zalambdalestes specimen (36.2-42.3).
106
The rotation of the plane of the basal turn of the cochlea from the plane of the
lateral semicircular canal is greater in every zhelestid labyrinth than it is in any of the
other Cretaceous specimens studied. The least amount of rotation of the basal plane of the
cochlea was observed in the bony labyrinth of URBAC 02-56 (Kulbeckia), followed by
Ukhaatherium. Among the non-zhelestids, the highest degree of rotation (19.2°) was
observed in the left ear of PSS-MAE 130 (Zalambdalestes), compared to the least amount
of rotation in any zhelestid specimen (28.5°), which was ZIN C. 85511.
The gross morphology of the vestibular apparatus is in general accordance among
all of the Cretaceous eutherians listed in Appendix 1, including zhelestids. The secondary
crus was observed in every specimen, and the radius of curvature and length of the
anterior semicircular canal always was the largest of the three canals within any single
labyrinth. However, some minute differences among the sample were detected in specific
dimensions of the inner ear. For example, the planes of the semicircular canals of
zhelestids are closer to 90° than any of the other Cretaceous taxa. The planes of the
anterior and posterior semicircular canals form the greatest angle between any canal
planes in Ukhaatherium (105°), but the planes between the lateral canal and both the
anterior and posterior canals of Ukhaatherium are near 90° (89 and 88° respectively).
MORPHOLOGICAL VARIATION
Variation was observed among all measurements taken on the Bissekty zhelestid
sample, as well as among the Kulbeckia kulbecke specimens and between the two
Zalambdalestes lechei individuals used in this study (Tables 4.1-4). In fact, the two skulls
of Zalambdalestes exhibit cranial asymmetry having dramatically different dimensions
and orientations between right and left labyrinths within a single skull.
107
Although variation was observed in the measurement data, there was no variation
in the presence or absence of structures. For instance, the secondary common crus was
present in every Cretaceous petrosal for which that particular region of the labyrinth was
preserved. Further, the cochlea had a complete coil in almost every specimen (the only
exception being the left ear of PSS-MAE 130, although the opposing ear completes over
one full turn), even if to differing extents, and the canal always curved medially from the
fenestra vestibuli.
Coefficients of variation were not calculated for every measurement, but rather
those that commonly are reported in literature. The highest CV for each measurement is
italicized in Table 4.6, in which reports the CV values for cochlear dimensions and
orientations of the semicircular canals are summarized. No single extant species
expressed the most variation (highest CV value) for all of the measurements, although
patterns are observed among the data. For example, the majority of high CV values were
calculated for Monodelphis domestica. Zhelestids as a whole did not have the highest CV
value for any measurement, and the variation within the labyrinth dimensions of
zhelestids falls within the range of a single species (based on the extant sample).
Coiling of the cochlea expressed relatively low levels of variation across the
sampled taxa, especially the CV calculated for Dasypus novemcinctus. The most
variation in cochlear coiling, as well as in length of the cochlear canal (basilar
membrane) was observed in Kulbeckia.
The arc radius of curvature of the semicircular canals expressed relatively low
amounts of variation. The CV of the arc radius of the lateral canal was less than 10.0% in
all of the specimens studied. Likewise, the angles between the average planes of the three
semicircular canals expressed low-level CV values. The same is not the case for the
angular deviation of the canals, where these coefficients expressed the highest degree of
108
TABLE 4.6. Coefficients of variation (CV) for measurements of labyrinth dimensions of
Cretaceous eutherians (including zhelestids) and selected extant taxa listed in Appendix
1. Coefficient calculated as standard deviation divided by mean and multiplied by 100
(reported as percentage). Largest CV for each measurement is italicized. Abbreviations:
A, anterior semicircular canal; A-L, angle between planes of anterior and lateral
semicircular canals; A-P, angle between planes of anterior and posterior semicircular
canals; C-L, angle between planes basal turn of cochlea and lateral semicircular canal; L,
lateral semicircular canal; LP, angle between lateral and posterior semicircular canals; P,
posterior semicircular canal.
Specimen
Cochlea
Coiling
Length
C-L
Zhelestid
4.42
7.43
9.19
Kulbeckia kulbecke
9.17
19.0
4.96
Zalambdalestes lechei
3.80
3.07
29.9
Dasypus novemcinctus
0.923
5.69
4.10
Monodelphis domestica
4.38
9.15
31.6
Tadarida brasiliensis
1.15
4.97
6.41
Plane Angles
A-L
A-P
L-P
Zhelestid
7.79
7.22
5.53
Kulbeckia kulbecke
15.5
3.43
8.59
Zalambdalestes lechei
8.02
3.26
3.58
Dasypus novemcinctus
14.5
7.55
12.8
Monodelphis domestica
7.43
7.30
6.02
Tadarida brasiliensis
7.08
6.22
4.32
Deviations
A
L
P
Zhelestid
32.4
55.5
55.1
Kulbeckia kulbecke
-
13.3
8.32
Zalambdalestes lechei
53.1
41.2
16.4
Dasypus novemcinctus
22.0
21.6
20.0
Monodelphis domestica
55.6
98.3
38.2
Tadarida brasiliensis
60.8
86.4
82.3
Length
A
L
P
Zhelestid
8.52
3.89
3.06
Kulbeckia kulbecke
-
4.83
2.50
Zalambdalestes lechei
4.12
2.39
5.04
Dasypus novemcinctus
6.97
3.89
3.06
Monodelphis domestica
10.6
11.9
12.6
Tadarida brasiliensis
6.27
4.82
4.45
109
TABLE 4.6. (Continued)
Specimen
Arc Radii
A
L
P
Zhelestid
3.66
3.49
3.33
Kulbeckia kulbecke
-
1.72
3.45
Zalambdalestes lechei
3.15
2.29
0.112
Dasypus novemcinctus
7.70
5.53
6.36
Monodelphis domestica
6.00
9.58
4.81
Tadarida brasiliensis
2.28
5.68
4.13
110
variation among the coefficients that were calculated. In particular, the angular deviation
of the lateral semicircular canal of Monodelphis was nearly the maximum amount (over
98%).
DISCUSSION
Morphological variation is a natural occurrence to which no anatomical system is
immune. Structures of the inner ear that express variation pertain to function of the
cochlea and vestibular apparatus, and other structures have been used to assert broad
evolutionary patterns concerning the history of mammal lineages through deep time.
Before anatomy of the otic region can be used to interpret physiology and/or phylogeny,
the issue of variation must adequately be explored.
Variation Considerations
The observed variation within the Bissekty zhelestid labyrinth could be the result
of several factors, including ontogeny, phylogeny, or individual variation. While the
external surface of the petrosal exhibits ontogenetic variation, as exemplified by a growth
series of petrosals referred to Elephantoidea (Proboscidea; Chapter 2), dimensions of the
internal surface of the petrosal (bony labyrinth) do not change significantly once the
walls of the labyrinth are ossified early in ontogeny (Chapter 3). Accordingly, variation
within the inner ear of zhelestids is not the result of differences in maturity of individuals
examined.
The isolated petrosals of zhelestids were collected from the same horizons within
the Bissekty Formation as teeth that represented different zhelestid taxa, and it is likely
that the zhelestid petrosals represent more than one of the Bissekty species. The
111
coefficient of variation (CV) was calculated to compare amounts of variation among
individual extant species and zhelestids. If the CV values of zhelestids are higher than
those for extant species, then it is possible that the variation observed in the zhelestid
labyrinths reflects multiple species within the sample (see Cope and Lacy, 1995;
Carrasco, 1998).
The amount of variation observed within the Bissekty zhelestid inner ears is no
greater than that in the original observations of the extant species Dasypus novemcinctus,
Monodelphis domestica, and Tadarida brasiliensis. The variation observed in the extant
sample parallels that calculated from measurements of additional placental species
reported elsewhere (published taxa and literature sources are listed in Appendix 2).
Although the variation within the bony labyrinth of zhelestids is within the range of
extant species, this does not necessarily signify that all of the petrosals are the same
species, or even a particular higher taxon. Although several previous workers
documented variation within the inner ear of single species (e.g., Muren et al., 1986), no
study has attempted to determine strictly at which taxonomic level the inner ear varies.
To what degree the arc radius of curvature varies at the generic level, or at the ordinal
level for that matter, is unknown.
The results of this variation study are significant for one important reason.
Although the morphology of the inner ear varies (which is not significant in and of itself),
the variation does not appear to be great within a species for the most part (keeping in
mind that only three species were evaluated). Variation does not pose a problem when
using measurements, such as the arc radius of a semicircular canal or the angles between
two canals, to make behavioral interpretations of extinct mammal species. However, the
planarity of single canals is a variable feature. The total angular deviation may not be
much, but the variation of this measure is considerably high. The physiological
112
significance of canals that deviate from a single plane is not well known, but such
deviation likely affects the sensitivity of the canals. Whatever correlation the deviation a
canal from its plane has with physiology should be approached with caution, especially
given that the standard deviation of the lateral canal varies almost as much as possible
within Monodelphis; there is little room for the planarity of the lateral canal to vary any
more than it already does. Fortunately, the other dimensions described here can be used
for physiology and evolutionary studies without much issue.
Physiological Considerations
The arc radius of a semicircular canal is positively correlated to the afferent
sensitivity of the canal (Yang and Hullar, 2007). That is, semicircular canals with large
arc radii are more sensitive than those with smaller radii. The largest canals among the
Cretaceous sample were observed in Zalambdalestes lechei, signifying that that species
possessed the most sensitive canals. Absolute size of canals might relate to overall body
size (larger species have larger canal radii; Spoor et al., 2007), but the average total
volume of the bony labyrinth of the two Zalambdalestes specimens (6.31 mm
3
) is nearly
identical to that of the Bissekty zhelestids (6.28 mm
3
). Large canals not only are
correlated to afferent sensitivity, but agile animals tend to have larger canals than slower
animals (Spoor et al., 2007).
Based on variable molar sizes in zhelestids (Nessov et al., 1998), body size is
thought to vary among zhelestid taxa. In order to eliminate size from the comparison, a
ratio was taken between the arc radius of each canal over the total labyrinth volume,
which may exhibit a correlation to body mass (preliminary unpublished data). The ratios
are presented in Table 4.7. If the size of the semicircular canal arc is related to locomotor
113
ability, and total volume of the bony labyrinth is correlated to body size, then a higher
ratio would signify a ‘fast’ animal under the system developed by Spoor et al. (2007).
The ratio of arc radius to inner ear volume indeed is larger for Zalambdalestes than for
zhelestids, suggesting that Zalambdalestes would have been a much more agile animal. In
fact, the highest ratios were observed in Ukhaatherium nessovi, suggesting that this
animal was the most agile, even though its arc radii were the smallest (Ukhaatherium
likely is the smallest Cretaceous eutherian included in the present study).
The vertical (anterior and posterior) semicircular canals of aquatic species of non-
mammalian amniotes tend to have lower canal aspect ratios than closely related terrestrial
species (Georgi and Sipla, 2008). Although the Bissekty zhelestids are not hypothesized
as aquatic animals, it is worth comparing the morphology in question with these potential
correlations. The aspect ratios of the vertical canals are similar among all of the
Cretaceous taxa considered in the present report in that none of the taxa have canals with
drastically smaller aspect ratios (Table 4.7). Although this result might suggest that all of
the Cretaceous species were terrestrial (or else all aquatic), the correlation observed by
Georgi and Sipla (2008) was found for closely related species. Among the Cretaceous
sample, Kulbeckia and Zalambdalestes share the closest ancestry, but neither species has
vertical canals with noticeably low aspect ratios.
The aspect of the cochlea in profile (height divided by width) plays a role in the
sense of hearing. A positive correlation between the ratio and both best frequency and
low frequency limit was recovered by Gosselin-Ildari (2006). In short, mammals with
planispiral cochleae have lower frequency limits. All of the Cretaceous labyrinths
possessed nearly planispiral cochleae, which suggests that the hearing ability was similar
among these taxa. A low aspect ratio likely is plesiomorphic for Eutheria, although a low
114
TABLE 4.7. Ratios between dimensions of semicircular canals. Volume is total volume
of bony labyrinth. Canal aspect ratio calculated as height of arc over width. Length is
calculated for the slender canal. Abbreviations: A, anterior semicircular canal; L, lateral
semicircular canal; P, posterior semicircular canal.
Taxon
Radius/Volume
Aspect Ratio
Radius/Length
A
L
P
A
L
P
A
L
P
Zhelestid
0.191
0.127
0.137
94.7
74.6
88.5
4.96
4.40
5.15
Kulbeckia kulbecke
0.221
0.169
0.183
102
96.5
102
4.80
4.29
4.75
Ukhaatherium nessovi
0.382
0.341
0.320
94.0
94.6
90.0
4.55
4.28
4.88
Zalambdalestes lechei
0.247
0.204
0.204
108
87.7
97.5
4.77
4.36
4.53
115
ratio might be a result of the low degree of coiling. Both Zalambdalestes and
Ukhaatherium have cochleae that complete only a single turn, which might result in a
smaller aspect ratio. However, a broader sampling of taxa is needed before a correlation
between the aspect ratio of the cochlea in profile and coiling of the cochlea can be
determined.
The ratio of the slender canal length over arc radius was computed by Boyer and
Georgi (2007) in order to investigate changes in frequency ranges transduced from head
movements in three Tertiary eutherians, Aphronorus orieli, Eoryctes melanus, and
Pantolestes longicaudus. The values that Boyer and Georgi calculated are similar to those
measured for the Cretaceous sample here (Table 4.7), although the significance of this
measure and any role that it plays in the locomotor behavior of mammals is yet to be
determined.
Phylogenetic Considerations
The number of turns completed by the cochlea is a character that commonly is
used in phylogenetic analyses incorporating Mesozoic mammals (e.g., Rougier et al.,
1998; Archibald et al., 2001). Among the Cretaceous fossils examined here, the cochlea
completes around one and a half turns (in zhelestids) or less (in the remaining taxa). This
value is consistent with those reported for other Mesozoic eutherians, including those
collected from the Bug Creek Anthills of Montana (Meng and Fox, 1995), which
complete one and a half turns, as well as the natural endocast of Zalambdalestes
described by Kielan-Jaworowska (1984), which completes a single coil. Other Cretaceous
eutherians for which the morphology of the inner ear is known include Prokennalestes
trofimovi (~360°; Wible et al., 2001), and Uchkudukodon nessovi (reassigned from
116
Daulestes nessovi by Archibald and Averianov, 2006), which completes slightly less than
360° (McKenna et al., 2000).
Interestingly, one labyrinth of Zalambdalestes fails to complete a single coil, as
well (left labyrinth of PSS MAE-130). Although the variation is not great (low CV
values) for coiling of the cochlea, variation is observed. If the average degree of coiling
for a species is 360° and the coiling varies, some individual cochleae will fall below the
average. Only a single cochlea of Uchkudukodon is described, but it can be hypothesized
that other cochleae of Uchkudukodon would complete more than 360°. However, this
hypothesis cannot be tested until additional petrosals are referred to Uchkudukodon.
Despite the variation, the degree of coiling observed within Cretaceous eutherians
is significantly less than that of in most extant eutherians. No single cochlea with fewer
than two turns was reported by West (1985), and cochleae with two or more coils were
observed in the majority of the placentals examined by Gray (1906, 1907, 1908). The
exceptions Gray noted mainly include lipotyphlans, a pipistrelle bat, a porpoise, and a
manatee, in all of which the cochlea coils between 1.5 and 1.75 turns.
A single turn in the cochlea is plesiomorphic for Eutheria (Meng and Fox, 1995),
and the data from the current study support this hypothesis. But coiling beyond one turn
likely developed more than once within the placental lineage. The phylogenetic affinities
of zhelestids are contentious, but the most recent and thorough phylogenetic analysis that
incorporated zhelestids (Wible et al., 2007) placed zhelestids near the bottom of the
eutherian tree. Given the topology recovered by Wible et al. (2007), a cochlea coiled to
540° and above would have evolved at least twice within Eutheria – once with the
Bissekty zhelestids, and a second time within Placentalia.
In addition to the coiling of the cochlea, the secondary common crus between the
lateral and posterior semicircular canals is a plesiomorphic structure for Eutheria (Meng
117
and Fox, 1995). The structure was observed in all of the Mesozoic taxa examined here, as
well as in the eutherians examined by Meng and Fox (1995). The secondary crus was
figured for Prokennalestes by Wible et al. (2001:fig. 2), although the structure is not
labeled in the figure, nor discussed in the text. A secondary common crus is not reported
for Uchkudukodon (McKenna et al., 2000), but the specimen is broken along the posterior
aspect of the lateral semicircular canal. Within Placentalia, a secondary common crus is
observed in several species of Canis and Felis (Hyrtl, 1845), Eumetopias jubatus, and
Orycteropus afer (personal observation), but is absent in most members of the crown.
The structure was lost, possibly multiple times, within Placentalia.
Whether or not the membranous lateral and posterior semicircular ducts
themselves were joined in the Cretaceous taxa is unclear. Joining of the ducts is not
uncommon within Mammalia. For example, the ducts fuse to form a membranous
secondary common crus in Trichosurus (Gray, 1908). However, the presence of an
osseous secondary common crus does not necessarily indicate that the membranous ducts
are joined, because the posterior and lateral ducts are separate in Canis (Gray, 1907;
Evans, 1993) despite the presence of a bony secondary common crus (Evans, 1993;
personal observation). A similar state was observed in the marsupial Caluromys
philander (Sánchez-Villagra and Schmelzle, 2007).
The measurements of the Kulbeckia specimen URBAC 98-113 are considerably
smaller than the other specimens referred to the same species (also discussed by Ekdale et
al., 2004). Two size morphs based on dentition were identified within Kulbeckia
(Archibald and Averianov, 2003). The difference in size, as hypothesized by Archibald
and Averianov, is the result of either sexual dimorphism or else a second, smaller species
of Kulbeckia. Changes through ontogeny is not a likely source of the observed
differential in size, because there is no correlation between most dimensions of the inner
118
ear and maturity of the individual once the bony labyrinth is ossified (for further
discussion, see Chapter 3). Alternatively, URBAC 98-113 might represent
Uchkudukodon. The size of the petrosal is consistent with that of Uchkudukodon
(McKenna et al., 2000), but the morphology is consistent with that of the other petrosals
referred to Kulbeckia (Ekdale et al., 2004). Specifically, Uchkudukodon possesses a
distinct groove for the transpromontorial artery on the tympanic surface of the petrosal
(McKenna et al., 2000, fig. 8). Such a sulcus is absent in URBAC 98-113, as well as the
petrosals of other zalambdalestids, including Zalmbdalestes (Wible et al., 2004, fig. 37).
CONCLUSIONS
The results of this study support the hypothesis of the ancestral morphology that
the bony labyrinth of eutherians includes a low degree of coiling (less than one and a half
turns, with the zhelestids having a higher degree of coiling of any identified Cretaceous
eutherian) and a secondary common crus. The results also indicate the generalities and
low degree of variation among Cretaceous eutherian taxa. Although the debate on
placental origins cannot be addressed, the data support a few functional interpretations of
the ear of Cretaceous eutherians. The cochleae of Cretaceous eutherians were not
specially derived for particularly high or low frequency hearing, as determined from the
structure of the cochlear spiral. Zalambdalestes and Ukhaatherium likely were more agile
animals than the Bissekty zhelestids, because they possessed relatively large semicircular
canal arcs for the size of their labyrinths.
Whether Placentalia originated within the Mesozoic or Cenozoic era, the inner ear
experienced a diversification of morphologies and physiologies after the end of the
Cretaceous, as reflected in the great diversity of ear anatomy and function among extant
119
placentals. Combination of the results detailed here with future research into the ear
regions of extinct eutherians, both inside and outside crown Placentalia, as well as from
both sides of the Cretaceous-Tertiary boundary, may clarify eutherian relationships, and
significantly increase our knowledge of the evolution of the senses of balance and
hearing of placental mammals.
120
CHAPTER 5: THE BONY LABYRINTH OF PLACENTAL
MAMMALS
ABSTRACT
The morphological diversity of the bony labyrinth of the inner ear (contained
within the petrosal bone) across a broad range of placental mammal taxa is documented,
and patterns of variation among the taxa are identified. Comparisons were made using
digital endocasts constructed using high-resolution X-ray computed tomography (CT)
imagery. The descriptions are organized taxonomically, covering the major placental
clades within Afrotheria, Xenarthra, Laurasiatheria, and Euarchontoglires, wherein linear,
angular, and volumetric dimensions were measured for the endocasts. The size of the
bony labyrinth is correlated to the overall body mass of individuals, where large bodied
mammals have absolutely longer and more voluminous inner ear cavities. The arc radius
of curvature averaged over the three semicircular canals is not correlated with agility as
has been hypothesized elsewhere, but the ratio between the average arc radius and body
mass of aquatic species is substantially lower than the ratios of related terrestrial taxa.
Additionally, the volume percentage of the vestibular apparatus of aquatic mammals
tends to be less than that calculated for terrestrial species. A significantly reduced
vestibule is unique to Cetacea among cetartiodactyls, and a cochlear spiral that is taller
than it is wide likely is a synapomorphy for caviomorph rodents, a low position of the
plane of the lateral semicircular canal compared to the posterior canal in Cetacea and
Carnivora, the lateral semicircular canal enters the vestibule at its posterior end in
Placentalia and into the posterior ampulla in Cetacea and Carnivora, and the cochlea has
121
a low aspect ratio in Primatomoprha. Thus, aspects of bony labyrinth morphology appear
to be phylogenetically informative.
INTRODUCTION
The vertebrate otic region, which functions in hearing via the cochlea as well as
balance and equilibrium via the vestibule and semicircular canals, is one of the most
intensively studied sensory systems. The external morphology of the petrosal bone,
which surrounds the delicate structures of the inner ear, is a common source of characters
used in phylogenetic analyses (e.g., Geisler and Luo, 1998; Rougier et al., 1998;
Archibald et al., 2001; Ladevèze, 2004; Wible et al., 2007). Because petrosals preserve
readily in the fossil record (see Archibald, 1979), the otic region is a valuable resource
for paleontologists when making biological inferences about extinct mammals (Fleischer,
1976; Spoor et al., 1994; Meng and Wyss, 1995; Witmer et al., 2003; Alonso et al., 2004;
Clarke, 2005).
The internal cavities within the petrosal comprise the bony labyrinth of the inner
ear, including the cochlea anteroventrally and the vestibular apparatus (with semicircular
canals) posterodorsally. The dimensions of inner ear structures are correlated to the
physiological capabilities of the otic region, including both hearing and balance. Ratios
between measurements of the cochlea are related to auditory frequency limits (Gosselin-
Ildari, 2006; Manoussaki et al., 2006, 2008; Gosselin-Ildari and Kirk, 2007; Kirk and
Gosselin-Ildari, 2009), which correlate with vocalization and social behavior, and the
dimensions of the semicircular canals relate to the sensitivity of the canals (Yang and
Hullar, 2007), which may in turn correlate to agility and locomotor behaviors (Spoor et
al., 2002, 2007; Georgi and Sipla, 2008).
122
The labyrinth of the inner ear is difficult to study because the inner ear structures
are completely surrounded by bone, and removal of this bone is necessary in order to
observe the inner ear cavities (Figure 5.1). The structures of the inner ear can be exposed
via dissolution of the surrounding bone (Gray, 1907, 1908; West, 1985) or through
serially sectioning the petrosal (Luo and Marsh, 1996; Novacek, 1986). Alternatively,
non-destructive techniques, such as high resolution X-ray computed tomography (CT),
can be used to digitally image the internal cavities of the petrosal bone (Georgi and Sipla,
2008; Spoor and Zonneveld, 1995; Witmer et al., 2003).
Morphology of the inner ear is phylogenetically informative at both more- and
less-inclusive taxonomic levels. For example, the cochlea completes at least one
complete 360° turn in living therian mammals, but less in monotremes (Gray, 1908;
Rowe, 1988). The bony labyrinth of Mesozoic therians exhibit ancestral morphologies,
such as a fusion of the posterior and lateral semicircular canals to form a secondary
common crus, which is lost in several clades within crown Theria (Meng and Fox, 1995;
Chapter 4). Within Primates, dimensions of the semicircular canals differ between the
great apes and other primates (Spoor and Zonneveld, 1998). Further phylogenetic
information can be found in the cochleae of squamate reptiles (Shute and Bellairs, 1953;
Schmidt, 1964; Miller, 1966a, b, 1968).
Given the functional and phylogenetic importance of this region of the skull, it is
surprising that broad comparisons of the inner ear of mammals are lacking (the most
notable exceptions are the works of Gray, 1906, 1907, and 1908). Furthermore, most
authors, owing to the functional division between the cochlea and vestibular apparatus,
decouple the structural continuity within the labyrinth. Functional studies therefore are
restricted either to the cochlea and the sense of hearing (Fleisher, 1976; Ramprashad et
al., 1979; West, 1985; Manley, 2000; Manoussaki et al., 2008), or to the vestibular
123
FIGURE 5.1. Petrosal of Dasypus novemcinctus (TMM M-1885) within which sits
endocast of bony labyrinth. A, tympanic view of petrosal bone; B, bone rendered semi-
transparent to reveal bony labyrinth; C, endocast of bony labyrinth. Abbreviations: am,
ampulla; ant, anterior direction (approximate); co, cochlea; cr, common crus; fc, fenestra
cochleae; fv, fenestra vestibuli; med, medial direction (approximate); pr, promontorium;
sc, semicircular canal; vb, vestibule.
A B C
pr
fv
fc
pr
co
vb
sc
co
vb
sc
1 mm1 mm
cr
ant
med
am
fv
fc
fc
fv
124
apparatus and the sense of balance (Jones and Spells, 1963; Blanks et al., 1975; Hullar
and Williams, 2006; Spoor et al., 2007). Rarely is the labyrinth considered as a whole and
compared across a large number of species. Such a comparison for the bony labyrinth of
placental mammals is provided here, along with potential functional and phylogenetic
considerations.
Systematic Context
As a point of departure for comparison, the bony labyrinth of a marsupial,
Didelphis virginiana, is described. The opossum is considered in many respects to retain
ancestral morphologies for Theria (Gaudin and Biewener, 1992; Vaughn et al., 2000;
Nilsson et al., 2003; Beck et al., 2008; however, see Clemens, 1968), and didelphids hold
a basal position on the marsupial phylogeny (Springer et al., l998; Bininda-Emonds et al,
2007). Moreover, Didelphis commonly is used as a marsupial outgroup in phylogenetic
analyses investigating placental relationships (e.g., Geisler and Luo, 1998; O’Leary and
Uhen, 1999; Amrine-Madsen et al., 2003; Asher, 2007). Whereas certain aspects of the
cranial morphology of the opossum are apomorphic (e.g., reduced pterygoids),
comparisons of the bony labyrinth suggest the otic morphology largely is plesiomorphic
(see Wible, 2003).
From Didelphis, descriptions of the labyrinths of eutherians (which includes
crown Placentalia and all extinct therians more closely related to Placentalia than its
extant sister taxon, Marsupialia) are arranged taxonomically following the relationships
recovered for Mesozoic non-placental eutherians by Wible et al. (2007), and the
relationships recovered for extant placentals by Bininda-Emonds et al. (2007).
125
A composite tree following the results of both studies is illustrated in Figure 5.2.
The relationships proposed by Wible et al. (2007) and Bininda-Emonds et al. (2007) were
based on an extensive taxonomic sampling, and therefore are used here. The relationships
recovered by Bininda-Emonds et al. (2007) were presented as a supertree (see Sanderson
et al., 1998) of 2622 published source tree topologies that included 4,510 out of 4,554
(99%) of mammal species recognized. The resulting supertree is the most comprehensive
phylogeny of extant mammals produced thus far. Likewise, the phylogeny produced by
Wible et al. (2007) is the most recent study to include the largest combination of
eutherian taxa both within and outside of crown Placentalia.
The descriptions of the bony labyrinths of crown placental mammals begin with
Afrotheria, and follow in order with Xenarthra, Laurasiatheria, and Euarchontoglires
(Figure 5.2). The descriptions are organized taxonomically within these major divisions
to allow the reader skip ahead to the account of the species in which he or she is
interested (see Table 5.1 for a list of species examined).
MATERIALS AND METHODS
Specimens
At least one representative of the major clades of placental mammals recovered
by the phylogenetic analyses of Bininda-Emonds et al. (2007) was selected (Figure 5.2)
based on availability of specimens at the Texas Natural Science Center Recent mammal
collection at the University of Texas at Austin, as well as available CT imagery from
“Digital Morphology: a National Science Foundation Digital Library at The University of
Texas at Austin” (www.digimorph.org). The specimens used in this study are listed in
Table 5.1, along with institutional abbreviations. Anatomical terminology follows Sisson
126
FIGURE 5.2. Cladogram of Theria including taxa considered. Relationships follow
Bininda-Emonds et al. (2007) and Wible et al. (2007). Nodes: A, Theria; B, Eutheria; C,
Ukhaatherium+Zalambdalestidae+Placentlia; D, Zalambdalesidae+Placentalia; E,
Zalambdalestidae; F, Placentalia; G, Afrotheria; H, Afrosoricida+Macroscelides; I,
Afrosoricida; J, Paenungulata; K, Procavia+Trichechus; L, Boreoeutheria; M,
Laurasiatheria; N, Ungulata+Fera+Chiroptera; O, Ungulata; P, Cetartiodactyla; Q,
Sus+Cetacea; R, Cetacea; S, Ferae; T, Carnivora; U, Caniformia; V, Chiroptera; W,
Microchiroptera; X, Rhinolophus+Tadarida; Y, Eulipotyphla; Z, Euarchonoglires; a,
Glires; b, Rodentia; c, Lagomorpha; d, Primatomorpha; e, Primates.
127
Didelphis
Zhelestid
Ukhaatherium
Kulbeckia
Zalambdalestes
Chrysochloris
Hemicentetes
Macroscelides
Orycteropus
Procavia
Trichechus
Mammuthus
Dasypus
Bathygenys
Sus
Balaenopterid
Tursiops
Equus
Canis
Eumetopias
Felis
Manis
Pteropus
Nycteris
Rhinolophus
Tadarida
Atelerix
Sorex
Cavia
Mus
Lepus
Sylvilagus
Homo
Macaca
Cynocephalus
Tupaia
A
B
C
D
F M
P
O
R
U
V
W
X
Y
Z
a
b
c
d
e
T
S
L
N
Q
H
I
G
K
J
E
128
TABLE 5.1. Taxa examined and scanning parameters
a
Taxon
b
Slices
Space
FR
Pixel
Size
Marsupialia
Didelphis virginiana (TMM M-25527)
111
0.132
61
0.0596
1024
Eutheria
Kulbeckia kulbecke (URBAC 04-36)
387
0.016
14.9
0.0146
1024
Ukhaatherium nessovi (PSS-MAE 110)
59
0.080
15
0.0290
512
Zalambdalestes lechei (PSS-MAE 108)
66
0.113
24.5
0.0479
512
Zhelestid (URBAC 03-39)
536
0.016
14.9
0.0146
1024
Afrotheria
Afrosoricida
Chrysochloris sp. (AMNH 82372)
85
0.050
31
0.0303
1024
Hemicentetes semispinosum (AMNH 100837)
56
0.067
48
0.0469
1024
Macroscelidea
Macroscelides proboscideus (AMNH 161535)
151
0.055
22.5
0.0220
1024
Tubulidentata
Orycteropus afer (AMNH 51909)
136
0.202
95
0.0930
1024
Hyracoidea
Procavia capensis (TMM M-4351)
180
0.799
70
0.0683
1024
Sirenia
Trichechus manatus (MSW 03156)
c
229
0.300
80
0.1563
512
Proboscidea
Elephantoidea (TMM 933-950)
275
0.134
53
0.0518
1024
Xenarthra
Dasypus novemcinctus (TMM M-152)
494
0.291
25
0.0244
1024
Laurasiatheria
Ceatartiodactyla
Bathygenys reevesi (TMM 40209-198)
149
0.141
64
0.0625
1024
Sus scrofa (TMM M-2689)
601
0.033
31
0.0303
1024
Balaenopteridae (TMM 42958-35)
1131
0.072
64.8
0.0633
1024
Tursiops truncatus (SDSNH 21212)
346
0.128
40
0.0391
1024
Perissodactyla
Equus caballus (TMM M-171)
645
0.115
54
0.0567
1024
Carnivora
Canis familiaris (TMM M-150)
92
0.144
68
0.0664
1024
Eumetopias jubatus (TMM M-171)
645
0.115
54
0.0527
1024
Felis catus (TMM M-968)
627
0.033
31
0.0303
1024
Pholidota
Manis tricuspis (AMNH 53896)
101
0.116
35.5
0.0347
1024
Chiroptera
Pteropus lyelli (AMNH 237593)
188
0.447
41
0.0400
1024
129
TABLE 5.1. (Continued)
Nycteris grandis (AMNH 268369)
70
0.072
67
0.0654
1024
Rhinolophus ferrumequinum (AMNH
245591)
45
0.097
44
0.0430
1024
Tadarida brasiliensis (TMM M-3030)
380
0.010
9.9
0.0097
1024
Eulipotyphla
Atelerix albiventris (uncatalogued)
68
0.082
65
0.0635
1024
Sorex monticolus (uncatalogued)
d
130
0.265
12
0.0117
1024
Euarchontoglires
Rodentia
Cavia porcellus (TMM M-7283)
728
0.038
29.4
0.0287
1024
Mus musculus (TMM M-3196)
84
0.044
12.6
0.0246
512
Lagomorpha
Lepus californicus (TMM M-7500)
114
0.144
67
0.0654
1024
Sylvilagus floridanus (TMM M-2689)
325
0.034
30
0.0293
1024
Primates
Maraca emulate (TMM M-5987)
1121
0.033
31
0.0303
1024
Homo sapiens (UT Teach Coll)
636
0.027
24.84
0.0272
1024
Dermoptera
Cynocephalus volans
350
0.028
22
0.0215
1024
Scandentia
Topeka glass (TMM M-2256)
537
0.038
59
0.0576
1024
a
Definitions of parameters are as follows: FR, field of reconstruction refers to the
dimensions of an individual CT slice, expressed in millimeters; Pixel, interpixel spacing,
or vertical and horizontal dimensions of an individual pixel, expressed in millimeters, and
calculated as FR/Size; Size, number of pixels in a CT slice, either 512X512 or
1024X1024 pixels; Slices, number of CT slices through the ear collected in the coronal
(original) slice plane; Space, interslice spacing, or distance between consecutive slices,
expressed in millimeters.
b
Taxonomy and systematic arrangement follows Bininda-Emonds et al. (2007) and
Wible et al. (2007). Institutional abbreviations: AMNH, American Museum of Natural
History, New York; MSW, Mortality South West; PSS-MAE, Collections of Joint
Paleontological and Stratigraphic Section of the Geological Institute, Mongolian
Academy of Science, Ulaanbaatar – American Museum of Natural History, New York;
SDSNH, San Diego Society of Natural History, San Diego, CA; TMM, Texas Natural
Science Center, Austin, TX; URBAC, Uzbekistan/ Russian/ British/ American/ Canadian
joint paleontological expedition, Kyzylkum Desert, Uzbekistan, specimens in the Institute
of Zoology, Tashkent.
c
This specimen was the 156
th
Mortality South West in 2003, collected by S. Rommel at
University of North Carolina Wilmington.
d
Specimen information at http://digimorph.org/specimens/Sorex_monticolus/whole/
130
and Grossman (1938) and Evans (1993), and orientation terminology of the cochlea
follows Fleischer (1976) as depicted in Figure 5.3A-B.
Whenever possible, isolated petrosal bones were scanned to maximize the
resolution of the CT imagery (CT methods described below). The left petrosal was
examined for each taxon, with a few exceptions, for consistency. Although cranial
asymmetry is known within the ear region (Caix and Outrequin, 1979), the physiological
significance of such asymmetry is poorly understood. Images of the bony labyrinth are
reversed in figures in the cases where right petrosals were used instead of elements from
the left side of the skull for easy visual comparison.
All specimens were presumed mature, although no rigorous assessment of
maturity was done. Although the external surface of the petrosal changes through
accretionary growth, there is evidence that the structures of the inner ear do not change
significantly once the walls of the bony labyrinth have ossified (Hoyte, 1961; Chapter 3).
Based on those studies, the maturity of individuals used in this study should not affect the
observed morphology. Because post-ossification changes in the bony labyrinth only have
been investigated for the rabbit (Hoyte, 1961) and the marsupial Monodelphis domestica
(Chapter 3), but not all of Mammalia, the overall consistency of this pattern among all
therian mammal species cannot be assessed. Such a survey is beyond the scope of the
current study, and it is assumed that any variation in the mature and fully ossified bony
labyrinths used in the following comparisons does not affect the observation of characters
at the gross morphological level.
131
FIGURE 5.3. Measurement methods employed. A, coiling of cochlea; B, height and
width of cochlea used for calculation of aspect ratio; C, height, width, and length of
semicircular canal; D, arc radius of curvature calculated from height and width of arc; E,
linear deviation of semicircular canal, used with arc radius of curvature to calculate
angular deviation. Abbreviations: am, ampulla; ar, semicircular canal arc radius of
curvature; cl, length of cochlear canal; cp, plane of semicircular canal; fc, fenestra
cochleae; ht, height; ld, linear deviation of semicircular canal from its plane; rl, reference
line for measuring coiling of cochlea; sc, semicircular canal; scl, slender semicircular
canal length; sl, secondary bony lamina; vb, vestibule; wt, width.
A B
C D
E
cl
fc
ht
fc
sl
wt
ht
wt
scl
am
vb
ar
ld 1
cp
sc
scl
am
ld 2
rl
scl
132
Computed Tomography Methods
Digital imagery obtained through computed tomography (CT) was employed to
observe the internal chambers of the petrosal that constitute the bony labyrinth. The
majority of the specimens used for this study were scanned at the University of Texas
High-resolution X-ray CT facility (UTCT) in Austin, TX. The only specimen not scanned
at UTCT was Trichechus manatus (MSW 03156), which was scanned at Washington
University in Saint Louis, MO. Parameters for CT scanning and post-scanning image
processing are provided in Table 5.1.
Resolution of the CT data was not uniform among the datasets, in terms of the
number of voxels per slice (512 X 512 versus 1024 X 1024), interslice spacing, field of
reconstruction (which affects the size of individual voxels), as well as the number of
slices through the bony labyrinth (listed in Table 5.1). Differences in data resolution not
only affect the appearance of the bony labyrinth (see digital endocasts in Figures 5.4-
5.69), but minute structures may not appear on low-resolution datasets (see the
anatomical descriptions of the afrosoricid Hemicentetes and the carnivoran Canis
discussed below). Canals with diameters that are less than the size of a pixel may not be
observed, and there is a chance that the CT scanner will miss a structure smaller than the
interslice spacing. Furthermore, differences in dataset resolution can affect the values of
measurements in a significant manner (see Gosman and Ketcham, 2009). The absolute
number of slices through the ear region (ranging from 45 for the bat Rhinolophus
ferrumequinum to 1131 for a fossil balaenopterid whale) seems to have the most dramatic
effect on the data presented here.
The bony labyrinths were digitally segmented from the CT imagery into the
various partitions of the inner ear (e.g., cochlea and vestibule) in order to calculate partial
133
volumes of the osseous cavities, as well as create a 3-dimensional representation of the
osseous inner ear cavities. Segmentation was performed in the computer software
packages VGStudio Max 1.2
©
(Volume Graphics, 2004) and Amira 3.1
©
(Mercury
Computer Systems, 2003). The process of segmentation was achieved by isolating
individual voxels that represent a particular anatomical feature of interest on 2-
dimensional CT slices (in this case, the spaces within the petrosal). A series of individual
CT slices were “stacked”, and the isolated voxels were extracted from the surrounding
bone to create a 3-dimensional endocast of the segmented structure. The bony channels
for the vestibular and cochlear aqueducts were included in the segments of the cochlea
and vestibule respectively. The canals for branches of cranial nerve VIII were not
segmented. The boundaries between the components were kept as planar as possible. The
medial border of the fenestra vestibuli was used as the dividing line between the cochlea
and vestibule, where the entire fenestra is included within the segmented vestibule.
Thresholding during segmentation (determining the air to bone boundary) was
accomplished visually, modified from the half-maximum height protocol (HMH) of
Coleman and Colbert (2007).
Standard orientation of the bony labyrinth as a whole is not trivial or
straightforward. The endocasts constructed for this study are oriented with the plane of
the lateral semicircular canal parallel to the horizon. Such an orientation was selected
because the lateral semicircular canal usually is held horizontal when the animal is in a
state of alertness (de Beer, 1947). Although the lateral canal is not parallel to the earth-
horizon at all times in every animal (Hullar, 2006), a standard position is used in the
present study for comparative purposes. Anterior view is oriented down a line connecting
the ampullar aperture of the lateral semicircular canal and the vestibular aperture of the
posterior limb of the lateral canal (or vestibular aperture of the posterior ampulla if the
134
canal does not open directly into the vestibule at its posterior end). The labyrinth is
oriented with respect to this anterior position in all other views.
Measurement Methods
Methodologies for various measurements follow Fleisher (1976), Spoor and
Zonneveld (1995), Geisler and Luo (1996), and Jeffery and Spoor (2004). Angular,
linear, and volumetric measurements were made in the Amira software. The
measurements made on the digital endocasts, as well as ratios calculated from the
measurements, have been used to interpret the physiology and evolutionary relationships
in a wide variety of vertebrate species. Many of the measurements are depicted
graphically in Figure 5.3.
The total volume of the bony labyrinth is calculated here as a measure of overall
size of the inner ear. Volumes are calculated for individual compartments within the
internal cavities, including the cochlea, vestibule, and semicircular canals. In addition, the
linear length of the bony labyrinth was measured as the greatest distance from the center
of the lumen at the posteriormost point of the posterior semicircular canal to the center of
the lumen at the anteriormost point along the basal turn of the cochlea (Jeffery and Spoor,
2004).
Dimensions of the cochlea include the total degrees of coiling completed by the
canal, as well as the length of the cochlear spiral and widths of internal cochlear
structures. In order to calculate the degree of coiling exhibited by the cochlea (following
modified methods of West, 1985, and Geisler and Luo, 1996), the endocast of the
cochlear canal is positioned in vestibular (“map”) view, down the axis of rotation (Figure
5.3A). A line is drawn from the axis of rotation to a point at the base of the cochlea,
between the fenestrae cochleae and vestibuli (round and oval windows respectively). The
135
number of times that the cochlea crosses the line are counted and multiplied by 180° in
order to calculate a gross value of cochlear coiling. Added to this product is an additional
angle measured between the line drawn through the cochlea and a second line drawn
from the axis of rotation to the apex (most distal point) of the cochlear canal. The total
measurement of coiling of the cochlea is expressed both in degrees as well as total
number of turns (calculated as total degrees divided by 360°).
A shape index (aspect ratio) of the cochlear spiral was calculated by dividing the
height of the spiral by the width of the basal turn (Gosselin-Ildari, 2006). The width of
the basal turn (which has the largest diameter of all turns in the cochlea) is measured as
the greatest distance from the vestibular edge (closest to the apex of the spiral) of the
fenestra cochleae to the radial (outside) wall on the opposite side of the basal turn of the
cochlea, parallel to the plane of the basal turn of the cochlea (Figure 5.3B). The height of
the cochlea is measured as the greatest distance from the level of the most tympanal edge
of the fenestra cochleae to the vestibular-most wall of the cochlear spiral (within the
apical turn), perpendicular to the width. The aspect ratio of the cochlea in profile may
correlate with frequency limits (Gosselin-Ildari, 2006). A high aspect ratio is considered
to be above 0.55, following observations of “flattened” and “sharp-pointed” cochleae by
Gray (1907, 1908), where “flattened” cochleae have an aspect ratio 0.55 and below.
The total length of the cochlear canal from base to apex was measured using the
SplineProbe tool in the Amira software. In order to measure this, a curved line is fit to
points within the lumen of the cochlea at the center of the laminar gap (the space between
the primary and secondary bony spiral laminae that support the basilar membrane and
spiral organ of hearing in life). The endpoints of the line are at the base of the laminar
gap (where the primary and secondary lamina converge) and the apex (most distal point)
of the canal. The length of the cochlear canal approximates the length of the soft-tissue
136
basilar membrane, upon which the spiral organ of hearing sits, which may correlate to
audible frequencies (West, 1985).
The orientation of the cochlea with respect to the vestibule is quantified as the
angle between the planes of the basal turn of the cochlea and the lateral semicircular
canal. Often it is assumed that an alert animal will hold its head so that the lateral
semicircular canal is held parallel to the horizon (de Beer, 1947), although such is not
always the case with all species (see review provided by Hullar, 2006). The plane of the
basal turn of the cochlea intersects points at the center of the lumen of the basal turn, and
the plane of the lateral canal fits to points at the center of the lumen at the two ends of the
canal, as well as the midpoint of the arc. The angle is measured while the bony labyrinth
is oriented in the Amira software so that the intersection of the two planes is
perpendicular to the field of view.
Dimensions of the semicircular canals include the angles between the planes of
the canals, the angular deviation of each canal from its average plane, the arc radius of
curvature of each canal, diameter of the canal lumen, and the length and volume of the
slender canal (unampullated course). The angle between two semicircular canals is
measured when the intersection of the planes is perpendicular to the field of view. The
planes of the canals intersect a point at the midpoint of the canal arc, as well as the entry
of the canal into its respective ampulla at one end and the vestibule at the other.
The lengths of the slender semicircular canals are measured using the SplineProbe
tool in the Amira software, similar to the method used for measuring the length of the
cochlear canal, where the curved line intersects points at the center of the lumen across
the slender course of the canal. The slender canal is defined as the course of the
semicircular canal that does not include the ampulla (Boyer and Georgi, 2007). The
common crus is included with the slender paths of both the anterior and posterior
137
semicircular canals, and the posterior ampulla is included in the path of the lateral
semicircular canal when the posterior limb of the lateral semicircular canal does not enter
the vestibule directly.
The radius of curvature of a semicircular canal (the dimension “R” of Jones and
Spells, 1963, and Spoor and Zonneveld, 1995) is calculated as half the average between
the height and width of the canal arc (Figures 5.3C and 5.3D). The specific methods to
measure the height and width of the semicircular canal arc are adapted from Spoor and
Zonneveld (1995), and the measurements are taken on the digital endocasts in the Amira
software using the line measuring tool with the 3-dimensional capability enabled. The
height of the lateral semicircular canal is measured as the greatest distance from the wall
of the vestibule to the center of the lumen of the canal. The height of the anterior and
width of the posterior semicircular canals are measured when the lateral canal is viewed
parallel to the horizon, where the height of the anterior semicircular canal arc is measured
as the greatest distance from the wall of the vestibule to the center of the lumen of the
canal, and the width of the posterior canal is measured as the greatest distance between
the center of the lumen of the posterior limb of the canal and the center of the lumen of
the common crus.
The total angular deviation of a semicircular canal from its respective plane is
calculated trigonometrically using two linear measurements of the canal (adapted from
Calabrese and Hullar, 2006, and Hullar and Williams, 2006). The linear measures utilized
in the calculation of angular deviation are the arc radius of curvature of the canal and
total linear deviation of the canal from its plane (Figure 5.3E). The total linear deviation
is measured when the plane of the semicircular canal is perpendicular to the field of view
in the Amira software using the line measurement tool with the 2-dimensional capability
enabled. The greatest distance from the center of the lumen to the plane of the canal is
138
measured on both sides of the plane. Partial angular deviations are calculated as the
arcsine of the partial linear deviation divided by the arc radius of curvature. The total
angular deviation is the sum of the two partial angular deviations.
Substantial deviation of a semicircular canal from its plane is defined here as any
ratio of the total linear deviation over the cross-sectional diameter at the midpoint of the
semicircular canal being greater than 1. In short, if the total linear deviation of a canal
from its plane is greater than the diameter of the canal in cross-section, the deviation is
considered significant, in a non-statistical sense. The measure of significant deviation is
arbitrary and does not have any basis in the physiology of the semicircular canal system.
The functional implications of non-planar semicircular canals are poorly understood, and
the significance values are intended to emphasize the phenomenon of non-planarity,
rather than to make any functional interpretations at this time.
The sagittal labyrinthine index, which is defined as the percentage of the width of
the posterior semicircular canal arc below the plane of the lateral semicircular canal
(Spoor and Zonneveld, 1995), quantifies the relative positions of the lateral and posterior
semicircular canals. High sagittal labyrinthine indices separate the great apes from other
primates (Spoor and Zonneveld, 1998), and the index might be useful in the phylogenetic
assessment of other mammal groups.
Two additional indices are ratios that might relate to aquatic habitat and
locomotion. Specifically, the ratio of the slender canal length over arc radius (Boyer and
Georgi, 2007), as well as the aspect ratio (height over width) of the arcs of the vertical
(anterior and posterior) semicircular canals (Georgi and Sipla, 2008) distinguish aquatic
species from their terrestrial ancestors. These ratios were calculated for the present
sample.
139
The stapedial ratio describes the shape of the footplate of the stapes (Segall,
1970), which contacts the inner ear spaces via the fenestra vestibuli. Marsupial mammals
tend towards circular fenestrae with ratios below 1.8, whereas the fenestrae of placentals
are more elliptical. In absence of the stapedial footplate, the dimensions of the fenestra
vestibuli can be used. Although the footplate of the stapes is held in place by an annular
ligament, the stapes articulates tightly with the fenestra (personal observation). The
fenestra vestibuli is expressed on the digital endocasts, and the stapedial ratio is
calculated as the greatest distance along the long axis of the fenestra, divided by the
greatest distance measured along the short axis, perpendicular to the first measurement.
To ascertain whether the dimensions of the inner ear are correlated to overall
body size of the animal, specific measurements were plotted over body mass (all data
logarithmically transformed using the natural log) and the coefficient of correlation (“r”)
was calculated. Any coefficient greater than or equal to 0.7 is considered significant. If
the body mass of the specimen examined was not known, an average was calculated from
Silva and Downing (1995) for most non-human taxa (data for Eumetopias jubatus from
Loughlin et al., 1987), and values from Ogden et al. (2004) were used for Homo sapiens.
Body mass data are unavailable for the two fossils (Bathygenys reevesi and
Balaenopteridae). Further, a body mass was not used for Canis familiaris given the broad
range of body masses observed in domestic dogs (Galvão, 1947; Heusner, 1991). Further
correlations were investigated between dimensions of the cochlea, as well as dimensions
of the semicircular canals.
Ancestral states, both continuous and discrete, for the hypothetical common
ancestors of the clades pictured in Figure 5.2 were reconstructed in the computer program
Mesquite (Maddison and Maddison, 2008). Although discrete character states of
ancestors were reconstructed using the parsimony method in the Mesquite software, the
140
maximum likelihood method was utilized for continuous characters (following Martins,
1999). Characters that were traced across the cladogram are (1) entry of the lateral
semicircular canal into a secondary common crus, the posterior ampulla, or the vestibule,
(2) largest semicircular canal arc among the anterior, lateral, and posterior canals, (3)
aspect ratio of the cochlear spiral in profile, (4) degree of cochlear coiling, and (5)
contribution (percentage) of the volume of the cochlea to the entire labyrinth.
RESULTS –ANATOMICAL COMPARISONS
The osseous cavities that constitute the bony labyrinth of the inner ear contain a
series of membranous sacs and ducts that embody the membranous labyrinth. The
membranous labyrinth consists of the cochlear duct and saccule of the vestibule within a
ventral division of the labyrinth, and the utricle of the vestibule, along with the three
semicircular ducts with their respective ampullae and the common crus within a dorsal
division.
Theria
Theria includes the most recent common ancestor of extant marsupials (such as
Didelphis virginiana, which is used to represent Marsupialia) and extant placentals (such
as Homo sapiens) and all of the descendents of that ancestor. The bony labyrinth of the
hypothetical therian ancestor possessed a secondary common crus formed between the
lateral and posterior semicircular canals (see the labyrinth of Didelphis in Figure 5.4-5.5),
which likely was inherited from a much more distant mammalian ancestor (Ruf et al.,
2009).
The plane of the lateral canal is positioned low with respect to the ampullar
entrance of the posterior canal so that the area of the arc of the posterior canal is not
141
FIGURE 5.4. Bony labyrinth of Didelphis virginiana. A, stereopair and labeled line
drawing of digital endocast in anterior view; B, stereopair and labeled line drawing of
digital endocast in dorsal view; C, stereopair and labeled line drawing of digital endocast
in lateral view; D, line drawing of cochlea viewed down axis of rotation to display degree
of coiling; E, line drawing of cochlea in profile; F, vestibular apparatus displaying
secondary common crus. Abbreviations: aa, anterior ampulla; ac, anterior semicircular
canal; ant, anterior direction; av, bony channel for aqueduct of vestibule; cc, canaliculus
cochleae for aqueduct of cochlea; co, cochlea; cr, common crus; dor, dorsal direction; er,
elliptical recess of vestibule; fc, fenestra cochleae; fv, fenestra vestibuli; la, lateral
ampulla; lc, lateral semicircular canal; med, medial direction; pa, posterior ampulla; pc,
posterior semicircular canal; pl, primary bony lamina; pos, posterior direction; ps,
outpocketing for perilymphatic sac; scr, secondary common crus; sl, secondary bony
lamina; sr, spherical recess of vestibule; vb, vestibule.
142
co
fc
ps
fv
sr
er
aa
la
pa
cr
1 mm
1 mm
1 mm
ac
lc
pc
co
cr
aa
la
ac
er
lc
pa
scr
scr
pc
cc
av
pl
co
fc
sl
cc
av
pc
ac
cr
aa
la
fv
lc
scr
pls
sr
co
fc
cc
co
sl
fc
pls
cc
cr
scr
pa
la
aa
lc
ac
pc
vb
A
B
C
D E F
dor
med
pos
med
dor
ant
143
FIGURE 5.5. Original CT slices through ear region of Didelphis virginiana. Numbers
refer to specific CT slices. Abbreviations: aa, anterior ampulla; ac, anterior semicircular
canal; av, bony channel for aqueduct of vestibule; cc, canaliculus cochleae; cn, canal for
cranial nerve VIII; co, cochlea; cr, common crus; dor, dorsal direction; fc, fenestra
cochleae; fn, canal for cranial nerve VII; fv, fenestra vestibuli; la, lateral ampulla; lat,
lateral direction; lc, lateral semicircular canal; pa, posterior ampulla; pc, posterior
semicircular canal; pl, primary bony lamina; pos, posterior direction; pr, promontorium
housing cochlea; scr, secondary common crus; sl, secondary bony lamina; sr, spherical
recess of vestibule.
144
pr
dor
lat
co
pos
lat
38
42
46
50
54
58
62
66
70
78
74
1 mm
cn
vb
pl
1 mm
fn
co
co
fn
38 42
5046 54
6258 66
7470 78
pl
fv
fc
fv
pl
fn
cn
cn
sl
cn
cn
sr
av
av
av
ac
sa
cr
cr
cr
ac
lc
lc
ac
ac
ac
er
scr
pc
pc
pc
pc
pa
aa
ac
lc
la
cc
145
divided by the lateral canal in anterior view, as it is in most extant placentals (e.g.,
Chrysochloris or Sylvilagus). As observed in most mammal species, the arc of the
anterior semicircular canal is the largest among the three canals. The cochlea completed a
685° coil (nearly two turns) and contributed 66% of the total inner ear volume. The
aspect ratio of the cochlea of the ancestral therian is reconstructed as low, although the
aspect ratio in Didelphis (0.62) is higher than that calculated for basal taxa along the
eutherian lineage. The ancestor of Theria likely possessed a cochlea with a low aspect
ratio given the close similarities between basal metatherian and eutherian labyrinths (see
Meng and Wyss, 1995), and the ancestral state is reconstructed as such.
Marsupialia
The structure of the inner ear of Didelphis virginiana is described for comparison
with the inner ear structures of crown placentals and their Mesozoic eutherian relatives.
Dimensions of the bony labyrinth as a whole of Didelphis (and all other taxa) are
provided in Table 5.2. Dimensions of the cochlea are provided in Table 5.3, and
dimensions and orientations of the semicircular canals are reported in Tables 5.4-5.6.
Didelphis is a common animal in North America, despite it being the only North
American marsupial. The body mass of the specimen used (TMM M-2517) is 2.8 kg (see
Table 5.2), which is on the higher end of the mass range of the species (1.6-3.1 kg; Silva
and Downing, 1995). The bony labyrinth of D. virginiana is 5.15 mm in total length, and
the cochlea contributes 68.7% of the total volume of the inner ear (8.30 out of 12.1 mm
3
),
which is close to that calculated for the ancestral therian (66.0%). The cochlear spiral is
high in profile, with an aspect ratio of 0.62 (height equals 2.28 mm and width equals 3.68
mm). The cochlea completes nearly two and a quarter turns (790.7°), and the length of
146
TABLE 5.2. Body mass, skull length, and dimensions of the entire bony labyrinth of
placentals
a
Taxon
b
Body
Skull
Bony Labyrinth
Mass
Length
Volume
Length
Marsupialia
Didelphis
2800.00
106.65
12.08
5.15
Eutheria
Kulbeckia
NA
NA
5.37
4.73
Ukhaatherium
NA
NA
2.17
3.57
Zalambdalestes
NA
NA
6.07
5.31
Zhelestid
NA
NA
6.28
4.51
Afrotheria
Chrysochloris
44.40
23.85
4.11
3.93
Elephantoidea
NA
NA
1145.18
26.00
Hemicentetes
110.00
30.61
2.78
4.08
Macroscelides
38.40
32.26
9.19
4.31
Orycteropus
60000.00
245.48
107.32
14.95
Procavia
3800.00
75.87
19.36
8.50
Trichechus
500000.00
NA
621.04
19.29
Xenarthra
Dasypus
4753.75
NA
26.48
8.06
Laurasiatheria
Atelerix
866.38
38.13
4.58
5.46
Balaenopteridae
NA
NA
1075.51
19.67
Bathygenys
NA
90.71
29.83
7.40
Canis
NA
87.10
31.36
8.10
Equus
258324.32
530.00
165.16
16.52
Eumetopias
735000.00
NA
138.60
13.71
Felis
3407.86
NA
45.78
8.91
Manis
4500.00
75.05
28.53
6.66
Nycteris
29.27
27.80
2.13
3.39
Pteropus
435.00
65.23
7.01
6.19
Rhinolophus
17.21
24.28
5.89
3.76
Sorex
6.07
16.90
0.81
2.81
Sus
88285.71
240.00
61.86
9.95
Tadarida
12.13
NA
3.86
3.22
Tursiops
179500.00
543.02
167.98
10.08
Euarchontoglires
Cavia
728.00
67.80
22.22
7.13
147
TABLE 5.2. (Continued)
Taxon
b
Body
Skull
Bony Labyrinth
Mass
Length
Volume
Length
Cynocephalus
1000.00
NA
20.32
7.17
Homo
8000.00
NA
164.73
16.31
Lepus
2350.00
94.37
24.26
7.39
Macaca
4667.50
NA
41.64
11.23
Mus
15.48
20.81
1.47
2.71
Sylvilagus
1160.00
68.70
11.32
5.82
Tupaia
131.15
49.60
9.83
6.67
a
Body mass expressed in grams; Skull and bony labyrinth length expressed in
millimeters; bony labyrinth volume expressed in cubic millimeters.
b
Specimens used listed in Table 5.1. Taxonomy follows Bininda-Emonds et al. (2007)
and Wible et al. (2007). Values for Kulbeckia, Zalambdalestes, and the Zhelestid are
averages (reported in Chapter 4).
148
TABLE 5.3. Dimensions and orientations of the cochlea of placentals
a
Taxon
b
Volume
Coiling
2° Lamina
Length
Aqueduct
Ratio
Angle
Marsupialia
Didelphis
8.30
790.7
427.0
7.54
1.68
0.62
19.6
Eutheria
Kulbeckia
2.59
446.0
208.7
4.93
0.60
0.44
12.1
Ukhaatherium
1.23
380.0
76.8
2.77
0.36
0.35
6.63
Zalambdalestes
2.91
368.0
95.3
3.40
0.48
0.36
13.5
Zhelestid
4.15
545.0
197.7
4.93
0.37
0.46
34.0
Afrotheria
Chrysochloris
2.93
1191.0
301.0
6.65
0.45
0.63
41.9
Elephantoidea
350.66
765.0
NA
32.45
NA
0.42
48.5
Hemicentetes
1.39
540.0
240.3
3.79
.28
.38
18.4
Macroscelides
6.59
720.0
334.0
7.11
0.58
0.80
25.1
Orycteropus
59.31
709.0
390.0
14.86
4.82
0.45
31.9
Procavia
9.24
1363.0
190.0
14.97
1.21
0.72
45.4
Trichechus
441.63
407.0
NA
22.46
NA
0.55
27.7
Xenarthra
Dasypus
17.48
816.3
383.0
11.21
1.17
0.63
17.9
Laurasiatheria
Atelerix
2.28
623.7
240.0
4.99
0.77
0.69
53.8
Balaenopteridae
973.91
886.0
238.0
53.02
3.65
0.48
23.2
Bathygenys
16.17
667.0
NA
8.51
NA
0.32
26.8
Canis
20.72
1156.0
104.0
13.85
2.08
0.64
20.8
Equus
84.33
911.3
153.0
22.08
11.33
0.41
37.9
Eumetopias
74.17
795.4
249.0
19.25
4.16
0.68
31.6
Felis
31.12
1092.0
243.0
16.77
3.60
0.69
45.8
Manis
14.00
863.0
NA
9.64
2.85
0.54
20.3
Nycteris
1.42
801.0
316.0
6.66
0.66
0.61
47.2
Pteropus
4.13
656.0
335.0
7.66
0.73
0.61
36.2
Rhinolophus
5.24
1115.0
935.0
11.57
0.59
0.63
5.5
Sorex
0.37
493.0
179.0
2.52
0.23
0.47
9.41
Sus
36.25
1340.0
NA
22.89
2.64
0.71
23.8
Tadarida
2.80
751.6
659.0
6.95
0.12
0.52
29.2
Tursiops
157.11
661.0
396.0
24.01
6.47
0.47
21.3
Euarchontoglires
Cavia
12.26
1457.0
195.0
13.42
2.52
1.29
35.1
149
TABLE 5.3. (Continued)
Taxon
b
Volume
Coiling
2° Lamina
Length
Aqueduct
Ratio
Angle
Cynocephalus
9.83
953.7
65.4
12.20
0.90
0.50
34.6
Homo
71.49
889.0
22.2
22.49
10.86
0.36
62.4
Lepus
13.07
693.0
147.0
8.80
1.34
0.64
40.6
Macaca
20.96
1088.0
81.0
16.94
3.53
0.48
47.8
Mus
0.86
627.7
327.0
3.87
0.17
0.62
10.8
Sylvilagus
6.26
816.8
200.0
8.75
1.05
0.71
40.3
Tupaia
5.43
1125.0
220.0
10.51
0.66
0.66
28.9
a
Measurement methodologies provided in text: Volume, total volume of cochlear canal,
expressed in cubic millimeters; Coiling, the total degrees completed by the cochlea; 2°
Lamina, extension of secondary lamina through cochlea, expressed in degrees; Length;
length of canal, expressed in millimeters; Aqueduct, length of cochlear aqueduct,
expressed in millimeters; Ratio, aspect ratio calculated as height of spiral over width;
Angle, formed between basal turn of cochlea and lateral semicircular canal, expressed in
degrees.
b
Specimens used listed in Table 5.1. Taxonomy follows Bininda-Emonds et al. (2007)
and Wible et al. (2007). Values for Kulbeckia, Zalambdalestes, and the Zhelestid are
averages (reported in Chapter 4).
150
TABLE 5.4. Dimensions of vestibular elements and orientations of semicircular canals
a
Taxon
b
Aqueduct
Stapedial
Labyrinth
Semicircular Canal Angles
Length
Ratio
Index
Ant-Lat
Ant-Post
Lat-Post
Marsupialia
Didelphis
2.58
1.6
0.0
109.0
102.0
104.0
Eutheria
Kulbeckia
1.24
2.0
0.0
79.9
79.9
89.6
Ukhaatherium
NA
1.5
0.0
88.8
105.0
88.4
Zalambdalestes
1.74
1.7
0.0
81.0
93.6
85.6
Zhelestid
1.06
1.6
0.0
88.8
96.8
93.1
Afrotheria
Chrysochloris
0.37
2.8
21.7
65.6
86.9
96.7
Elephantoidea
13.90
1.6
0.0
66.3
73.7
92.6
Hemicentetes
NA
1.6
4.1
79.3
87.9
87.0
Macroscelides
2.08
1.9
32.7
100.0
90.7
95.7
Orycteropus
8.25
1.8
0.0
78.5
91.9
87.4
Procavia
3.39
2.1
44.9
87.4
112.0
86.3
Trichechus
12.30
1.6
0.0
52.2
84.9
77.5
Xenarthra
Dasypus
2.63
1.7
23.0
62.4
67.7
87.3
Laurasiatheria
Atelerix
NA
1.8
26.4
82.2
91.7
92.1
Balaenopteridae
3.83
1.5
0.0
71.6
105.0
75.6
Bathygenys
NA
NA
45.2
86.0
99.6
91.3
Canis
NA
1.3
0.0
80.4
101.0
89.1
Equus
11.70
1.7
10.5
84.7
93.3
90.1
Eumetopias
2.26
1.5
0.0
79.7
105.0
90.6
Felis
3.77
1.9
13.1
76.8
91.4
96.7
Manis
2.45
1.7
20.5
77.0
84.8
88.6
Nycteris
NA
1.0
0.0
85.9
112.0
94.9
Pteropus
1.62
1.8
29.7
84.9
98.3
90.4
Rhinolophus
1.40
1.4
38.3
79.9
104.0
87.9
Sorex
1.58
1.7
11.9
75.3
89.6
89.3
Sus
3.18
1.3
16.5
82.8
96.0
87.9
Tadarida
1.42
2.0
22.1
74.7
98.4
98.4
Tursiops
2.23
1.4
0.0
52.2
84.9
77.5
Euarchontoglires
Cavia
3.82
2.9
25.3
77.2
105.0
85.5
151
TABLE 5.4. (Continued)
Taxon
b
Aqueduct
Stapedial
Labyrinth
Semicircular Canal Angles
Length
Ratio
Index
Ant-Lat
Ant-Post
Lat-Post
Cynocephalus
1.80
2.0
30.8
92.2
90.0
91.8
Homo
5.47
3.0
55.8
98.9
100.0
89.8
Lepus
3.71
1.7
32.4
84.2
94.0
88.6
Macaca
3.76
2.5
50.1
83.1
100.0
89.0
Mus
1.28
1.9
25.8
88.8
94.4
95.6
Sylvilagus
2.08
1.5
33.9
92.7
97.5
77.9
Tupaia
2.61
2.6
13.1
82.3
106.0
102.0
a
Measurement methodologies provided in text: Aqueduct Length, expressed in
millimeters; Stapedial Ratio, height of fenestra vestibuli over width; Labyrinth Index,
sagittal labyrinthine index (Spoor and Zonneveld, 1995); Semicircular Canal Angles,
formed between planes of canals, expressed in degrees.
b
Specimens used listed in Table 5.1. Taxonomy follows Bininda-Emonds et al. (2007)
and Wible et al. (2007). Values for Kulbeckia, Zalambdalestes, and the Zhelestid are
averages (reported in Chapter 4).
152
TABLE 5.5. Linear dimensions of the semicircular canals
a
Taxon
b
Radius
Length
Lumen Diameter
Ant
Lat
Post
Ant
Lat
Post
Ant
Lat
Post
Marsupialia
Didelphis
1.46
0.88
1.23
8.24
5.07
7.53
0.26
0.30
0.28
Eutheria
Kulbeckia
1.19
0.92
0.96
5.70
3.94
4.55
0.18
0.20
0.20
Ukhaatherium
0.84
0.74
0.69
3.81
3.16
3.39
0.17
0.13
0.15
Zalambdalestes
1.46
1.21
1.20
6.92
5.20
5.85
0.19
0.17
0.18
Zhelestid
1.17
0.79
0.86
5.80
3.49
4.62
0.19
0.19
0.19
Afrotheria
Chrysochloris
1.10
0.67
0.71
4.71
2.62
3.60
0.15
0.18
0.16
Elephantoidea
4.99
2.67
5.51
24.57
12.50
24.28
1.85
1.69
1.77
Hemicentetes
1.10
0.68
0.89
4.96
2.44
4.79
0.13
0.15
0.09
Macroscelides
1.32
1.05
1.02
5.61
4.21
5.22
0.19
0.20
0.20
Orycteropus
3.10
3.27
3.50
15.40
16.40
18.86
0.58
0.53
0.55
Procavia
1.99
1.79
2.18
10.23
7.65
10.68
0.21
0.33
0.27
Trichechus
4.30
4.46
3.54
17.30
14.20
16.53
0.51
0.52
0.51
Xenarthra
Dasypus
1.64
1.60
1.92
9.69
7.38
11.30
0.22
0.23
0.23
Laurasiatheria
Atelerix
1.24
0.88
1.22
5.88
3.67
5.80
0.16
0.15
0.15
Balaenopteridae
2.54
2.11
1.92
10.65
8.54
9.46
0.32
0.51
0.41
Bathygenys
1.91
1.52
1.79
9.72
7.11
10.02
0.44
0.33
0.38
Canis
1.73
1.57
1.43
8.58
7.08
7.37
0.31
0.35
0.33
Equus
3.62
3.55
3.50
17.35
14.30
18.63
0.51
0.45
0.48
Eumetopias
3.00
3.13
2.86
12.99
14.80
14.08
0.38
0.53
0.45
Felis
1.92
1.68
1.91
8.78
7.48
9.39
0.26
0.26
0.26
Manis
1.46
1.06
1.66
6.59
3.71
7.03
0.55
0.62
0.59
Nycteris
0.97
0.87
0.79
4.34
3.40
4.36
0.12
0.14
0.13
Pteropus
1.57
1.28
1.35
6.86
5.86
7.03
0.17
0.24
0.20
Rhinolophus
0.83
0.69
0.74
3.52
3.21
3.90
0.07
0.09
0.08
Sorex
0.65
0.48
0.63
3.20
1.63
3.42
0.12
0.14
0.13
Sus
2.50
2.08
2.18
12.14
8.04
10.65
0.42
0.39
0.41
Tadarida
0.85
0.73
0.74
3.90
3.26
3.59
0.15
0.17
0.16
Tursiops
1.19
1.36
0.84
4.14
4.61
4.35
0.27
0.25
0.26
Euarchontoglires
Cavia
1.88
1.57
1.63
9.01
6.49
8.18
0.21
0.29
0.25
153
TABLE 5.5. (Continued)
Taxon
b
Radius
Length
Lumen Diameter
Ant
Lat
Post
Ant
Lat
Post
Ant
Lat
Post
Cynocephalus
1.93
1.47
1.70
9.93
6.99
8.38
0.27
0.37
0.32
Homo
2.94
2.35
3.10
13.55
10.30
14.73
0.92
0.86
0.89
Lepus
2.34
1.66
1.69
11.45
6.86
8.10
0.27
0.26
0.26
Macaca
2.70
2.47
2.54
12.79
10.60
13.05
0.33
0.50
0.41
Mus
0.78
0.60
0.67
3.86
2.48
3.60
0.15
0.15
0.15
Sylvilagus
1.86
1.29
1.44
8.98
5.65
7.38
0.12
0.24
0.18
Tupaia
1.73
1.44
1.50
9.24
7.85
8.07
0.18
0.22
0.20
a
Measurement methodologies provided in text; expressed in millimeters.
b
Specimens used listed in Table 5.1. Taxonomy follows Bininda-Emonds et al. (2007)
and Wible et al. (2007). Values for Kulbeckia, Zalambdalestes, and the Zhelestid are
averages (reported in Chapter 4).
154
TABLE 5.6. Deviations and aspect ratios of the semicircular canals
a
Taxon
Linear
Angular
Ratio
c
Ant
Lat
Post
Ant
Lat
Post
Ant
Lat
Post
Marsupialia
Didelphis
0.22
0.38
0.00
8.62
23.70
0.00
0.97
0.79
1.08
Eutheria
Kulbeckia
0.23
0.04
0.09
11.10
2.70
5.09
1.02
0.97
1.02
Ukhaatherium
0.06
0.04
0.12
8.22
6.21
9.92
0.94
0.95
0.90
Zalambdalestes
0.08
0.13
0.14
5.83
6.32
6.85
1.08
0.88
0.98
Zhelestid
0.23
0.11
0.23
12.90
6.88
15.20
0.95
0.75
0.89
Afrotheria
Chrysochloris
0.13
0.00
0.23
6.81
0.00
18.90
1.32
1.01
0.91
Elephantoidea
1.60
0.14
1.36
18.50
3.01
14.30
0.72
1.31
1.10
Hemicentetes
0.18
0.07
0.10
9.41
5.90
6.48
0.88
0.93
0.72
Macroscelides
0.26
0.06
0.24
11.40
3.27
13.50
0.91
0.75
0.82
Orycteropus
1.06
0.41
0.70
19.70
7.21
11.50
0.81
1.03
1.28
Procavia
0.27
0.18
0.23
7.79
5.78
6.06
0.68
0.72
0.79
Trichechus
0.59
0.69
0.00
7.86
8.87
0.00
0.91
0.89
1.19
Xenarthra
Dasypus
0.37
0.50
0.26
13.00
18.10
7.76
0.58
0.96
1.16
Laurasiatheria
Atelerix
0.23
0.29
0.31
10.60
18.90
14.60
0.87
0.99
0.97
Balaenopteridae
0.40
0.20
0.53
9.03
5.44
15.90
0.91
0.39
1.21
Bathygenys
0.27
0.21
0.42
8.10
7.92
13.50
0.86
0.99
0.95
Canis
0.18
0.14
0.27
5.98
5.10
10.80
0.82
1.01
0.98
Equus
0.14
0.29
0.35
2.22
4.68
5.74
0.93
1.15
1.04
Eumetopias
0.04
0.89
0.47
0.76
16.40
9.45
0.96
1.24
1.18
Felis
0.15
0.13
0.00
4.48
4.43
0.00
0.77
1.04
1.01
Manis
0.17
0.00
0.21
6.69
0.00
7.25
0.76
0.82
0.93
Nycteris
0.07
0.10
0.31
4.14
6.61
22.70
0.91
0.71
0.95
Pteropus
0.28
0.32
0.11
10.20
14.30
4.67
0.94
0.97
0.85
Rhinolophus
0.12
0.05
0.18
8.31
4.14
13.90
0.83
0.46
0.98
Sorex
0.09
0.00
0.23
7.94
0.00
21.20
1.63
0.88
0.72
Sus
0.00
0.08
0.10
0.00
2.20
2.63
0.78
0.83
0.74
Tadarida
0.03
0.06
0.00
2.03
4.69
0.00
0.81
0.58
0.91
Tursiops
0.00
0.21
0.00
0.00
8.86
0.00
0.95
0.96
1.60
Euarchontoglires
Cavia
0.62
0.43
0.86
19.10
15.80
30.70
0.75
0.49
0.99
155
TABLE 5.6. (Continued)
Taxon
Linear
Angular
Ratio
c
Ant
Lat
Post
Ant
Lat
Post
Ant
Lat
Post
Cynocephalus
0.45
0.09
0.09
13.40
3.51
3.04
0.82
0.85
1.05
Homo
0.99
0.29
0.68
19.50
7.08
12.70
0.86
0.85
1.08
Lepus
0.16
0.06
0.32
3.92
2.07
10.90
0.86
0.87
0.81
Macaca
1.23
0.33
0.52
26.40
7.68
11.80
0.87
0.89
0.98
Mus
0.18
0.02
0.04
13.30
1.90
3.43
0.67
0.92
0.75
Sylvilagus
0.16
0.12
0.62
4.95
5.34
25.30
0.97
0.84
0.94
Tupaia
0.69
0.21
0.28
23.10
8.41
10.80
0.85
0.71
0.96
a
Measurement methodologies provided in text; linear deviations expressed in
millimeters; angular deviations expressed in degrees.
b
Specimens used listed in Table 5.1. Taxonomy follows Bininda-Emonds et al. (2007)
and Wible et al. (2007). Values for Kulbeckia, Zalambdalestes, and the Zhelestid are
averages (reported in Chapter 4).
c
Aspect ratio, calculated as height of canal arc divided by width.
156
the canal is 7.54 mm. The vestibular wall of the cochlea is expanded behind the fenestra
cochleae to accommodate the perlimphatic sac. The bony canaliculus cochleae for the
aqueduct of the cochlea extends 1.68 mm from the swelling in the cochlea as a straight
tube.
The plane of the basal turn of the cochlea is rotated ventrally and anteriorly from
the plane of the lateral semicircular canal by 19.6°. The basal end of the cochlea is
inflected at the junction between the cochlea and the spherical recess of the vestibule.
The fenestra vestibuli, in which the stapes sits, is rounded in shape, with a width versus
height (stapedial) ratio of 1.6.
The division between the spherical and elliptical recesses within the bony
vestibule is not distinct in Didelphis, although the swelling of the spherical recess is
observed in anterior view of the labyrinth. The elliptical recess is bowed slightly
medially. The anterior and posterior ends of the elliptical recess are penetrated by two
large openings each, with the anterior (medial) and lateral (lateral) ampullae in the
anterior aspect and the common crus and posterior ampulla at the posterior extremity (the
opening for the common crus is medial to that of the posterior ampulla. The lateral
semicircular canal does not possess a separate opening into the vestibule. Rather, the
posterior limb of the lateral canal joins with the lateral limb of the posterior canal to form
a secondary common crus. Presence of the secondary crus in Didelphis is a plesiomorphic
condition inherited from the ancestor of Theria.
The bony channel for the vestibular aqueduct exits the vestibule ventral and
anterior to the vestibular aperture of the common crus. The aqueduct extends dorsally and
posteriorly, crossing the common crus in medial view, as a slender and straight tube
before widening as it curves medially and becomes flattened. The channel for the
vestibular aqueduct is more robust than the canaliculus cochleae, and the vestibular
157
aqueduct is over one and a half times longer than the cochlear aqueduct (2.58 mm versus
1.68 mm).
The planes of all three semicircular canals form obtuse angles with one another,
particularly between the anterior and lateral canals (109°). The plane of the posterior
canal forms an angle of 104° with the plane of the lateral canal, and 102° with the
anterior. The posterior canal fits onto a single plane. However, both the anterior and
lateral canals deviate from their average plane (8.24° and 5.07° respectively). The course
of the anterior semicircular canal diverges medially at its midpoint, and the lateral canal
is sigmoid when viewed with its plane parallel to the horizon. Although the total angular
deviation of the anterior semicircular canal plane is greater than that calculated for the
lateral canal, the anterior canal does not deviate significantly from its plane (ratio of the
total linear deviation over canal diameter equals 0.85), whereas the deviation exhibited by
the lateral canal is significant (ratio equals 1.29).
The anterior canal is the largest of the three, in terms of slender canal length (8.24
mm; 5.07 mm and 7.53 mm for the lateral and posterior respectively) and arc radius (1.46
mm; 0.93 mm and 1.23 mm for lateral and posterior respectively). However, the cross-
sectional diameter of the lumen of the lateral semicircular canal (0.30 mm) is greater than
either the anterior (0.26 mm) or posterior (0.24 mm) canals. The arcs of the anterior and
posterior approach circularity (aspect ratios are 0.97 for the anterior arc and 1.08 for the
posterior arc), although the arc of the lateral semicircular canal is lower (0.79), being
relatively wider than either the anterior or posterior canal arcs. The ratio of the slender
semicircular canal length over arc radius of curvature calculated for the posterior canal is
greatest among the three canals (6.11; ratios for anterior and lateral canals are 5.63 and
5.47 respectively).
158
Compared to the reconstructions for the therian ancestor, the bony labyrinth of
Didelphis retains several plesiomorphic therian characters, namely the presence of the
secondary common crus and a relatively larger anterior semicircular canal compared to
the lateral and posterior canals (see Meng and Fox, 1995; Chapter 4). A third character
that likely is ancestral for therians, or at least eutherians, is a cochlea that is coiled to
around 360°. However, the cochlea is derived for Didelphis in this regard, as it completes
over two turns (Table 5.3). Lastly, the position of the lateral semicircular canal of
Didelphis is similar to that of Cretaceous eutherians (Chapter 4) in that the lateral canal
does not divide the space enclosed by the posterior canal in anterior view, a condition that
is observed in many placentals, including the golden mole Chrysochloris (see below).
Eutheria
Eutheria is defined as the most recent common ancestor of crown Placentalia and
all taxa more closely related to Placentalia than to Marsupialia (the marsupial mammals).
a brief overview of the labyrinth of Kulbeckia kulbecke is provided here as a
representative of non-placental Mesozoic eutherians (which are thought to exhibit the
ancestral condition for Eutheria, if not Theria; Chapter 4), although data from three
additional non-placental eutherian taxa were used to reconstruct hypothetical ancestral
states. The non-placental eutherian taxa that were examined are from the Cretaceous of
Asia, and they include a representative of a monophyletic group of zhelestids from the
Bissekty Formation of Uzbekistan (see Chapter 4), Kulbeckia kulbecke (also from
Uzbekistan), Ukhaatherium nessovi, and Zalambdalestes lechei (the latter two taxa from
Mongolia). The relationships depicted for these taxa in Figure 5.2 follow Wible et al.
(2007), and more thorough descriptions of the bony labyrinths of these taxa are provided
159
elsewhere (Chapter 4). The gross morphology of the non-placental taxa does not vary
significantly among the taxa examined, and the values of measurements for Kulbeckia are
averages across a sample of four petrosals reported in Chapter 4 of this dissertation. The
bony labyrinth of Kulbeckia is illustrated in Figures 5.6-5.7.
The total length of the bony labyrinth of Kulbeckia is 4.73 mm (labyrinth length
could only be measured for two specimens, URBAC 00-16 and URBAC 04-36), and the
cochlea contributes 48.2% of the total volume of the inner ear (2.59 out of 5.37 mm
3
).
The contribution of the cochlea to the entire bony labyrinth is noticeably less in
Kulbeckia than in Didelphis. Likewise, both the aspect ratio of the cochlear spiral (0.44)
and the degree of coiling of the cochlea (446°) are smaller in Kulbeckia than in Didelphis.
The spiral length of the cochlear canal of Kulbeckia is 4.11 mm. The canaliculus cochleae
extends for 0.60 mm from a swelling of the cochlea near the fenestra cochleae.
The plane of the basal turn of the cochlea is rotated from the plane of the lateral
semicircular canal by an average of 12.1° in Kulbeckia. The basal end of the cochlea is
inflected before it joins the spherical recess of the vestibule near the fenestra vestibuli
(average stapedial ratio of 1.9; Ekdale et al., 2004). The spherical and elliptical recesses
are distinguished by a constriction of the vestibule lateral to the fenestra vestibuli.
As was observed in Didelphis, the posterior and lateral semicircular canals fuse to
form a secondary common crus, which in turn empties into the posterior ampulla. The
common crus between the anterior and posterior semicircular canals is situated medial to
the posterior ampulla. The bony channel for the vestibular aqueduct was observed
anteromedial to the common crus in two Kulbeckia specimens (URBAC 00-16 and 04-
36), extending for an average of 1.24 mm.
The planes of the lateral and posterior semicircular canals almost form a right
angle (89.6°), but the angles that each of these canals form with the anterior canal are
160
FIGURE 5.6. Bony labyrinth of Kulbeckia kulbecke. A, stereopair and labeled line
drawing of digital endocast in anterior view; B, stereopair and labeled line drawing of
digital endocast in dorsal view; C, stereopair and labeled line drawing of digital endocast
in lateral view; D, line drawing of cochlea viewed down axis of rotation to display degree
of coiling; E, line drawing of cochlea in profile. Abbreviations: aa, anterior ampulla; ac,
anterior semicircular canal; ant, anterior direction; cc, canaliculus cochleae for aqueduct
of cochlea; co, cochlea; cr, common crus; dor, dorsal direction; er, elliptical recess of
vestibule; fc, fenestra cochleae; fv, fenestra vestibuli; la, lateral ampulla; lc, lateral
semicircular canal; med, medial direction; pa, posterior ampulla; pc, posterior
semicircular canal; pl, primary bony lamina; pos, posterior direction; scr, secondary
common crus; sl, secondary bony lamina; sr, spherical recess of vestibule.
161
co
co
sl
sl
sl
pl
cc
cc
cc
fc
fc
fc
fv
co
co
cc
fc
co
la
aa
ac
cr
lc
la
scr
aa
er
ac
lc
cr
pa
pa
pc
pc
sr
er
aa
la
cr
ac
lc
pc
A
B
C
D E
1 mm
1 mm
1 mm
dor
pos
med
med
dor
ant
162
FIGURE 5.7. CT slices through ear region of Kulbeckia kulbecke. Numbers refer to
specific CT slices. Abbreviations: aa, anterior ampulla; ac, anterior semicircular canal;
ant, anterior direction; cn, canal for cranial nerve VIII; co, cochlea; cr, common crus; fc,
fenestra cochleae; fn, canal for cranial nerve VII; fv, fenestra vestibuli; la, lateral
ampulla; lc, lateral semicircular canal; med, medial direction; pa, posterior ampulla; pc,
posterior semicircular canal; pl, primary bony lamina; pr, promontorium housing
cochlea; sa, subarcuate fossa; scr, secondary common crus; sg, canal for spiral ganglion
within primary bony lamina; sl, secondary bony lamina; ven, ventral direction.
163
med
ven
178
209
240
271
302
333
364
395
426
488
452
1 mm
ant
pc
cr
pc
cr
cr
co
co
co
co
co
co
co
sg
sl
cn
cn
cn
pl
cc
fc
fv
fn
ac
ac
sa
sa
sa
pa
pa
la
aa
scr
lc
lc
lc
ven
5 mm
178 209
302271240
395364333
388457426
sg
sl
sg
pl
pl
pl
164
acute (79.9° for both). The anterior semicircular canal deviates the most from its plane
(11.1°), and the lateral canal is the most planar (deviation of 2.70°; posterior canal
deviates by 5.09°). Only the anterior canal deviates from its plane by a significant
amount.
The anterior semicircular canal is the largest in terms of radius (1.19 mm; 0.92
mm and 0.96 mm for lateral and posterior respectively) and slender canal length (5.70
mm; 3.94 mm and 4.55 mm for lateral and posterior respectively). However, the lateral
canal is the largest in terms of cross-sectional diameter (0.20 mm; 0.18 mm and 0.19 mm
for anterior and posterior respectively). The aspect ratios of the anterior and posterior
semicircular canals are identical (1.02) with arcs that are higher than they are wide. In
contrast, the arc of the lateral semicircular canal is wider than it is high (ratio equals
0.97). The ratio of the slender anterior semicircular canal length over arc radius of
curvature (4.80) is the largest ratio calculated among the three canals, although the ratio
for the posterior canal is close to that of the anterior (4.75; ratio for lateral canal is 4.29).
The inner ear morphology of Kulbeckia and the other Mesozoic taxa, as well as
Didelphis, were used to reconstruct the ancestral states of Eutheria. The bony labyrinth of
the ancestor of Eutheria retained the ancestral therian conditions in all respects. The
lateral semicircular canal formed a secondary common crus with the posterior canal, the
plane of the lateral canal was low compared to the ampullar entrance of the posterior
semicircular canal, the arc of the anterior semicircular canal was the largest among the
three semicircular canals, and the aspect ratio of the cochlea was low (below 0.55). All
ancestors at the nodes leading to crown Placentalia retained the ancestral eutherian states
for all discrete characters.
The contribution of the ancestral eutherian cochlea to the total inner ear volume
was 64%, which was less than that reconstructed for Theria (66%), and the percentage
165
decreased through time (59% for the most recent common ancestor of Ukhaatherium and
Placentalia; 56% for the most recent common ancestor of Zalambdalestidae, which
includes Kulbeckia and Zalambdalestes, and Placentalia). The contribution of the cochlea
of the ancestral zalambdalestid was 51%.
The ancestral eutherian cochlea completed 580°, which was less than that
reconstructed as the ancestral therian condition (685°). The most recent common ancestor
of Ukhaatherium and Placentalia completed 510°, whereas the cochlea of the ancestor of
Zalambdalestidae and Placentalia completed 570°. The ancestral condition of the cochlea
reconstructed for Zalambdalestidae was 461°.
Placentalia
Placentalia includes the most recent common ancestor of extant placental
mammals (e.g., Hemicentetes semispinosum, Dasypus novemcinctus, and Homo sapiens)
plus all of its descendants. Placentalia is divided into the three major lineages Afrotheria,
Xenarthra, and Boreoeutheria, which in turn is divided into Laurasiatheria and
Euarchontoglires (Murphy et al., 2001b; Bininda-Emonds et al., 2007).
Entry of the lateral semicircular canal directly into the vestibule in absence of a
secondary common crus is the single unambiguous otic synapomorphy for Placentalia,
which is a condition not found outside of the crown (at least within Eutheria). The vast
majority of placental taxa lack a secondary common crus (only exceptions among
sampled taxa are Orycteropus afer and Canis familiaris). The cochlea of the ancestor of
placental mammals completes 738° (over two turns), which is almost one half turn
greater than the cochlea of the most recent common ancestor of Placentalia and
166
Zalambdalestidae (570°). The volumetric contribution of the cochlea to the entire
labyrinth (58%) is less than that of the ancestral eutherian (64%).
The arc of the anterior semicircular canal is the largest among the three canal arcs,
which is retained from the ancestor of Theria. The reconstructed states of both the
position of the plane of the lateral semicircular canal compared to the ampullar entrance
of the posterior canal and the aspect ratio of the cochlea in profile are equivocal owing to
variation in the position of the lateral canal within Afrotheria and variation in the shape
of the cochlear spiral in both Afrotheria and Boreoeutheria.
Afrotheria
Afrotheria is a clade of placentals endemic to Africa that includes the groups
Afrosoricida (tenrecs and golden moles), Macroscelidea (elephant shrews), Tubulidentata
(aardvark), Hyracoidea (hyraxes), Sirenia (dugongs and manatees), and Proboscidea
(elephants). Monophyly of Afrotheria is controversial, primarily because it was not
recognized in classical morphological studies of placentals, whether based on strict
cladistic methodologies or not (Gregory, 1910; Simpson, 1945; McKenna, 1975;
Novacek, 1986; Novacek and Wyss, 1986b; Novacek et al., 1988; Novacek, 1992a, b,
1993; McKenna and Bell, 1997). Monophyletic Afrotheria (including the afrosoricids and
macroscelids) was first proposed by Springer et al. (1997), although the first use of the
name “Afrotheria” was by Stanhope et al. (1998).
The earliest support for Afrotheria as a whole was restricted to molecular
evidence (Springer et al., 1997; Stanhope et al., 1998; Springer et al., 1999; Madsen et
al., 2001; Murphy et al., 2001a; van Dijk et al., 2001). Although more recent
morphological evidence has been proposed to support the clade (Asher, 2001; Mess and
167
Carter, 2006; Sánchez-Villagra et al., 2007; Seiffert, 2007), strict morphological analyses
fail to recover afrotherian monophyly (Asher et al., 2003; Wible et al., 2007).
The members of Afrotheria studied here are Macroscelides proboscideus
(Macroscelidea), Orycteropus afer (Tubulidentata), a fossil elephantoid proboscidean
(either Mammut or Mammuthus; see Chapter 2), Trichechus manatus (Sirenia), Procavia
capensis (Hyracoidea), and the two afrosoricids Chrysochloris sp. (Chrysochloridae) and
Hemicentetes semispinosum (Tenrecidae). There is a broad range in body mass among
these taxa (Table 5.2), from 44 grams in Chrysochloris (Silva and Downing, 1995) to
upwards of 8,000 kg in extinct elephantoids (Christiansen, 2004). Likewise, the inner ear
cavities vary in size. The overall volume of the bony labyrinth within Afrotheria ranges
from 4.11 mm
3
in Chrysochloris to 26.0 mm
3
in the fossil elephantoid. Dimensions of the
bony labyrinths of afrotherians are provided in Table 5.2. Dimensions of the cochlea are
provided in Table 5.3, and dimensions and orientations of the semicircular canals are
reported in Tables 5.4-5.6.
Afrotheria often is placed as the sister taxon to all other placentals (e.g., Murphy
et al., 2001a, b), although the results of Bininda-Emonds et al. (2007) include Afrotheria
in a basal polytomy with Xenarthra and a clade comprised of the remaining placentals.
Three major lineages are included within Afrotheria, which are Tubulidentata
(aardvarks), Paenungulata (hyraxes, manatees, and elephants), and a clade including
Macroscelidea (elephant shrews) and Afrosoricida (golden moles and tenrecs). The three
major lineages are placed include these lineages within a polytomy at the base of
Afrotheria (see Bininda-Emonds et al., 2007; Figure 5.2).
The bony labyrinth of the ancestor of Afrotheria retained the ancestral
morphology of Placentalia in that the lateral semicircular canal entered into the vestibule
directly and the arc of the anterior semicircular canal was the greatest among the three
168
canal arcs. The reconstructed states of the position of the lateral semicircular canal
compared with the posterior canal, as well as the aspect ratio of the cochlea, are
equivocal. The states reconstructed for all of the nodes within Afrotheria are identical to
that of the afrotherian ancestor, except the state for the largest semicircular canal arc is
equivocal for the clade consisting of Procavia and Trichechus (the posterior arc is largest
for Procavia and the lateral is largest for Trichechus; see below).
The volumetric contribution of the cochlea to the total labyrinthine volume of the
ancestral afrotherian was 56.0%, which was close to that reconstructed for the ancestor of
Placentalia (58.0%). The ancestral cochlear contribution of the paenungulate clade
consisting of Procavia and Trichechus was the same as that of the afrotherian ancestor
(56.0%), although the contribution of the cochlea of the ancestor of Paenungulata was
almost ten percent less (48.0%). Contributions of 63.0% and 64.0% were reconstructed
for the ancestors of Afrosoricida and the more inclusive clade that also includes
Macroscelidea, respectively. The ancestral afrotherian cochlea coiled 751°, which was
greater than the ancestral placental condition, but less than the values reconstructed for
the nodes within Afrotheria (768° for the clade consisting of Afrosoricida and
Macroscelides; 833° for Afrosoricida; 790° for Paenungulata; 853° for the clade
consisting of Procavia and Trichechus).
Afrosoricida
The group Afrosoricida contains Tenrecidae (tenrecs) and Chrysochloridae
(golden moles). Although traditional classifications (e.g., Simpson, 1945) group tenrecids
and chrysochlorids with other insectivorous mammals, such as the lipotyphlans
Erinaceus (hedgehog) and Sorex (shrew), the results of more recent molecular studies
(e.g., Springer et al., 1997; Stanhope et al., 1998), ally Tenrecidae and Chrysochloridae
169
with other placentals within the clade of African endemic mammals Afrotheria. Each
afrosoricid group is represented by a single taxon, Chrysochloris sp. (Chrysochloridae)
and Hemicentetes semispinosum (Tenrecidae).
The bony labyrinths of Chrysochloris and Hemicentetes differ in several ways,
one of which is absolute size (Figures 5.8-5.11), where the former tends to be smaller
than the latter (also observed in body mass where Chrysochloris sp. is 44.4 grams and H.
semispinosum is 110 grams; Silva and Downing, 1995). For example, the labyrinth is
4.08 mm long in Hemicentetes, and 3.93 mm in Chrysochloris. However, the volume of
the bony labyrinth is smaller in Hemicentetes (2.78 mm
3
) than in Chrysochloris (4.11
mm
3
).
The cochlear canal is not only longer in Chrysochloris (6.65 mm versus 3.79
mm), but the spiral also completes a greater degree of coiling (1191° or 3.3 turns versus
540° or 1.5 turns). The absolute volume of the cochlea is larger in Chrysochloris (2.93
mm
3
versus 1.39 mm
3
in Hemicentetes), and the proportion of the total labyrinth volume
is greater in Chrysochloris (71.3%) than in Hemicentetes (49.9%).
The cochlea is more planispiral in Hemicentetes (aspect ratio of 0.38) than it is in
Chrysochloris (aspect ratio of 0.63). The second turn of the cochlea in Chrysochloris is
nearly equal in diameter to the basal turn, and obscures most of the basal turn when the
cochlea is viewed tympanally (from down the axis of rotation), although the apical turn
nearly fits within the arc of the basal turn of the cochlea of Hemicentetes. The plane of
the basal turn of the cochlea of Hemicentetes also is rotated less from the plane of the
lateral semicircular canal than it is in Chrysochloris (18.4° versus 41.9°).
The bony canaliculus cochleae for the aqueduct of the cochlea is shorter in
Hemicentetes (0.28 mm versus 0.45 mm in Chrysochloris), and the cochlea is expanded
170
FIGURE 5.8. Bony labyrinth of Chrysochloris sp. A, stereopair and labeled line drawing
of digital endocast in anterior view; B, stereopair and labeled line drawing of digital
endocast in dorsal view; C, stereopair and labeled line drawing of digital endocast in
lateral view; D, line drawing of cochlea viewed down axis of rotation to display degree of
coiling; E, line drawing of cochlea in profile. Abbreviations: aa, anterior ampulla; ac,
anterior semicircular canal; ant, anterior direction; co, cochlea; cr, common crus; dor,
dorsal direction; er, elliptical recess of vestibule; fc, fenestra cochleae; fv, fenestra
vestibuli; la, lateral ampulla; lc, lateral semicircular canal; med, medial direction; pa,
posterior ampulla; pc, posterior semicircular canal; pl, primary bony lamina; pos,
posterior direction; ps, outpocketing for perilymphatic sac; sr, spherical recess of
vestibule.
171
ac
pc
pa
cr
lc
fc
ps
er
sr
fv
co
co
aa
la
sr
er
aa
la
ac
1 mm
1 mm
1 mm
lc
pc
cr
ps
pl
pa
co
er
sr
ps
cc
aa
la
cr
pa
ac
pc
lc
fc
co
fc
ps
co
ps
A
B
C
D E
dor
med
pos
med
dor
ant
172
FIGURE 5.9. CT slices through ear region of Chrysochloris sp. Numbers refer to specific
CT slices. Abbreviations: aa, anterior ampulla; ac, anterior semicircular canal; cn, canal
for cranial nerve VIII; co, cochlea; cr, common crus; dor, dorsal direction; er, elliptical
recess of vestibule; fn, canal for cranial nerve VII; fv, fenestra vestibuli; la, lateral
ampulla; lc, lateral semicircular canal; med, medial direction; pc, posterior semicircular
canal; pl, primary bony lamina; pos, posterior direction; sr, spherical recess of vestibule;
st, stapes within fenestra vestibuli; vb, vestibule.
173
7
1 mm
1 mm
15
2922 36
5043 57
7164 78
pos
med
ma
ma
ma
aa
ac
ma
ma
dor
med
co
7
15
22
29
36
43
50
57
64
78
71
co
cn
pl
co
co
co
pl
fn
cn
sr
vb
er
er
cr
ac
lc
er
cr
ac
lc
lc
pc
pc
cn
cn
fv
st
er
pc
ac
lc
la
174
FIGURE 5.10. Bony labyrinth of Hemicentetes semispinosum. A, stereopair and labeled
line drawing of digital endocast in anterior view; B, stereopair and labeled line drawing
of digital endocast in dorsal view; C, stereopair and labeled line drawing of digital
endocast in lateral view; D, line drawing of cochlea viewed down axis of rotation to
display degree of coiling; E, line drawing of cochlea in profile. Abbreviations: aa,
anterior ampulla; ac, anterior semicircular canal; ant, anterior direction; cc, canaliculus
cochleae; co, cochlea; cr, common crus; dor, dorsal direction; er, elliptical recess of
vestibule; fc, fenestra cochleae; fv, fenestra vestibuli; la, lateral ampulla; lc, lateral
semicircular canal; med, medial direction; pa, posterior ampulla; pc, posterior
semicircular canal; pos, posterior direction; ps, outpocketing for perilymphatic sac; sl,
secondary bony lamina; sr, spherical recess of vestibule.
175
co
sl
fv
ps
sr
aa
la
cr
ac
pc
lc
pa
fc
co
sl
fc
cc
pa
pc
1 mm
1 mm
1 mm
er
la
aa
lc
ac
cr
ps
co
craa
la
lc
er
fc
cc
cc
ps
pa
ac
pc
co
fc
fc
co
sl
ps
ps
A
B
C
D E
dor
med
pos
med
dor
ant
176
FIGURE 5.11. CT slices through ear region of Hemicentetes semispinosum. Numbers
refer to specific CT slices. Abbreviations: aa, anterior ampulla; ac, anterior semicircular
canal; co, cochlea; cr, common crus; dor, dorsal direction; er, elliptical recess of
vestibule; fn, canal for cranial nerve VII; la, lateral ampulla; lc, lateral semicircular canal;
med, medial direction; pa, posterior ampulla; pc, posterior semicircular canal; pos,
posterior direction; sa, subarcuate fossa; vb, vestibule.
177
1 mm
1 mm
292 296
304300 308
316312 320
328324 332
pos
dor
med
dor
292
296
300
304
308
312
316
320
324
332
328
ac
ac
ac
aa
aa
sa
sa
sa
vb
vb
vb
cr
sa
fn
fn
co
ac
er
la
lc
pc
pc
pc
pc
pc
er
lc
lc
lc
lc
co
co
co
co
vb
vb
pa
co
co
178
behind the fenestra cochleae in Chrysochloris for the perilymphatic sac in both
afrosoricid species. The swelling hooks posteriorly before the canaliculus cochleae exits
the cochlea, and the canaliculi of these taxa are not as delicate as that observed in
Didelphis. The fenestra vestibuli is more oval in Chrysochloris, with a stapedial ratio of
2.8. The ratio is 1.6 in Hemicentetes.
The spherical and elliptical recesses are distinguishable within the vestibule of
Chrysochloris, where the former projects anteriorly towards the cochlea. As a whole, the
spherical recess is ovoid in shape. The elliptical recess is smaller than the spherical recess
in Chrysochloris, and forms a gently curved tube with openings for the semicircular canal
system in dorsal view. Each end of the tube is extended into a chamber dorsally, and each
extension is expressed as a short pedestal in the endocast. The anterior and posterior
ampullae open into the anterior chamber of the elliptical recess. The posterior chamber is
penetrated by three major apertures, which lead to the posterior ampulla, the common
crus, and the posterior limb of the lateral semicircular canal. The latter of these apertures
opens into the vestibule near the anterodorsal edge of the opening for the posterior
ampulla.
The recesses are less distinguishable in Hemicentetes, and the vestibule forms a
continuous cavity, albeit irregular in shape. The vestibule of Hemicentetes is penetrated
by four major openings only, as opposed to five in Chrysochloris. The openings into the
vestibule of Hemicentetes, in addition to the junction of the vestibule and cochlea, lead to
the ampullae of the three semicircular canals, as well as the common crus. As in
Didelphis, the posterior limb of the lateral semicircular canal does not have its own
aperture into the vestibule in Hemicentetes, as is the case with Didelphis. However, the
lateral canal of Hemicentetes does not join with the posterior semicircular canal to form a
secondary common crus which is observed in the opossum,; rather the lateral canal
179
empties into the posterior ampulla. A groove on the anterior wall of the posterior ampulla
that presumably accommodated the membranous lateral semicircular duct in life extends
from the lateral semicircular canal to the vestibule. Such a condition suggests the lateral
duct is separate from the membranous posterior ampulla in this species, although such
cannot be determined with certainty through examination of the bony labyrinth alone.
No trace of the bony channel for the vestibular aqueduct was observed in the CT
slices of Hemicentetes. The aqueduct likely is present (there is no record of any mammal
lacking this structure) but likely is small and too narrow to be captured on the CT
imagery. The channel for the aqueduct is observed in CT data of Chrysochloris, in which
the channel exits the elliptical recess medial to the common crus. An indistinct groove for
the endolymphatic duct along the medial wall of the elliptical recess (expressed as a ridge
on the endocast) extends from the channel for the vestibular aqueduct to the junction
between the elliptical and spherical recesses.
The planes of the semicircular canals do not form right angles with one another in
either afrosoricid species examined. In Chrysochloris, the largest angle was measured
between the posterior and lateral canals (96.7°), and the smallest was measured between
the anterior and lateral canals (65.6°). The angle between the anterior and posterior
semicircular canal planes is 86.9°. The widest angle measured in Hemicentetes is between
the anterior and posterior canals (87.9°), which not only is the largest angle between two
semicircular canals in either taxon, but it is also the closest angle to 90° in the labyrinths
of either Chrysochloris or Hemicentetes. The smallest angle in Hemicentetes is between
the posterior and lateral canals (79.3°), and the angle between the anterior and lateral
semicircular canals in this species is 87.0°.
The anterior canal is the largest of all semicircular canals in terms of volume,
slender canal length, and arc radius for both afrosoricid taxa included in the present study
180
(0.031 mm
3
, 4.71 mm, and 1.10 mm respectively in Chrysochloris; 0.37 mm
3
, 4.96 mm,
and 1.10 mm respectively in Hemicentetes). Likewise, the lateral semicircular canal was
the smallest in both species in at least slender canal length and arc radius (2.62 mm and
0.67 mm respectively for Chrysochloris; 4.69 mm and 1.38 mm respectively for
Hemicentetes).
Not only is the anterior semicircular canal the largest among all of the canals, it
has the highest aspect ratio in Chrysochloris (1.32 versus 0.88 in Hemicentetes),
indicating that the height of the arc is larger in proportion to the width than it is in other
canals. The height of the lateral semicircular canal arc is nearly equal to the width of that
arc in Chrysochloris (ratio of 1.01), and nearly so in Hemicentetes (ratio of 0.93). The
posterior canal arc has a ratio of 0.91 in Chrysochloris, and 0.72 in Hemicentetes. In fact,
the aspect ratio of the posterior canal arc is the lowest among all of the canals between
the two species. The ratios between the length of a slender semicircular canal and the
radius of its arc for Chrysochloris are 4.30 for the anterior canal, 3.89 for the lateral
canal, and 5.07 for the posterior canal. A similar pattern is observed in Hemicentetes
where the posterior semicircular canal has the highest canal length to arc radius ratio
(5.41), and the lateral canal has the lowest (3.59; ratio for anterior canal equals 4.52).
The angular deviation of the anterior and lateral semicircular canals from their
planes in Chrysochloris are less than that observed for the same canals in Hemicentetes,
although the lateral semicircular canal of Chrysochloris is the only planar canal between
the taxa (the total angular deviation of the lateral canal in Hemicentetes is 5.90°). The
least planar canal in Chrysochloris is the posterior (18.9°; 6.48 ° in Hemicentetes), and
the posterior is the only canal to deviate substantially from its plane in Chrysochloris
(ratio of the total linear deviation over cross-sectional diameter of the posterior canal is
1.31; ratio for the anterior canal is 0.87). Both the anterior and posterior canals of
181
Hemicentetes deviate substantially from their planes (ratios are 1.38 and 1.11
respectively), although the ratio is only 0.47 for the lateral semicircular canal. The
anterior semicircular canal of Hemicentetes deviates from its plane by a total of 9.4°,
whereas the same canal deviates by 6.8° in Chrysochloris.
Lastly, the plane of the lateral semicircular canal is high with respect to the
posterior canal in both Chrysochloris and Hemicentetes to an extent that it divides the
space enclosed by the arc of the posterior semicircular canal into dorsal and ventral
sections when the labyrinth is oriented in anterior view. Within Afrotheria, a similar
condition is observed in Macroscelides and Procavia as described below. The sagittal
labyrinthine index (following Spoor and Zonneveld, 1995) of Chrysochloris is 21.7 and
4.1 in Hemicentetes, both of which are lower than the other afrotherians exhibiting this
feature of the bony labyrinth (see below). The labyrinthine index of Hemicentetes is
smaller than that calculated for any other mammal in this study (in which the lateral canal
divides the space enclosed by the posterior canal arc when the labyrinth is in anterior
view).
The bony labyrinths of both afrosoricid taxa retain the ancestral placental
condition in that the lateral semicircular canal does not form a secondary common crus
(although the canal opens into the posterior ampulla rather than the vestibule in
Hemicentetes), and the anterior semicircular canal has the greatest radius among the three
canals. Although the cochlea of Chrysochloris exhibits a great degree of coiling (over
three complete turns at 1191°; Table 5.3), the coiling in Hemicentetes (540° or 1.5 turns)
is only slightly greater than the average calculated for zhelestids from the Bissekty
Formation (518° or 1.4 turns; Chapter 4), and nearly 200° (over one half turn) less than
the ancestral placental condition.
182
Both Hemicentetes and Chrysochloris are derived with respect to the ancestral
eutherian condition in the placement of the lateral semicircular canal that visually divides
the space enclosed by the posterior semicircular canal when the labyrinth is in anterior
view. Such a condition is not observed in Didelphis or any Mesozoic eutherian, including
Kulbeckia as described above (also see Chapter 4 of this dissertation).
Macroscelidea
Macroscelidea contains the elephant shrews or sengis. The phylogenetic affinities
of Macroscelidea are contentious, although the analyses of Bininda-Emonds et al. (2007),
as well as other molecular studies (e.g., Springer et al., 1998; Murphy et al., 2001a) group
macroscelideans inside Afrotheria. Within Afrotheria, Macroscelides holds a sister
relationship with Afrosoricida (Bininda-Emonds et al., 2007).
Only one species of Macroscelidea (Macroscelides proboscideus) was examined
in the present study (Figures 5.12-5.13). The average body mass of Macroscelides (38.4
grams) is less than either afrosoricid taxon examined (44.4 grams for Chrysochloris sp.
and 110 grams for Hemicentetes semispinosum; mass for Chrysochloris is an average
across the genus because the species of Chrysochloris used is unknown) as reported by
Silva and Downing (1995). However, the dimensions of the bony labyrinth of
Macroscelides tend to be intermediate between the afrosoricids. The volume of the
combined inner ear cavities in Macroscelides is 9.19 mm
3
(4.11 mm
3
for Chrysochloris;
2.78 mm
3
for Hemicentetes) and the labyrinth is 4.31 mm in length (3.93 mm for
Chrysochloris; 4.08 mm for Hemicentetes). Likewise, the skull length of the
Macroscelides specimen (32.26 mm) is intermediate between Chrysochloris (23.85 mm)
and Hemicentetes (30.61 mm).
183
FIGURE 5.12. Bony labyrinth of Macroscelides proboscideus. A, stereopair and labeled
line drawing of digital endocast in anterior view; B, stereopair and labeled line drawing
of digital endocast in dorsal view; C, stereopair and labeled line drawing of digital
endocast in lateral view; D, line drawing of cochlea viewed down axis of rotation to
display degree of coiling; E, line drawing of cochlea in profile. Abbreviations: aa,
anterior ampulla; ac, anterior semicircular canal; ant, anterior direction; av, bony channel
for aqueduct of vestibule; cc, canaliculus cochleae for aqueduct of cochlea; co, cochlea;
cr, common crus; cv, canal for cochlear vein; dor, dorsal direction; er, elliptical recess of
vestibule; fc, fenestra cochleae; fv, fenestra vestibuli; la, lateral ampulla; lc, lateral
semicircular canal; med, medial direction; pa, posterior ampulla; pc, posterior
semicircular canal; pl, primary bony lamina; pos, posterior direction; ps, outpocketing
for perilymphatic sac; sl, secondary bony lamina; sr, spherical recess of vestibule.
184
sr
fv
fc
la
aa
er
cr
ac
pc
lc
pl
sl
co
av
cr
pa
pc
fv
fc
aa
la
ac
lc
sr
er
fc
fv
la
aa
pa
cr
av
lc
ac
pc
co
cc
co
cv
co
sl
sl
ps
ps
fc
cv
cc
cc
cv
A
B
C
D E
co
1 mm
1 mm
1 mm
dor
med
pos
med
dor
ant
185
FIGURE 5.13. CT slices through ear region of Macroscelides proboscideus.
Abbreviations: aa, anterior ampulla; ac, anterior semicircular canal; av, bony channel for
aqueduct of vestibule; cc, canaliculus cochleae; cn, canal for cranial nerve VIII; co,
cochlea; cr, common crus; dor, dorsal direction; er, elliptical recess of vestibule; fc,
fenestra cochleae; la, lateral ampulla; lc, lateral semicircular canal; med, medial
direction; pa, posterior ampulla; pc, posterior semicircular canal; pl, primary bony
lamina; pos, posterior direction; sa, subarcuate fossa; sl, secondary bony lamina; sr,
spherical recess of vestibule; st, stapes within fenestra vestibuli.
186
54
co
co
co
co
co
cn
cn
sl
sl
sl
sl
cn
st
st
sr
sr
la
la
ac
ac
ac
ac
sa
sa
sa
sa
sa
lc
lc
lc
pa
cr
cr
cr
pc
pc
pc
pc
lc
eraa
av
pa
er
fc
aa
pl
pl
pl
63
8172 90
10899 117
135126 144
63
72
81
90
99
108
117
126
144
135
pos
med
1 mm
dor
med
1 mm
ac
av
av
av
pl
cv
cc
187
The cochlea of Macroscelides contributes 71.3% of the total volume of the inner
ear (6.59 mm
3
out of 9.19 mm
3
; Figure 5.4). This proportion is almost identical to that of
Chrysochloris. Furthermore, the cochlea of Macroscelides not only has a higher aspect
ratio than Hemicentetes (0.80 versus 0.38; ratio of Chrysochloris is 0.63), but the ratio in
Macroscelides is higher than that calculated for any other afrotherian examined in this
study. The second turn of the cochlea in Macroscelides sits upon the basal whorl, as is
observed in Chrysochloris, but not Hemicentetes. The cochlear spiral of Macroscelides
completes two whorls (720°), and the total length of the canal is 7.11 mm.
The secondary bony lamina for the basilar membrane is well developed on the
radial wall of the basal turn of the cochlea (expressed as the distinct groove on the
endocast), and is prolonged beyond three quarter turns of the basal whorl, but ends before
the basal whorl is complete. The secondary lamina curves sharply at the basal end of the
cochlear canal, between the fenestrae cochleae and vestibuli, the latter of which has an
aspect ratio of 1.9 (similar to that observed in Hemicentetes). The plane of the basal turn
of the cochlea deviates from that of the lateral semicircular canal by 25.1° in
Macroscelides, which is greater than Hemicentetes (18.4°), but not as great as
Chrysochloris (41.9°).
A triangular outpocketing for the perilymphatic sac that leads to two bony canals
is situated medial to the fenestra cochleae. One of these canals is situated ventral to the
other. The dorsal canal extending from the triangular outpocketing is the bony
canaliculus cochleae for the aqueduct of the cochlea, and the ventral channel transmits a
cochlear vein in life (which is observed traveling with the membranous aqueduct of
various mammal species; Gray, 1907, 1908). The ventral canal for the cochlear vein is
more delicate than the canaliculus cochleae, and forms a straight tube that widens slightly
before it opens on the medial surface of the petrosal. Unlike the canal for the cochlear
188
vein, the canaliculus cochleae widens rapidly as it extends away from the cochlea,
forming a pyramid shaped conduit for the aqueduct of the cochlea. The canaliculus
cochleae is 0.58 mm in length, which is larger than the channel in either Chrysochloris
(0.45 mm) or Hemicentetes (0.51 mm). The canaliculus is more easily observed in the CT
scans of Macroscelides than in the data of afrosoricids, also.
The division between the spherical and elliptical recesses is not well defined in
Macroscelides, although there is a slight constriction in the vestibule lateral to the
fenestra vestibuli. The elliptical recess of the vestibule is not curved in Macroscelides,
but the chamber contains five major openings, as also is observed in the bony labyrinth of
Chrysochloris, for the common crus, ampullae of the semicircular canals, and a separate
opening for the posterior limb of the lateral semicircular canal near the vestibular
aperture of the posterior ampulla.
The bony channel for the aqueduct of the vestibule exits the spherical recess
anterior to the common crus at a greater distance than it does in Didelphis. The channel is
a straight tube of uniform diameter across most of its length (2.08 mm, which is
significantly longer than the canaliculus cochleae at 0.58 mm) until it turns
posterolaterally and flattens into a fissure-like chamber before opening onto the
endocranial surface of the petrosal (near the union of the posterior and anterior
semicircular canals at the apex of the common crus).
The semicircular canals form gentle curves, and the planes of the anterior and
posterior canals nearly form a right angle (90.7°). The planes of the lateral and other two
semicircular canals form obtuse angles (95.7° between lateral and posterior; 100°
between lateral and anterior). None of the semicircular canals fit onto a single plane,
particularly the anterior canal where a significant deviation of 26.4° was measured (ratio
of total linear deviation over cross-sectional diameter is 1.37). The posterior canal is the
189
most planar of the three (total deviation equals 5.9°; lateral canal deviates 7.7°), although
the deviation of the posterior canal is significant, whereas the deviation of the lateral
canal is not (ratios of linear deviation to canal diameter are 1.36 and 0.30 respectively).
The anterior semicircular canal is the longest of the three canals (5.61 mm; lateral
equals 4.21 mm; posterior equals 5.22 mm), and the radius of the anterior canal arc (1.32
mm) is larger than either the lateral (1.05 mm) or posterior (1.02 mm) arc canals.
However, the cross-sectional diameter of the lumen of the posterior canal is larger than
the anterior canal (0.47 mm versus 0.33 mm), although the largest diameter was
measured for the lateral canal (0.50 mm).
The ratio of the slender canal length over arc radius is greatest for the posterior
semicircular canal (5.10). The ratio for the anterior canal is 4.24, and the ratio for the
lateral canal is 4.00. The aspect ratio of the anterior semicircular canal arc (0.91; height
equals 2.51 mm; width equals 2.77 mm) is higher than the ratio of the arcs of either the
lateral (0.75; height equals 1.81 mm; width equals 2.40 mm) or posterior canal (0.82;
height equals 1.85 mm; width equals 2.25 mm). This result signifies that the arc of the
anterior semicircular canal approaches a perfect circle more so than the other canal arcs.
The plane of the lateral semicircular canal is positioned high with respect to the
posterior semicircular canal in Macroscelides so that the lateral canal divides the space
enclosed by the posterior semicircular canal arc when the bony labyrinth is viewed
anteriorly. The sagittal labyrinthine index for Macroscelides is 32.7, which is greater than
that measured for Chrysochloris. A high index indicates a more dorsal position of the
lateral semicircular canal.
Although Macroscelides holds a sister relationship with afrosoricids in the
supertrees reconstructed by Bininda-Emonds et al. (2007), there are no unambiguous otic
synapomorphies uniting the clade. The lateral semicircular canal is derived relative to the
190
ancestral eutherian state in that it takes a high position relative to the posterior canal, as is
observed in Chrysochloris and Hemicentetes. The afrosoricids and Macroscelides lack
secondary common crura, which are present in Cretaceous eutherians (Meng and Fox,
1995; Chapter 4). Absence of a secondary common crus is derived for Placentalia, and
therefore a feature that the clade consisting of Macroscelidea and Afrosoricida inherited
from its placental ancestor. The anterior semicircular canal is the largest in terms of
radius, indicating that the bony labyrinth of Macroscelides retains the ancestral therian
condition in this regard.
Tubulidentata
Only one extant species, Orycteropus afer, contributes to the group Tubulidentata
(aardvarks), although two additional genera of aardvarks (Leptorycteropus and
Myorycteropus) were present during the Neogene of sub-Saharan Africa (Holroyd and
Mussell, 2005). Morphology does not support any strong systematic placement of
Tubulidentata within Eutheria, either placing it either in a basal polytomy with most
placental lineages (Novacek, 1986; Novacek and Wyss, 1986b) or with weak associations
to ungulates (Novacek, 1992b; Shoshani and McKenna, 1998). Molecular evidence
suggests a close relationship between aardvarks some ungulates (Miyamoto and
Goodman, 1986), particularly Paenungulata (de Jong et al., 1981), which includes
Hyracoidea, Sirenia, and Proboscidea (sensu Simpson, 1945). Both Tubulidentata and
Paenungulata are included within Afrotheria, although the relationships among these taxa
within the African clade are ambiguous (Springer et al., 1997; Eizirik et al., 2001;
Murphy et al., 2001 a, b; Murata et al., 2003; Bininda-Emonds et al., 2007).
The average body mass of Orycteropus afer is significantly greater than any of
the other afrotherians described thus far (50.5 kg; Silva and Downing, 1995), and the
191
mass of the specimen used in this study (AMNH 51909) is above the species average (60
kg). The large body size of Orycteropus is reflected in the bony labyrinth of the inner ear
(total volume equals 107 mm
3
; total length equals 14.95 mm
3
).
The cochlea of Orycteropus contributes a similar volumetric proportion (55.4%;
volume equals 59.3 mm
3
) to that calculated for the ancestor of Afrotheria (56.0%), and
the spiral of the cochlea is fairly flat (Figure 5.14; aspect ratio of 0.45; height equals 4.23
mm; width equals 9.49 mm). The cochlea completes nearly two turns (709°), but the
diameter of the second whorl is smaller than the basal turn (Figure 5.15), and therefore
does not obscure the basal turn when the cochlea is viewed down the axis of rotation. The
total length of the canal is 14.86 mm.
The basal plane of the cochlea is rotated by 31.9° from the plane of the lateral
semicircular canal. The fenestra cochleae opens into the tympanic cavity at the end of a
short stalk. A groove is situated behind the fenestra cochleae (expressed on the endocast
as a curved ridge) that leads to the bony canaliculus cochleae. The canaliculus, which
accommodates the aqueduct of the cochlea in life, is developed in Orycteropus as a
slightly curved tube that widens at the tip. The curve of the 4.82 mm long canaliculus is
best observed in medial view. The secondary lamina that supports the basilar membrane
in life extends between one half and three quarters of the basal turn of the cochlear canal.
A gentle constriction of the vestibule lateral to the fenestra vestibuli (stapedial
ratio of 1.8) divides the spherical and elliptical recesses, although the elliptical recess
within the vestibule of Orycteropus is not divided into anterior and posterior chambers as
is observed within the labyrinth of Chrysochloris. One striking feature of the vestibular
apparatus of Orycteropus is that the lateral and posterior semicircular canals are joined to
form a secondary common crus, a structure that is present in Didelphis (Figure 5.4). The
secondary common crus is flattened, and bony ridges (expressed as thin grooves on the
192
FIGURE 5.14. Bony labyrinth of Orycteropus afer. A, stereopair and labeled line
drawing of digital endocast in anterior view; B, stereopair and labeled line drawing of
digital endocast in dorsal view; C, stereopair and labeled line drawing of digital endocast
in lateral view; D, line drawing of cochlea viewed down axis of rotation to display degree
of coiling; E, line drawing of cochlea in profile. Abbreviations: aa, anterior ampulla; ac,
anterior semicircular canal; ant, anterior direction; av, bony channel for aqueduct of
vestibule; cc, canaliculus cochleae for aqueduct of cochlea; co, cochlea; cr, common
crus; dor, dorsal direction; er, elliptical recess of vestibule; fc, fenestra cochleae; fv,
fenestra vestibuli; la, lateral ampulla; lc, lateral semicircular canal; med, medial
direction; pc, posterior semicircular canal; pl, primary bony lamina; pos, posterior
direction; ps, outpocketing for perilymphatic sac; scr, secondary common crus; sl,
secondary bony lamina; sr, spherical recess of vestibule.
193
A
B
C
D E
co
sl
fv
sr
cc
av
av
aa
fc
la
scr
lc
pc
ac
cr
er
co
pl
cc
av
er
aa
la
scr
fc
pc
lc
ac
fv
fc
ps
5 mm
5 mm
5 mm
cc
co
er
cr
la
aa
lc
pc
ac
scr
co
cc
ps
fc
co
sl
ps
fc
cc
dor
med
med
pos
dor
ant
194
FIGURE 5.15. CT slices through ear region of Orycteropus afer. Abbreviations: aa,
anterior ampulla; ac, anterior semicircular canal; av, bony channel for aqueduct of
vestibule; cc, canaliculus cochleae; cn, canal for cranial nerve VIII; co, cochlea; cr,
common crus; dor, dorsal direction; fc, fenestra cochleae; fn, canal for cranial nerve VII;
fv, fenestra vestibuli; la, lateral ampulla; lc, lateral semicircular canal; med, medial
direction; pa, posterior ampulla; pc, posterior semicircular canal; pl, primary bony
lamina; pos, posterior direction; scr, secondary common crus; sl, secondary bony lamina;
sr, spherical recess of vestibule; vb, vestibule.
195
26
33
40
47
54
61
68
75
82
96
89
26
fn
fn
fn
aa
ac
ac
ac
ac
av
av
av
ac
cn
vb
vb
cn
co
pl
pl
pl
cc
cr
pc
pc
pc
pc
scr
sr
la
lc
cn
co
co
sl
fv
fc
pa
fv
33
4740 54
6861 75
8982 96
pos
med
5 mm
dor
5 mm
med
av
lc
196
endocast) divide the crus into a section for the lateral semicircular canal anteriorly, and
the posterior canal posteriorly, although the two sections are continuous.
The primary common crus between the anterior and posterior canals in
Orycteropus appears squat in relation to the semicircular canal arcs when compared to
the labyrinth of Macroscelides proboscideus (Figure 5.12). The bony channel for the
aqueduct of the vestibule exits medial to the vestibular aperture for the common crus. The
8.25 mm long channel forms a straight tube for a little over half of its length, whereupon
it bifurcates into an anterior and a posterior projection. The anterior projection is wider
than the posterior. The bifurcation of the channel for the aqueduct of the vestibule is
unique to Orycteropus among all of the mammals considered in this study. No mammal
observed by Gray (1907, 1908) possessed a bifurcated membranous aqueduct, and it is
unlikely that the aqueduct of Orycteropus is forked, although the membranous labyrinth
of the species is poorly known (the aardvark was not among the species examined by
Gray).
The planes between the anterior and posterior semicircular canals approach a right
angle (91.9°), and the angle between the posterior and lateral canals are not far from 90°,
either (87.4°). The angle between the anterior and lateral semicircular canal planes is
noticeably acute (78.5°). The lateral semicircular canal is the most planar among the three
with a total angular deviation of 7.2° and a ratio of the total linear deviation over cross-
sectional diameter of the canal of 0.78. The anterior canal deviates most from its plane, at
19.7°, while the posterior canal deviates a total of 11.5° from its plane. The deviation of
both the anterior and posterior semicircular canals are significant with ratios of linear
deviation over canal diameter of 1.82 and 1.06 respectively.
The posterior semicircular canal is larger than the anterior and lateral in terms of
slender canal length (18.86 mm; anterior equals 15.40 mm; lateral equals 16.44 mm),
197
diameter of canal in cross-section (0.66 mm; anterior equals 0.58; lateral equals 0.53),
and arc radius of curvature (3.50 mm; anterior equals 3.10 mm; lateral equals 3.27 mm).
This pattern is not observed in the afrosoricids, nor in Macroscelides, where the anterior
canal tends to be the largest. The dimensions of the anterior semicircular canal of
Orycteropus are the smallest among the three canals, including the aspect ratio of the
anterior semicircular canal arc (0.81). The arc of the lateral semicircular canal approaches
a perfect circle, with an aspect ratio of 1.03, and the height of the posterior semicircular
canal arc is greater than the width (aspect ratio equals 1.28). The ratio between the
slender semicircular canal length and arc radius for the anterior, lateral, and posterior
semicircular canals are 4.96, 5.03, and 5.39 respectively.
The bony labyrinth of Orycteropus possesses two features that are observed in
Kulbeckia and other Cretaceous eutherians (Meng and Fox, 1995; Chapter 4), which are
the secondary common crus and the low position of the lateral semicircular canal relative
to the posterior canal. Such plesiomorphic features support a basal position of
Tubulidentata within Afrotheria, as reconstructed by Eizirik et al. (2001) and Malia et al.
(2002), and are not inconsistent with the relationships recovered by Bininda-Emonds et
al. (2007), in which Tubulidentata is placed in a basal polytomy within Afrotheria.
If Orycteropus is the sister taxon to all other afrotherians, then absence of the
secondary common crus (or separate openings for the lateral and posterior semicircular
canals into the vestibule) might be a synapomorphy uniting the remaining afrotheres.
However, the absence of the secondary common crus is reconstructed as a synapomorphy
for all of Placentalia based on the phylogeny used here (Figure 5.2). Thus, the presence of
the secondary crus is an autapomorphy for Orycteropus. It is interesting, however, that
the bony labyrinth of Orycteropus possesses a few apomorphies (when compared to stem
198
eutherians). For example, the cochlea completes over one and a half turns (almost two),
and the posterior semicircular canal has the largest radius, rather than the anterior canal.
Hyracoidea
A close relationship between Hyracoidea (hyraxes) and ungulates, particularly
either Perissodactyla or Tethytheria (Sirenia + Proboscidea), is a classical hypothesis.
Although some morphological data support a sister relationship between Hyracoidea and
Perissodactyla (Owen, 1848; McKenna, 1975; Fischer and Tassy, 1993), the majority of
morphological (Novacek, 1986; Novacek and Wyss, 1986b; Shoshani, 1986; Novacek
1992a, b; Rasmussen et al., 1990) and molecular data (Shoshani, 1986; Stanhope et al.,
1998; Springer et al., 1999; Murphy et al., 2001a, b; van Dijk et al., 2001; Bininda-
Emonds et al., 2007) support a pairing of Hyracoidea and Tethytheria within the group
Paenungulata. The results of most phylogenetic analyses recover a sister relationship
between Hyracoidea and a monophyletic Tethytheria, but the results of a few recent
analyses, including the supertree constructed by Bininda-Emonds et al. (2007), support a
closer relationship between Hyracoidea and Sirenia, rendering Tethytheria paraphyletic.
A digital endocast of Procavia capensis was constructed (Figure 5.16) to examine
the labyrinth of Hyracoidea. Procavia is a little larger than a house cat in overall body
mass (3.8 kg; Silva and Downing, 1995). The total length of the bony labyrinth of the
hyrax is 8.50 mm, and the total volume of the inner ear cavities is 19.34 mm
3
, of which
the cochlea contributes 47.7% (9.22 mm
3
). The percentage of volume of the bony
labyrinth that is made up by the cochlea is relatively low among the afrotherians
investigated, although larger than that of the elephantoid (the cochlea of which comprises
30.6 %, as described below).
199
FIGURE 5.16. Bony labyrinth of Procavia capensis. A, stereopair and labeled line
drawing of digital endocast in anterior view; B, stereopair and labeled line drawing of
digital endocast in dorsal view; C, stereopair and labeled line drawing of digital endocast
in lateral view; D, line drawing of cochlea viewed down axis of rotation to display degree
of coiling; E, line drawing of cochlea in profile. Abbreviations: aa, anterior ampulla; ac,
anterior semicircular canal; ant, anterior direction; av, bony channel for aqueduct of
vestibule; cc, canaliculus cochleae for aqueduct of cochlea; co, cochlea; cr, common
crus; dor, dorsal direction; er, elliptical recess of vestibule; fc, fenestra cochleae; fv,
fenestra vestibuli; la, lateral ampulla; lc, lateral semicircular canal; med, medial
direction; pa, posterior ampulla; pc, posterior semicircular canal; pl, primary bony
lamina; pos, posterior direction; ps, outpocketing for perilymphatic sac; sl, secondary
bony lamina; sr, spherical recess of vestibule.
200
co
1 mm
1 mm
1 mm
fc
cc
fv
sr
er
aa
la
cr
ac
pc
av
pa
lc
sl
co
sl
pl
er
aa
la
ac
lc
cr
pa
av
pc
co
fv
fc
cc
sl
la
aa
cr
lc
ac
pc
av
pa
fc
sl
co
ps
co
sl
fc
ps
A
B
C
D E
dor
med
pos
med
dor
ant
201
The aspect ratio of the cochlear spiral is 0.72, which is large when compared to
most of the other afrotherians examined, although the aspect ratio of Macroscelides
proboscideus is larger (0.80). However, the overall shape of the cochlea is different
between Procavia and Macroscelides. Whereas the apical turn sits upon the basal turn of
the cochlea in Macroscelides, the apical turns of Procavia fit within the arc formed by the
basal turn. Such morphology gives the cochlea of Procavia a conical appearance, rather
than the cylindrical shape of Macroscelides and Chrysochloris.
The cochlea of Procavia completes over three and three quarters turns (1363°), a
greater degree than in any other afrotherian. The secondary lamina of Procavia (Figure
5.17) is not as well developed as the structure is in Macroscelides, and only extends a
short distance past the first half of the basal turn of the cochlea. The plane of the basal
turn deviates from that of the lateral canal by 45.4°, which is greater than all other
afrotherians except the elephantoid (48.5°). The degree of rotation exhibited by the
cochlea might be a synapomorphy for Paenungulata, although the cochlea of Trichechus
only deviates from the plane of the lateral canal by 27.7°. The bony canaliculus cochleae
for transmission of the cochlear aqueduct (1.21 mm) is very slender and is not observed
easily in the CT slices, having a maximum diameter of a single pixel (around 0.07 mm) at
several points along its path. The canaliculus is not a straight tube, but rather hooks
laterally. The canaliculus cochleae exits the bony labyrinth from a bulge posteromedial to
the fenestra cochleae.
The spherical recess of the vestibule is distinguished easily from the elliptical
recess particularly at the anterior aspect of the vestibule, dorsal to the fenestra vestibuli
(stapedial ratio is 2.1). Despite its name, the spherical recess is ovoid in shape. However,
the elliptical recess is much more elongated, and it gently curves laterally. The anterior
end of the elliptical recess opens into the anterior and lateral ampullae, and the posterior
202
FIGURE 5.17. CT slices through ear region of Procavia capensis. Abbreviations: aa,
anterior ampulla; ac, anterior semicircular canal; av, bony channel for aqueduct of
vestibule; cn, canal for cranial nerve VIII; co, cochlea; cr, common crus; dor, dorsal
direction; fc, fenestra cochleae; fn, canal for cranial nerve VII; fv, fenestra vestibuli; la,
lateral ampulla; lc, lateral semicircular canal; med, medial direction; pa, posterior
ampulla; pc, posterior semicircular canal; pl, primary bony lamina; pos, posterior
direction; vb, vestibule; ven, ventral direction.
203
60
70
80
90
100
110
120
130
140
160
150
pos
60 70
9080 100
120110 130
150140 160
dor
med
co
co
co
co
pl
pl
cn
ac
aa
fn
fn
fv
fc
vb
vb
vb
lc
cn
cn
vb
ven
1 mm
1 mm
pl
pl
fc
ac
lc
lc
la
lc
ac
ac
ac
cr
lc
lc
pc
pc
pc
pc
pc
pc
av
av
pa
lc
lc
204
end of the recess leads to the vestibular apertures of the posterior ampulla, common crus,
and posterior limb of the lateral semicircular canal. The lateral canal enters the vestibule
closer to the common crus than it does to the posterior ampulla, which causes the lateral
semicircular canal to divide the space enclosed by the arc of the posterior semicircular
canal into dorsal and ventral regions in anterior view. The sagittal labyrinthine index for
Procavia is 44.9, which is over twice that observed in Chrysochloris (21.7), and greater
than that in Macroscelides (32.7).
The bony channel for the aqueduct of the vestibule exits medial and ventral to the
vestibular aperture of the common crus, and it is 3.39 mm in length, almost three times
longer than the canaliculus cochleae. The channel curves laterally along the posterior
border of the base of the common crus, but the aqueduct does not cross the rise of the
common crus when the bony labyrinth is viewed medially. The channel for the aqueduct
is a uniformly subcircular tube in cross-section, until it flares and flattens into a fissure
nearly on the plane of the posterior semicircular canal arc.
The greatest angle between the planes of two semicircular canals in Procavia was
measured between the anterior and posterior canals (112°). The angle between the
anterior and lateral canals is almost as acute as the angle between the lateral and posterior
canals is obtuse (87.4° and 94.9° respectively). No canal fits onto a single plane in
Procavia, although the angular deviation is not great for any canal. The anterior canal
shows the largest angular deviation of a canal from its plane (7.79°), although this is not
much different than the deviations of the lateral (5.78°) and posterior canals (6.06°). Even
so, only the deviation of the anterior canal is considered significant, with the ratio of total
linear deviation of the anterior canal over the cross-sectional diameter of the canal
equaling 1.28 (ratios for the lateral and posterior canals are 0.54 and 0.88 respectively).
205
As in Orycteropus afer, the posterior semicircular canal of Procavia is the most
voluminous (0.41 mm
3
), as well as the longest of the three canals (10.68 mm). The arc
radius of curvature of the posterior semicircular canal arc is greater for the posterior canal
(2.18 mm) than for either the anterior (1.99 mm) or lateral semicircular canals (1.79 mm).
The slender anterior canal of Procavia is 10.24 mm long, whereas the length of the
slender lateral canal is 7.65 mm. Both the anterior and lateral semicircular canals have a
volume of 0.37 mm
3
each. Among afrotherians, the posterior canal is the largest only in
Procavia and Orycteropus and is a potential synapomorphy uniting the two taxa. A sister
relationship between aardvarks and hyraxes has not been proposed, although Hyracoidea
has been placed in a polytomy along with Tubulidentata and Sirenia (sister taxon to
polytomy is Proboscidea; de Jong et al., 1981).
The arcs of the three semicircular canals are wider than they are high, and the
aspect ratio of the arcs of the anterior and lateral canals are similar (0.68 and 0.72
respectively). The aspect ratio of the posterior semicircular canal arc is higher at 0.79.
The ratio of the slender canal length over semicircular canal arc radius is greatest for the
anterior canal, at 5.14 (lateral equals 4.28; posterior equals 4.90).
There is no unambiguous support for paenungulate monophyly in the bony
labyrinth. Hyraxes retain the ancestral placental condition in that the secondary common
crus is absent, but are derived from the placental ancestor in that the posterior canal is
largest in terms of arc radius, rather than the anterior canal. Among the remaining
afrotherians, only Orycteropus shares the condition of having the largest arc in the
posterior semicircular canal. The labyrinth of Procavia is derived from the ancestral
eutherian condition in that the lateral semicircular canal is positioned dorsal with respect
to the ampullar entrance of the posterior semicircular canal.
206
Sirenia
Dugongs and manatees comprise the clade Sirenia. Sirenia and Cetacea are the
two exclusively aquatic groups of extant mammals (the other is Cetacea), although they
are not closely related despite similar lifestyles and superficial resemblances, such as a
fusiform body and short neck. Rather, there is a closer connection between Sirenia and
Proboscidea, which is a relationship that has been recognized for several centuries
(Linnaeus, 1758; Simpson, 1945; McKenna, 1975).
Monophyly of Tethytheria, which is the clade that includes Sirenia, Proboscidea,
as well as the extinct groups Desmostylia, “Anthrocobunidae”, and Embrithopoda
(McKenna, 1975; Gheerbrant et al., 2005), is supported by more recent morphological
(Novacek, 1986; Novacek and Wyss, 1986b; Court, 1990; Fischer and Tassy, 1993) and
molecular evidence (Lavergne et al., 1996; Murphy et al., 2001a). However, the results of
a few recent molecular analyses (Murphy et al., 2001b; Amrine-Madsen et al., 2003;
Bininda-Emonds et al., 2007) refute tethytherian monophyly, while still recovering a
close relationship between Sirenia and Proboscidea within Paenungulata.
The Florida manatee, Trichechus manatus, represents Sirenia. The manatee has an
average species body mass of 500 kg (Silva and Downing, 1995), and this is reflected in
the total length of the bony labyrinth (19.29 mm) of the Trichechus specimen used, as
well as in total volume of the inner ear cavities (1069.4 mm
3
).
The most notable feature observable on the digital endocast of Trichechus is the
absence of the bony canaliculus cochleae for transmission of the aqueduct of the cochlea
(Figure 5.18). Rather, the canaliculus and fenestra cochleae are fused to form an
undivided perilymphatic foramen, which is unique to Sirenia and Proboscidea among
extant mammals. Although the three living species of manatees and the dugong possess
an undivided perlymphatic foramen, the bony canaliculus cochleae is separate from the
207
FIGURE 5.18. Bony labyrinth of Trichechus manatus. A, stereopair and labeled line
drawing of digital endocast in anterior view; B, stereopair and labeled line drawing of
digital endocast in dorsal view; C, stereopair and labeled line drawing of digital endocast
in lateral view; D, line drawing of cochlea viewed down axis of rotation to display degree
of coiling; E, line drawing of cochlea in profile. Abbreviations: aa, anterior ampulla; ac,
anterior semicircular canal; ant, anterior direction; av, bony channel for aqueduct of
vestibule; co, cochlea; cr, common crus; dor, dorsal direction; fv, fenestra vestibuli; la,
lateral ampulla; lc, lateral semicircular canal; med, medial direction; pa, posterior
ampulla; pc, posterior semicircular canal; pf, perilymphatic foramen; pl, primary bony
lamina; pos, posterior direction; vb, vestibule.
208
5 mm
5 mm
5 mm
A
B
C
D E
co
pf
fv
av
cr
ac
pc
lc
aa
la
pa
vb
co
pl
pf
av
av
fv
la
lc
aa
pa
pc
ac
co
vb
la
aa
cr
ac
pc
pa
lc
co
co
pf
dor
med
pos
med
dor
ant
209
fenestra cochleae in the Eocene sirenian Prorastamus, suggesting that the undivided
perilymphatic foramen either evolved independently in Sirenia and Proboscidea (Court,
1990, 1992a), or a reversal in the Eocene sirenian.
The cochlea is large (441.6 mm
3
) with respect to the entire bony labyrinth,
contributing 71.1% of the entire labyrinthine volume. Although this is a greater
contribution than the cochleae of either the ancestral afrotherian (56.0%) or paenungulate
(48.0%), the dimension is not much different than the volumetric percentages calculated
for Chrysochloris sp. (71.3%) and Macroscelides proboscideus (71.7%). However, the
aspect ratio of the spiral of the cochlea (0.55) is lower in Trichechus than in either
Chrysochloris (0.63) or Macroscelides (0.80). The cochlea of Trichechus completes just
over a single turn (407°), which is less than any other afrotherian examined, and almost a
complete turn lower than that calculated for the ancestor of Afrotheria (751°). A low
degree of coiling may be a synapomorphy for Sirenia, given that a similar degree of
coiling is observed in the fossil Hydrodamalis gigas (Kaiser, 1974). The length of the
canal from the base to the apex in Trichechus is 22.46 mm, and the plane of the basal turn
of the cochlea deviates from that of the lateral semicircular canal by 27.7°. As in
proboscideans, the secondary lamina is not present in Trichechus (Figure 5.19).
The fenestra vestibuli has an aspect (stapedial) ratio of 1.6, which signifies a less
elliptical window than other afrotherians. The spherical recess of the vestibule, which
communicates with the tympanic cavity via the fenestra vestibuli, is poorly developed,
and not distinguishable from the elliptical recess. In fact, the vestibule as a whole is
mediolaterally compressed as can be seen when the bony labyrinth is in anterior view.
The thickest part of the vestibule is at the anterior end, where a slight laterodorsal
projection leads to the anterior and lateral ampullae. The ampullae of the semicircular
210
FIGURE 5.19. CT slices through ear region of Trichechus manatus. Abbreviations: ac,
anterior semicircular canal; av, bony channel for aqueduct of vestibule; cn, canal for
cranial nerve VIII; co, cochlea; cr, common crus; dor, dorsal direction; fv, fenestra
vestibuli; la, lateral ampulla; lat, lateral direction; lc, lateral semicircular canal; med,
medial direction; pa, posterior ampulla; pc, posterior semicircular canal; pf,
perilymphatic foramen; pl, primary bony lamina; pos, posterior direction; vb, vestibule;
ven, ventral direction.
211
59
66
73
80
87
94
101
108
115
129
122
59
co
co
pf
cn
pl
pl
pa
pc
pc
ac
ac
lc
lc
la
cr cr
lc
av
av
co
66
8073 87
10194 108
122115 129
pos
med
lat
dor
1 mm
ac
pf
cn
pl
pf
cn
vb
vb
fv
212
canals are proportionately smaller in Trichechus than in other taxa examined, such as
Macroscelides and Procavia.
Apertures for the posterior ampulla, common crus, and the posterior limb of the
lateral semicircular canal are situated at the posterior end of the vestibule, with the
common crus as the medial-most opening. The bony channel for the aqueduct of the
vestibule exits the bony labyrinth ventromedial to the vestibular aperture of the common
crus. The channel for the aqueduct extends from the vestibule as a round tube for a very
short distance before opening into a broad fissure that flares posterodorsally. The bony
channel is 12.33 mm in length, which is nearly two thirds as long as the total length of the
bony labyrinth (19.29 mm).
The vestibular aperture of the lateral semicircular canal opens near the base of the
posterior ampulla in Trichechus, similar to the state observed in Macroscelides. However,
the lateral canal enters the vestibule lateral and ventral to the posterior ampulla in
Trichechus, which is on the opposite side of the posterior ampulla from Macroscelides
and other taxa where the opening for the canal is well separated from the ampulla, such
as Procavia. Even in Orycteropus, where the lateral and posterior canals fuse to form a
secondary common crus, the lateral canal is situated dorsal and slightly medial to the
posterior canal. The morphology observed in Trichechus places the plane of the lateral
semicircular canal relatively low on the vestibule.
All of the planes of the semicircular canals of Trichechus form acute angles with
each other. The angle between the planes of the anterior and lateral semicircular canals
(52.2°) is the smallest angle measured between any two canals in any afrotherian
specimen. Within Trichechus, the only canals that approach a right angle are the anterior
and posterior canals (84.9°). The angle between the posterior and lateral semicircular
canal planes is 77.5°. The posterior semicircular canal does not deviate from its plane,
213
and the anterior canal is more planar than the lateral canal, with total angular deviations
of 7.86° and 8.87° respectively. The deviations exhibited by both the anterior and lateral
semicircular canals are significant (ratios of total linear deviation over cross-sectional
diameter are 1.17 and 1.33 respectively).
No single semicircular canal within the bony labyrinth of Trichechus is the
greatest in all dimensions measured. The radius of the arc of the lateral semicircular canal
(4.46 mm) is the larger than the arc of the posterior canal by nearly a millimeter (3.54
mm), but only slightly larger than the arc of the anterior canal (4.30 mm). Both the
diameter of the lumen of the lateral semicircular canal (0.52 mm), as well as the volume
of the canal (2.5 mm
3
) are the largest among the three canals (dimensions for the anterior
canal equal 0.51 mm and 2.3 mm
3
; dimensions for the posterior canal equal 0.50 mm and
1.8 mm
3
). However, the length of the slender canal of the lateral semicircular canal
(14.20 mm) is noticeably smaller than both the anterior (17.31 mm) and posterior (16.53
mm) semicircular canals.
The posterior limbs of the anterior and lateral semicircular canals form steeper
slopes than the anterior limbs. Although the anterior canal is curved along its entire
course, the anterior limb of the lateral canal is flattened, giving the arc of the canal an
angular appearance at its midpoint, before the posterior limb curves gradually to meet the
vestibule. The arc of the posterior semicircular canal is noticeably higher than the other
two canals, with an aspect ratio of 1.18. The aspect ratio of the anterior canal arc is 0.91,
which is similar to that of the lateral canal (0.89). The ratio of the slender canal length
over semicircular canal arc radius is greatest for the posterior canal (4.67), followed by
the anterior canal (4.02). The ratio for the lateral canal equals 3.18.
Although the results of several recent molecular analyses, including those of
Bininda-Emonds et al. (2007), do not support the monophyly of Tethytheria, the structure
214
of the inner ear supports a sister relationship between Sirenia and Proboscidea among the
paenungulates. Notable labyrinthine features that are shared by Sirenia and Proboscidea
within Paenungulata are a low position of the lateral semicircular canal compared to the
posterior canal (the lateral canal is high in Procavia), and a low cochlear spiral. The large
radius of the lateral semicircular canal is an autapomorphy for Trichechus compared to
all other afrotherians. The bony labyrinth of Trichechus retains the ancestral placental
condition of the lateral semicircular canal opening directly into the vestibule in the
absence of a secondary common crus.
Proboscidea
The bony labyrinth of a specimen of he extinct elephantoid was described
elsewhere (Chapter 2), but for the sake of comparison, a brief overview of the inner ear
anatomy of Proboscidea is provided here. Not only are proboscideans the largest
afrotherians, as is reflected in the volume (1145.2 mm
3
) and length (26.00 mm) of the
inner ear, they are the largest extant terrestrial mammal. Because the species of the
proboscidean used for this study is not known with certainty (see Chapter 2), the body
mass of the individual could not be estimated. Extant Proboscidea is not a taxonomically
diverse clade, with no more than three species (Wilson and Reeder, 2005; Reeder et al.,
2007), but proboscidean diversity was much greater throughout the Tertiary period
(McKenna and Bell, 1997).
As mentioned in the description of the bony labyrinth of the sirenian Trichechus
manatus, the canaliculus cochleae for the cochlear aqueduct is absent in the elephantoid
(Figure 5.20), which is an apomorphic condition for Tethytheria. Rather, both the
elephantoid and Trichechus share a secondarily undivided perilymphatic foramen in lieu
of a fenestra cochleae (although this condition may have an independent derivation in
215
FIGURE 5.20. Bony labyrinth of the fossil elephantoid proboscidean. A, stereopair and
labeled line drawing of digital endocast in anterior view; B, stereopair and labeled line
drawing of digital endocast in dorsal view; C, stereopair and labeled line drawing of
digital endocast in lateral view; D, line drawing of cochlea viewed down axis of rotation
to display degree of coiling; E, line drawing of cochlea in profile. Abbreviations: aa,
anterior ampulla; ac, anterior semicircular canal; ant, anterior direction; av, bony channel
for aqueduct of vestibule; co, cochlea; cr, common crus; dor, dorsal direction; er,
elliptical recess of vestibule; fv, fenestra vestibuli; la, lateral ampulla; lc, lateral
semicircular canal; med, medial direction; pa, posterior ampulla; pc, posterior
semicircular canal; pf, perilymphatic foramen; pos, posterior direction; ps, outpocketing
for perilymphatic sac; sr, spherical recess of vestibule; vb, vestibule.
216
A
B
C
D E
av
ac
pc
lc
la
aa
sr
er
cr
5 mm
5 mm
5 mm
co
fv
pf
co
av
vb
cr
aa
la
lc
pc
ac
co
fv
pf
ps
lc
la
aa
ac
cr
av
pc
pa
er
co
pf
co
ps
pf
dor
med
pos
med
dor
ant
217
both clades; see Court 1990, 1992a). The stapedial ratio measured from the fenestra
vestibuli of the elephantoid is 1.60, which is similar to that calculated for Trichechus
(1.64). A round fenestra also is characteristic of the extinct embrythopod Arsinotherium,
which is closely related to Tethytheria (McKenna and Bell, 1997; Gheerbrant et al.,
2005), if not within Tethytheria itself as the sister taxon of Proboscidea (Court, 1992a).
The cochlea of the elephantoid completes a little over two whorls (765°) and
contributes only 30.6% of the total volume of the inner ear, which is the smallest
contribution observed among the afrotherian sample investigated. Although the total
volume of the inner ear is greater in the elephantoid than any of the other afrotherians, the
volume of the cochlea (350.7 mm
3
) is less than that measured for Trichechus (441.6
mm
3
). The secondary lamina is not developed in Proboscidea (Figure 5.21), and the
cochlea is fairly planispiral in the elephantoid with an aspect ratio of 0.42. The only other
afrotherian to have a lower aspect ratio is Hemicentetes (0.38). The aspect ratios of the
cochleae of the other paenungulates are greater than that of the elephantoid (Trichechus is
0.55 and Procavia is 0.72).
The vestibular aperture of the posterior limb of the lateral semicircular canal is
situated anterior to the posterior ampulla in the elephantoid. The bony channel for the
aqueduct of the vestibule leaves the vestibule medial to the common crus, and the
aqueduct extends 13.94 mm before exiting the petrosal via a fissure on the endocranial
surface of the bone. The posterior limb of the lateral semicircular canal enters the
vestibule separately from the posterior ampulla.
The angle between the planes of the basal turn of the cochlea and the lateral
semicircular canal is greater for the elephantoid (48.5°) than any other afrotherian. The
most acute angle between the planes of two semicircular canals is between the anterior
and lateral canals (66.3°), and the most obtuse angle was measured between the posterior
218
FIGURE 5.21. CT slices through ear region of the fossil elephantoid proboscidean.
Abbreviations: aa, anterior ampulla; ac, anterior semicircular canal; av, bony channel for
aqueduct of vestibule; cn, canal for cranial nerve VIII; co, cochlea; cr, common crus;
dor, dorsal direction; fn, canal for cranial nerve VII; fv, fenestra vestibuli; la, lateral
ampulla; lat, lateral direction; lc, lateral semicircular canal; pc, posterior semicircular
canal; pf, perilymphatic foramen; pos, posterior direction; vb, vestibule.
219
150
172
194
216
238
260
282
304
326
370
348
dor
cn
cn
cn
cn
co
fn
fn
fn
pc
pc
pc
cr
fn
fn
1 mm
lat
pos
lat
5 mm
150 172
216194 238
282260 304
348326 370
co
fn
co
co
fv
pf
vb
fv
la
lc
ac
ac
pc
cr
av
av
pf
aa
vb
av
av
ac
ac
220
and lateral canals (92.6°). The angle between the planes of the anterior and posterior
semicircular canals is 73.7°. Although the canals deviate from their planes, the deviation
is not significant (ratio of total linear deviation over cross-sectional area for the anterior
canal equals 0.87; lateral canal equals 0.08; posterior canal equals 0.71). The total
angular deviations of the anterior, lateral, and posterior semicircular canals are 18.5°,
3.0°, and 14.3° respectively.
The arc radius of the posterior semicircular canal (5.51 mm) is larger than both
the anterior (4.99 mm) and lateral canals (2.67 mm), as is the diameter of the posterior
canal in cross-section (1.91 mm; anterior equals 1.85 mm; lateral equals 1.69 mm).
However, the length of the slender canal of the anterior semicircular canal (24.57 mm) is
greater than either the posterior (24.28 mm) or lateral canals (12.54 mm).
The arcs of the lateral and posterior semicircular canals are higher than they are
wide in the elephantoid with aspect ratios of 1.31 and 1.10 respectively. The aspect ratio
of the arc of the lateral semicircular canal is less (0.72). The ratios between the length of
the slender canal and arc radius of the anterior canal is 4.93, which is the largest ratio
among the three canals, and 4.41 for the posterior canal, which is the smallest value. The
ratio for the lateral canal is 4.70.
The bony labyrinth of the elephantoid retains the primitive eutherian morphology
observed in Kulbeckia, although the mammoth inner ear is derived in the absence of the
secondary common crus. There are no unambiguous characters within the bony labyrinth
to support monophyly of Tethytheria to the exclusion of all other afrotherians, although
among the paenungulates, both the elephantoid and Trichechus share a flattened cochlea
and a low position of the lateral semicircular canals (which are both the ancestral
condition for Eutheria), although the ancestral paenungulate state for both of these
characters is equivocal.
221
Xenarthra
There are two major groups of xenarthrans, the armadillos and extinct glyptodonts
that belong to Cingulata, and the anteaters and sloths, which make up the clade Pilosa
(McKenna and Bell, 1997). Xenarthra often occupies a basal position in placental
mammal phylogenies reconstructed using both morphological (e.g., Novacek and Wyss,
1986b) and molecular (e.g., Murphy et al., 2001a, b) analyses.
A close relationship between Xenarthra and Pholidota (pangolins) within a group
called Edentata has been proposed (as recently as the mid-1980’s by Novacek, 1986)
based on morphology, but such anatomical similarities, which include adaptations for a
fossorial lifestyle and a reduction in teeth, are considered homoplastic in more recent
phylogenetic analyses, including Bininda-Emonds et al. (2007). The validity of Edentata
as a natural grouping was discussed by Rose et al. (2005). Further, nearly all molecular
analyses group Pholidota with other placental clades separate from Xenarthra (e.g.,
Miyamoto and Goodman, 1986; Honeycutt and Adkins, 1993; Springer et al., 1997;
Shoshani and McKenna, 1998; Liu et al., 2001; Murphy et al., 2001a, b; van Dijk et al.,
2001; Amrine-Madsen et al., 2003; Bininda-Emonds, 2007; one exception is McKenna,
1992, which recovers a sister relationship between Xenarthra and Pholidota).
The nine-banded armadillo, Dasypus novemcinctus, which is the only xenarthran
found in the United States, represents Xenarthra in this study. Dasypus as a genus is
known from the Pliocene to Recent in North, Central, and South America (McKenna and
Bell, 1997), and D. novemcinctus itself has the largest biogeographical distribution of any
xenarthran species (McBee and Baker, 1982).
Intraspecific variation within the inner ear of D. novemcinctus was discussed in
Chapter 4, and a more thorough description of the bony labyrinth of this species is
provided here (Figures 5.22-5.23). Dimensions of the bony labyrinth of Dasypus are
222
FIGURE 5.22. Bony labyrinth of Dasypus novemcinctus. A, stereopair and labeled line
drawing of digital endocast in anterior view; B, stereopair and labeled line drawing of
digital endocast in dorsal view; C, stereopair and labeled line drawing of digital endocast
in lateral view; D, line drawing of cochlea viewed down axis of rotation to display degree
of coiling; E, line drawing of cochlea in profile. Abbreviations: aa, anterior ampulla; ac,
anterior semicircular canal; ant, anterior direction; av, bony channel for aqueduct of
vestibule; cc, canaliculus cochleae for aqueduct of cochlea; co, cochlea; cr, common
crus; dor, dorsal direction; er, elliptical recess of vestibule; fc, fenestra cochleae; fv,
fenestra vestibuli; la, lateral ampulla; lc, lateral semicircular canal; med, medial
direction; pa, posterior ampulla; pc, posterior semicircular canal; pos, posterior direction;
ps, outpocketing for perilymphatic sac; sl, secondary bony lamina; sr, spherical recess of
vestibule.
223
A
B
C
D E
co
sl
sr
aa
la
pa
lc
ac
pc
cr
fv
fc
er
sl
co
cc
av
pc
cr
er
aa
la
ac
fc
lc
fc
cc
av
cr
aa
ac
fv
co
sl
sr
la
lc
ps
pc
pa
co
fc
ps
co
sl
ps
1 mm
1 mm
1 mm
fc
cc
dor
med
pos
med
dor
ant
224
FIGURE 5.23. CT slices through ear region of Dasypus novemcinctus. Abbreviations: aa,
anterior ampulla; ac, anterior semicircular canal; av, bony channel for aqueduct of
vestibule; cc, canaliculus cochleae; cf, foramina within cribriform plate; cn, canal for
cranial nerve VIII; co, cochlea; cr, common crus; dor, dorsal direction; fc, fenestra
cochleae; fn, canal for cranial nerve VII; fv, fenestra vestibuli; la, lateral ampulla; lc,
lateral semicircular canal; med, medial; pc, posterior semicircular canal; pl, primary
bony lamina; pos, posterior direction; sl, secondary bony lamina; vb, vestibule.
225
144
173
202
231
260
289
318
347
376
434
405
dor
ac
pc
ac
av
pa
cc
cc
cf
co
fc
cn
cn
pl
sl
sl
pl pl
fn
fn
cr
lc la
vb
vb
vb
pc
ac
ac
aa
lc
cr
1 mm
med
pos
med
1 mm
144 173
231202 260
318289 347
405376 434
pc
lc
lc
fv
pc
pl pl
pl
ac
co
sl
226
provided in Table 5.2. The total length of the bony labyrinth is 8.06 mm, and the volume
of the combined inner ear cavities is 26.48 mm
3
. The cochlea itself contributes 17.48
mm
3
(see Table 5.3) to the volume of the bony labyrinth (66.0%), which is larger than
that reconstructed for the ancestors of both Placentalia (58.0%) and Afrotheria (56.0%).
Further dimensions of the cochlea are provided in Table 5.3, and dimensions and
orientations of the semicircular canals are reported in Tables 5.4-5.6. Average body mass
for the species is 4.8 kg (Silva and Downing, 1995).
The cochlea completes nearly two and a quarter turns (816.3°), and the total
length of the cochlear canal is 11.21 mm. The diameters of the apical whorls of the
cochlea are smaller than the basal turn of the cochlea (unlike the condition observed in
Macroscelides), although the successive whorls sit upon the basal turn (Figure 5.22). The
aspect ratio of the cochlear spiral in profile is 0.63, which is the same that was calculated
for the afrosoricid Chrysochloris.
As was observed within the cochlea of Macroscelides, the secondary bony lamina
is well developed in Dasypus (Figure 5.23), and the structure extends past the basal turn
(383.1°). The lamina is expressed as the distinct groove along the radial wall of the
digital endocast. The secondary lamina curves around the dorsal border of the fenestra
cochleae, and it defines the posterior border of an inflation of the scala vestibuli between
to the fenestra cochleae and fenestra vestibuli (stapedial ratio equals 1.74). Medial to the
fenestra cochleae, an outpocketing of the scala tympani for the perilymphatic duct leads
to a robust canaliculus cochleae for transmission of the aqueduct of the cochlea (1.17 mm
in length).
The angle between the plane of the basal turn of the cochlea and the lateral
semicircular canal is low for Dasypus (17.9°) when compared to other taxa that are
227
described above. The angle is 18.4° in Hemicentetes, but the cochlea deviates from the
plane of the lateral canal to a greater degree in all other taxa discussed so far.
The spherical recess of the vestibule is distinguishable from the elliptical recess as
the former bulges medially toward the axis of rotation of the cochlea. The anterior and
lateral ampullae open into a small anterior chamber of the elliptical recess (expressed on
the endocast as a short pedestal). The posterior limb of the lateral semicircular canal
opens into the vestibule dorsal to the posterior ampulla, which gives the lateral canal a
high position with respect to the rest of the vestibule. The lateral canal divides the space
enclosed by the arc of the posterior semicircular canal when the labyrinth is in anterior
view, and the sagittal labyrinthine index is 23.0. This index is slightly larger than that
calculated for Chrysochloris (21.7).
The common crus appears stout with respect to the arcs of the semicircular canals.
The bony channel for the aqueduct of the vestibule exits the labyrinth from a triangular
excavation on the medial wall of the spherical recess, medioventral to the common crus.
The canal for the aqueduct of the vestibule is longer than the canaliculus cochleae (2.63
mm versus 1.17 mm), but is a more delicate structure overall. The aqueduct only crosses
the base of the common crus on its posterodorsal course.
All of the semicircular canal planes of Dasypus form acute angles with each other,
although the angle between the planes of the posterior and lateral canals approaches a
right angle (87.3°). The plane of the anterior semicircular canal forms an angle of 62.4°
with the plane of the lateral canal, and an angle of 67.7° with the plane of the posterior
canal. The posterior semicircular canal is the most planar of the three, with a total angular
deviation of 7.8° from its plane, and the lateral canal is the least planar (18.1°). The
angular deviation of the anterior semicircular canal from its plane is 13.0°, and the ratio
228
of the total linear deviation over cross-sectional diameter is significant for all three canals
(anterior is 1.68; lateral is 2.13; posterior is 1.18).
The posterior semicircular canal is the largest of the three canals in terms of
slender canal length (11.30 mm; anterior equals 9.69 mm; lateral equals 7.38 mm) and
arc radius of curvature (1.92 mm; anterior equals 1.64 mm; lateral equals 1.60 mm).
However, the diameter of the lumen of the lateral semicircular canal (0.24 mm) is greater
than either the anterior or posterior canal (each with a diameter of 0.22 mm).
The area enclosed by the arc of the anterior semicircular canal is elliptical, as
expressed by the aspect ratio of the arc (0.58), whereas the area enclosed by the arc of the
posterior semicircular canal is circular (aspect ratio equals 0.96). The aspect ratio of the
arc of the lateral semicircular canal is 1.16, signifying that the height of the arc is greater
than the width. The ratio of the length of the slender semicircular canal over the arc
radius is 5.91 for the anterior semicircular canal, 4.63 for the lateral semicircular canal,
and 5.88 for the posterior semicircular canal.
The bony labyrinth of Dasypus is derived in all respects to that of the ancestral
eutherian, but retains the direct vestibular entry of the lateral semicircular canal from its
placental ancestor. The plane of the lateral canal is high relative to the ampullar entry of
the posterior canal, which is derived with respect to Eutheria, but the ancestral placental
condition is unknown. Furthermore, the posterior semicircular canal of Dasypus is largest
in terms of arc radius, rather than the anterior canal arc, and the aspect ratio of the
cochlear spiral is high, giving the cochlea a “sharp-pointed” appearance. Although the
labyrinth of Dasypus is derived with respect to the eutherian ancestral condition, there are
no unambiguous characters within the labyrinth to support a closer relationship between
Xenarthra and either Afrotheria or Boreoeutheria. The cochlea of Dasypus coils to a
greater degree than the ancestor of Placentalia (816.3° versus 738.2°), and the cochlea
229
contributes a larger percentage of the total inner ear volume than the placental ancestor
(66.0% versus 58.0%).
Boreoeutheria
The non-afrotherian and non-xenarthran placentals, or Boreoeutheria, are divided
into two sister clades, the Euarchontoglires and Laurasiatheria (Figure 5.2). The lateral
semicircular canal of the ancestral boreoeutherian entered the vestibule directly without
forming a secondary common crus with the posterior semicircular canal (a state inherited
from the placental ancestor), and the arc of the anterior semicircular canal was the largest
among the three arcs, which is a state retained from the therian ancestor. The plane of the
lateral semicircular canal was positioned high compared to the ampullar opening of the
posterior semicircular canal, which is derived from the ancestors of both Theria and
Eutheria, although the state in the placental ancestor is unknown. A high position of the
lateral canal in Boreoeutheria is shared with Dasypus, which might support a sister
relationship between Xenarthra and Boreoeutheria, but the ancestral state in Afrotheria
could not be reconstructed unequivocally. Owing to variation of the aspect ratio of the
cochlear spiral within Laurasiatheria and Euarchontoglires, the condition for the ancestor
of Boreoeutheria is equivocal between the high and low conditions.
The degree of coiling of the ancestor of Boreoeutheria (815.4°) is almost identical
to that of Dasypus (816.3°), both of which are greater than that reconstructed for
Afrotheria (751.3°). Such a degree of coiling might support a Xenarthra plus
Boreoeutheria pairing. However, the volumetric contribution of the cochlea to the entire
labyrinth of Boreoeutheria (55%) nearly is identical that reconstructed for Afrotheria
(56%), both of which are less than that calculated for Dasypus (66%).
230
Laurasiatheria
Laurasiatheria encompasses a great diversity of placental mammals in terms of
body size, ranging from the smallest extant mammal, the hog-nosed bat (Craseonycteris
thonglongyai) at around 2 g, to the largest, the blue whale (Balaenoptera musculus) at
around 150000 kg (Silva and Downing, 1995). Some laurasiatherians are specialized for
efficient cursoriality, such as the cheetah (Acinonyx jubatus) or Thomson’s gazelle
(Eudorcas thomsoni), while others are adapted for fossorial lifestyles, such as the
European mole (Talpa europaea). Furthermore, volant bats and fully aquatic cetaceans
are included within Laurasiatheria. As a whole, the clade Laurasiatheria is composed of
Cetartiodactyla (represented here by Sus scrofa, Bathygenys reeves, Tursiops truncatus,
and an extinct member of Balaenopteridae), Perissodactyla (represented by Equus
caballus), Carnivora (represented by Canis familiaris, Eumetopias jubatus, and Felis
catus), Chiroptera (represented by Pteropus lyelli, Nycteris grandis, Rhinolophus
ferrumequinum, and Tadarida brasiliensis), and Eulipotyphla (represented by Atelerix
albiventris and Sorex monticolus). Dimensions of the bony labyrinths of laurasiatheres
are provided in Table 5.2. Dimensions of the cochlea are provided in Table 5.3, and
dimensions and orientations of the semicircular canals are reported in Tables 5.4-5.6.
Most character states reconstructed for the ancestral laurasiatherian were retained
from its boreoeutherian ancestor. That is, the lateral semicircular canal entered directly
into the vestibule without forming a secondary common crus, the arc of the anterior
semicircular canal has the greatest radius, and the plane of the lateral semicircular canal
was high compared to the junction of the posterior canal and its ampulla. The state of the
aspect ratio of the cochlea was reconstructed as equivocal, as was calculated for
Boreoeutheria. The ancestral degree of coiling of the cochlea of Laurasiatheria (751.0°)
was less than that reconstructed for Boreoeutheria (815.4°), but the contribution of the
231
cochlea to the entire labyrinthine volume of Laurasiatheria (55%) was similar to that of
the boreoeutherian ancestor (56%).
The relationships within Boreoeutheria that were recovered by Bininda-Emonds
et al. (2007) place Chiroptera in a polytomy with Ungulata (Cetartiodactyla plus
Perissodactyla) and Ferae (Carnivora and Pholidota). The states reconstructed for the
bony labyrinth of the most recent common ancestor of this polytomy were the same as
those calculated for the ancestor of Laurasiatheria. That is, the lateral semicircular canal
entered the vestibule directly in the absence of a secondary common crus, the plane of the
lateral canal was high compared to the ampullar entry of the posterior semicircular canal,
and the arc of the anterior semicircular canal was the largest among the three canal arcs.
The ancestral degree of coiling for the ungulate-feran-chiropteran polytomy was 815.1°,
which was similar to that of the ancestral boreoeutherian condition (815.4°), and the
cochlea contributed 56.0% of the total labyrinthine volume, which also was inherited
from the ancestor of Boreoeutheria (50.0%).
Systematic analyses of mammals based on morphology (Novacek, 1986, 1992a,
b; McKenna and Bell, 1997) group cetartiodactyls (although separated into monophyletic
Artiodactyla with the exclusion of Cetacea) and perissodactyls in a group called Ungulata
along with Sirenia, Hyracoidea, and Proboscidea (and often Tubulidentata). Ungulate
monophyly has been brought into question by more recent molecular results that not only
separate hyracoids, sirenians, proboscideans, and tubulidentates from the perissodactyls,
artiodactyls, and cetaceans into Afrotheria (as discussed above), but some recover a close
relationship between the cetartiodactyls and perissodactyls with a Carnivora+Pholidota
clade (Liu et al., 2001; Murphy et al., 2001a), and at times placing Perissodactyla as the
sister taxon to the Carnivora+Pholidota grouping (Springer et al., 1997; Stanhope et al.,
1998; Murphy et al., 2001b). However, most molecular analyses, including those of
232
Bininda-Emonds et al (2007), recover a Perissodactyla+Cetartiodactyla clade, whether
Cetacea falls within Artiodactyla or not (e.g., Honeycutt and Adkins, 1993; Madsen et al.,
2001).
The only state reconstructed for the ungulate ancestor that differs from that of the
ancestor of Boreoeutheria was the aspect ratio of the cochlea, which was low in the
ancestor of Ungulata. The state of the boreoeutherian cochlea was equivocal, although the
aspect ratio of the cochlea was low in the ancestral therian. The bony labyrinth of the
ancestor of Ungulata had a lateral semicircular canal that opened into the vestibule
directly (retained from the placental ancestor), a position of the plane of the lateral canal
high compared to the posterior canal (retained from the boreoeutherian ancestor), and an
anterior semicircular canal arc as the largest of the three arcs (retained from the therian
ancestor). The ancestral coiling of the cochlea of Ungulata was 857.4°, which was greater
than that reconstructed for the ancestor of the ungulate-feran-chiropteran polytomy
(815.1°), and the ancestral ungulate cochlea contributed 55.0% of the total labyrinthine
volume, which was a value retained from the boreoeutherian ancestor (55.0%).
The ancestor of Ferae (Carnivora plus Pholidota as supported by the results of
numerous analyses; Murphy et al., 2001a, b; Amrine-Madsen et al., 2003; Bininda-
Emonds et al., 2007) retained labyrinthine morphology similar to the most recent
common ancestor of Ungulata, Ferae, and Chiroptera. The lateral semicircular canal
entered the vestibule directly in absence of a secondary common crus, the anterior
semicircular canal arc was the largest among the three arcs, and the lateral canal was
positioned high compared to the ampullar opening of the posterior semicircular canal.
The aspect ratio of the cochlea was equivocal, but the ancestral feran cochlea coiled
888.2° and contributed 56.0% of the total labyrinthine volume. The volumetric
contribution of the cochlea of Ferae was retained from the boreoeutherian ancestor, but
233
the degree of coiling was greater than that reconstructed for its ancestors within
Boreoeutheria.
There are no unambiguous otic synapomorphies that support any relationships
between Ungulata, Ferae, and Chiroptera. Ancestral states reconstructed for the ancestors
of clades within Ungulata, Ferae, and Chiroptera, as well as the ancestral states for
Chiroptera as a whole, are provided in separate sections below.
Terrestrial Cetartiodactyla
The origins of Cetacea were mired in controversy in the past (see Gingerich,
2005, for a brief historical review of cetacean systematics), but most evidence supports a
close relationship between cetaceans and even-toed ungulates (traditionally classified as
Artiodactyla). Both morphology (McKenna and Bell, 1997; Geisler and Luo, 1998;
O’Leary, 1999; O’Leary and Geisler, 1999; O’Leary and Uhen, 1999; Thewissen and
Madar, 1999; Geisler, 2001; Gingerich et al., 2001; Thewissen et al., 2001; Theodor and
Foss, 2005) and molecules (Boyden and Gemeroy, 1950; Graur and Higgins, 1994;
Gatesy et al., 1996; Ursing and Arnason, 1998; Gatesy et al., 1999; Kleineidam et al.,
1999; Madsen et al., 2001; Murphy et al., 2001a, b; Bininda-Emonds et al., 2007) have
been used to suggest a common origin for cetaceans and their terrestrial hoofed relatives.
Although a large amount the genetic data support a nesting of Cetacea within
Artiodactyla, only recently have such relationships found support from morphology
(Naylor and Adams, 2001; Geisler and Uhen, 2003; Geisler and Theodor, 2009). Because
the name Cetartiodactyla is commonly used in scientific literature, including those studies
that include Cetacea within Artiodactyla (e.g., Bininda-Emonds et al., 2007), the name
Cetartiodactyla is used throughout the remainder of the present paper.
234
The terrestrial members of Cetartiodactyla (non-cetacean even-toed ungulates) are
divided into the three major extant groups, which are Suiformes (pigs and hippos),
Tylopoda (camels and llamas), and Ruminantia (deer and cows) (Theodor et al., 2005).
The extinct oreodont Bathygenys reevesi and its closest relative in the sample, the extant
pig Sus scrofa, represent the terrestrial cetartiodactyls here.
Oreodonts (classified under Tylopoda by Theodor et al., 2005) are common
members of the North American mammal biota during the Tertiary (MacFadden and
Morgan, 2003). Bathygenys reevesi itself is a small oreodont from the Airstrip and Little
Egypt local faunas in the Trans-Pecos region of Texas (Wilson, 1971), and inclusion of
the ear region of this taxon extends the temporal range of the placental sample into the
Oligocene (Prothero and Emry, 2004). Preservation of the ear region in the skull of
Bathygenys was such that the matrix filling the inner ear cavities appears very similar to
the bone in the CT scans. Because of this, small structures, such as the bony channels for
the aqueducts for the cochlea and vestibule, are not visible in the digital imagery.
Furthermore, the boundaries of the fenestrae cochleae and vestibuli are ill defined, and
measurements, such as the stapedial ratio calculated from the dimensions of the fenestra
vestibuli, were not taken. However, the gross anatomy of the bony labyrinth of
Bathygenys was segmented (Figures 5.24-5.25) and described here along with Sus scrofa
(Figures 5.26-5.27).
The bony labyrinth of Sus is larger than the labyrinth of Bathygenys, both in terms
of the length of the inner ear (9.95 mm versus 7.40 mm), as well as the gross volume of
the inner ear cavities (61.86 mm
3
versus 29.83 mm
3
). Overall, Sus is a larger animal than
Bathygenys, given that the length of the skull of Sus (240.00 mm) is over two and a half
times greater than that measured for Bathygenys (90.71 mm). The body mass of
235
FIGURE 5.24. Bony labyrinth of Bathygenys reevesi. A, stereopair and labeled line
drawing of digital endocast in anterior view; B, stereopair and labeled line drawing of
digital endocast in dorsal view; C, stereopair and labeled line drawing of digital endocast
in lateral view; D, line drawing of cochlea viewed down axis of rotation to display degree
of coiling; E, line drawing of cochlea in profile. Abbreviations: aa, anterior ampulla; ac,
anterior semicircular canal; ant, anterior direction; av, bony channel for aqueduct of
vestibule; co, cochlea; cr, common crus; dor, dorsal direction; fc, fenestra cochleae; la,
lateral ampulla; lc, lateral semicircular canal; med, medial; pa, posterior ampulla; pc,
posterior semicircular canal; pos, posterior direction; sr, spherical recess of vestibule.
236
A
B
C
D E
co
sr
fc
la
pa
aa
cr
lc
pc
ac
co
cr
aa
la
ac
lc
pc
1 mm
1 mm
1 mm
pa
sr
er
co
co
co
aa
la
lc
cr
ac
pc
pa
dor
med
pos
med
dor
ant
237
FIGURE 5.25. CT slices through ear region of Bathygenys reevesi. Abbreviations: aa,
anterior ampulla; ac, anterior semicircular canal; co, cochlea; cr, common crus; dor,
dorsal direction; la, lateral ampulla; lc, lateral semicircular canal; med, medial; pa,
posterior ampulla; pc, posterior semicircular canal; pos, posterior direction; vb, vestibule.
238
505
510
515
520
525
530
535
540
545
555
550
pos
dor
1 mm
med
dor
co
1 mm
505 510
520
515 525
535530 540
550545 555
co
co
co
co
vb
vb
aa
aa
ac
ac
la
lc
lc lc
lc
lc
co
ac
ac
cr
ac
cr
vb
vb
pa
pc
pc
pc
pc
239
FIGURE 5.26. Bony labyrinth of Sus scrofa. A, stereopair and labeled line drawing of
digital endocast in anterior view; B, stereopair and labeled line drawing of digital
endocast in dorsal view; C, stereopair and labeled line drawing of digital endocast in
lateral view; D, line drawing of cochlea viewed down axis of rotation to display degree of
coiling; E, line drawing of cochlea in profile. Abbreviations: aa, anterior ampulla; ac,
anterior semicircular canal; ant, anterior direction; av, bony channel for aqueduct of
vestibule; cc, canaliculus cochleae; co, cochlea; cr, common crus; dor, dorsal direction;
er, elliptical recess; fc, fenestra cochleae; fv, fenestra vestibuli; la, lateral ampulla; lc,
lateral semicircular canal; med, medial; pa, posterior ampulla; pc, posterior semicircular
canal; pos, posterior direction; ps, outpocketing for the perilymphatic sac.
240
A
B
C
D E
co
cc
fv
ps
1 mm
1 mm
1 mm
fc
la
aa
cr
ac
lc
pc
pa
co
cc
av pa
pc
fc
aa
cr
la
lc
er
sr
aa
la
lc
ac
cr
pc
av
pa
fc
fv
co
ps cc
co
cc
ps
fc
ps
fc
cc
co
dor
med
med
pos
dor
ant
241
FIGURE 5.27. CT slices through ear region of Sus scrofa. Abbreviations: aa, anterior
ampulla; ac, anterior semicircular canal; av, bony channel for aqueduct of vestibule; ca,
canal for vasculature through petrosal; cc, canaliculus cochleae; cf, foramina within
cribriform plate; cn, canal for cranial nerve VIII; co, cochlea; cr, common crus; dor,
dorsal direction; fc, fenestra cochleae; fn, canal for cranial nerve VII; iam, internal
auditory meatus; la, lateral ampulla; lat, lateral direction; lc, lateral semicircular canal;
pa, posterior ampulla; pc, posterior semicircular canal; pl, primary bony lamina; pos,
posterior direction; sa, subarcuate fossa; sr, spherical recess of vestibule; st, stapes within
fenestra vestibuli; vn, canal for vestibular branch of cranial nerve VIII.
242
pos
cf
cc
lat
100
134
168
202
236
270
304
338
372
440
406
5 mm
dor
co
co
fn
fn
fn
vn
pc
cn
fn
iamiam
fn
fc
st
sr
sr
pa
pa
la
aa
aa
ca
sa
ac
ac
ac
sa
ca
aa
cn
cn pl
fn
lc
lc
pc
pc
pc
ac
av
av
cr
ca
cr
cr
cf
co
cn pl
pl
1 mm
lat
372 406 440
270 304 338
168 202 236
100 134
ac
st
co
co
co
fc
243
Bathygenys was not estimated, but the average body mass of Sus is 88 kg (Silva and
Downing, 1995).
Although the cochlea of Sus is more voluminous (36.25 mm
3
) than the cochlea of
Bathygenys (16.17 mm
3
), the structure contributes a similar amount to the total volume of
the bony labyrinth in both species (54.2% in Bathygenys; 58.6% in Sus). However, this is
the only similarity between the cochleae of the two taxa. The cochlear canal is
significantly longer (22.89 mm) and coils to a much greater degree (1274°) in Sus than is
observed in Bathygenys (canal is 8.51 mm long; canal coils 665°). Furthermore, the
cochlear spiral of Sus has a much higher aspect ratio (0.71) than the cochlea of
Bathygenys (0.32). The aspect ratio of the cochlea of Bathygenys is similar the ratio
calculated for Hemicentetes semispinosum (0.38). The apical whorl of the
cochlea of Bathygenys sits upon the basal turn, although the diameter of the apical whorl
is smaller than the basal, as can be seen when the cochlea is in vestibular view.
The shape of the cochleae of Sus and Procavia are strikingly similar (Figure
5.16). The aspect ratios calculated for the cochlear spirals are identical between the two
taxa, and both spirals form a pyramid-like structure, where the apical whorls do not sit
directly upon the basal whorl, but rather fit within the basal turn when the cochlea is in
vestibular view. The diameters of the second and third turns are similar, so that the third
whorl shields most of the second from view when the cochlea is viewed down the axis of
rotation.
Preservation of the bony labyrinth of Bathygenys was such that presence or
absence of a secondary bony lamina within the cochlea could not be determined (Figure
5.25). However, the CT scans through the ear region of Sus (Figure 5.27) demonstrate
that the secondary lamina was not present in the cochlea of the pig. The scala tympani of
Sus is expanded posterodorsal to the fenestra cochleae, and a very robust bony
244
canaliculus cochleae projects posteromedially from the excavation. The canaliculus is
triangular in cross-section and 2.64 mm in length. As stated above, the canaliculus
cochleae was not observed for Bathygenys, owing to preservation of the fossil specimen.
The canaliculus cochleae likely was present in the taxon, given that the structure is
observed in all other cetartiodactyls examined here and elsewhere (e.g., Gray, 1907), as
well as most placental taxa.
The angles formed between the planes of the basal turn of the cochlea and the
lateral semicircular canal are not much different between Bathygenys (26.8°) and Sus
(23.8°). The degree of angular deviation observed in these taxa is similar to that measured
in the elephant shrew, Macroscelides proboscideus (25.1°). The stapedial ratio could not
be measured for Bathygenys, but a ratio of 1.28 was calculated for Sus. The spherical
recess of the vestibule, through which the labyrinth communicates with the middle ear
cavity via the fenestra vestibuli and stapes, is well defined in Sus. The recess in the pig is
a sphere that is bisected, and the fenestra vestibuli is situated on the cut surface. The
spherical recess cannot be distinguished from the elliptical recess in Bathygenys, where
the vestibule is developed as a continuous yet irregularly shaped cavity.
The elliptical recess of the vestibule of Sus is elongate, with a stout anterior
projection that is expressed as a pedestal for the anterior and lateral ampullae on the
digital endocast. A similar projection is observed in the vestibule of Bathygenys. All three
ampullae in Sus are rounded, teardrop-shaped excavations within the vestibule. The
posterior limb of the lateral semicircular canal opens directly into the vestibule dorsal to
the posterior ampulla in both cetartiodactyls. Because of this, the lateral canal divides the
space enclosed by the arc of the posterior semicircular canal when either labyrinth is
viewed anteriorly. The sagittal labyrinthine index is nearly the same for Bathygenys
(17.4) as it is for Sus (16.5).
245
The bony channel for the aqueduct of the vestibule is not observed in Bathygenys,
but the channel is a thin canal in Sus that exits the vestibule ventromedial and slightly
anterior to the vestibular aperture of the common crus. The channel for the aqueduct of
the vestibule is expressed on the endocast as a fine thread before expanding as the
aqueduct opens into a fissure near the endocranial surface of the petrosal. Although the
channel for the aqueduct of the vestibule is more delicate than the canaliculus cochleae
for the aqueduct of the cochlea, the vestibular channel is slightly longer, with a length of
3.18 mm (as opposed to 2.64 mm for the canaliculus cochleae).
Although the specific measured values of the angles between the planes of the
three semicircular canals are different for Sus and Bathygenys, the basic pattern is the
same for both of these species. For example, the angle between the planes of the anterior
and posterior semicircular canals is the greatest for both Bathygenys (99.6°) and Sus
(96.0°), and the angle between the anterior and lateral canals is the smallest (Bathygenys
equals 86.0°; Sus equals 82.8°). The angles between the planes of the posterior and lateral
semicircular canals are the closest to 90° for both Bathygenys (91.3°) and Sus (87.9°).
The semicircular canals of Sus are more planar (fit better onto a single plane) than
the canals of Bathygenys. In fact, the anterior semicircular canal is perfectly planar in
Sus, whereas the anterior canal deviates from its plane by a total of 8.10°. The posterior
canal is the least planar of the three for both taxa, where it deviates from its plane by
2.63° in the labyrinth of Sus and by 13.5° in Bathygenys. The lateral semicircular canal of
Sus deviates from its plane by a total of 2.20° and the lateral canal of Bathygenys deviates
by 7.92°. Only the posterior semicircular canal of Bathygenys is considered significant
(ratio of total linear deviation over cross-sectional diameter is 1.23 versus 0.61 for
anterior and 0.64 for lateral; ratios for anterior, lateral, and posterior canals of Sus are
0.00, 0.20, and 0.24 respectively).
246
A similar pattern is observed in the radius of the semicircular canal arcs, as well
as in the cross-sectional diameter of the canals. That is, the radius of the anterior
semicircular canal arc is the largest in both Sus (2.50 mm; lateral equals 2.08 mm;
posterior equals 2.18 mm) and Bathygenys (1.91 mm; lateral equals 1.52 mm; posterior
equals 1.79 mm). The diameter of the lumen of the anterior semicircular canal is greater
than the other two canals in both Sus (0.42 mm; lateral equals 0.36 mm; posterior equals
0.42 mm) and Bathygenys (0.44 mm; lateral equals 0.33 mm; posterior equals 0.34 mm).
However, this pattern is not observed universally across all dimensions of the
semicircular canals. For example, the slender canal length of the anterior semicircular
canal is the greatest in Sus (12.14 mm; lateral equals 8.04 mm; posterior equals 10.65
mm), but the posterior canal is the longest in Bathygenys (10.01 mm; anterior equals 9.72
mm; posterior equals 7.11 mm).
The aspect ratio of the lateral semicircular canal is the greatest in both Bathygenys
(0.99) and Sus (0.83), although the aspect ratio of the anterior canal was the smallest in
Bathygenys (0.86; Sus equals 0.78), whereas the aspect ratio of the posterior canal was
the smallest in Sus (0.74; B. reevesi equals 0.95). The ratio between the slender canal
length and arc radius for the posterior semicircular canal was the greatest for both
Bathygenys (5.59; anterior equals 5.08; lateral equals 4.68) and Sus (4.89; anterior equals
4.86; lateral equals 3.87).
The bony labyrinth of the ancestor of Cetartiodactyla was similar to that
reconstructed for the ancestor of Ungulata. The lateral semicircular canal opened into the
vestibule directly in absence of a secondary common crus, the arc of the anterior
semicircular canal was the largest among the three, the lateral semicircular canal was
positioned high compared to the posterior canal, and the aspect ratio of the cochlea was
low. The cochlea of Cetartiodactyla coiled to a lesser degree than Ungulata (845.8°
247
versus 857.4°), but the cochlear canal contributed a greater percentage to the overall
labyrinthine volume (59% versus 55%).
The labyrinths of the two terrestrial cetartiodactyls retain the ancestral
cetartiodactyl condition of the anterior canal possessing the largest arc radius. The
cochlea of Bathygenys is flattened (low aspect ratio), which is the ancestral condition,
although the cochlea of Sus has a high profile. Both labyrinths retain the ancestral
cetartiodactyl condition of the high position of the lateral semicircular canal as compared
to the posterior canal, and a vestibular entrance of the lateral canal, rather than formation
of a secondary common crus.
Although the cladogram presented in Figure 5.2 (modified from the supertree of
Bininda-Emonds et al., 2007, with additional cetartiodactyl information from Theodor et
al., 2005) depicts a closer relationship between Sus and Cetacea, there are no
unambiguous otic synapomorphies supporting this relationship. Both Sus and Bathygenys
share a high position of the lateral semicircular canal that is absent in Cetaceans
(discussed in the following section), but this state was ancestral for crown Placentalia as a
whole. Nonetheless, the most recent common ancestor of Sus and Cetacea possessed a
bony labyrinth with the lateral semicircular canal opening directly into the vestibule, the
anterior semicircular canal arc with the largest radius among the three canals, a high
position of the lateral semicircular canal compared to the posterior canal, and a low
aspect ratio of the cochlea in profile. The ancestral cochlear coiling of Sus and Cetacea
was 1013.1°, and the contribution of the cochlea to the entire labyrinth is 67.0%.
Cetacea
With the exception of Sirenia (the bony labyrinth of which was described above
in the Afrotheria section), cetaceans are the only fully aquatic extant mammals. Two
248
major cetacean clades recognized are the baleen whales, or Mysticeti, which includes the
largest living mammal (Balaenoptera musculus), and the toothed whales, Odontoceti,
which includes porpoises and dolphins such as Tursiops truncatus. The bony labyrinth of
the bottlenose dolphin Tursiops is described, along with the labyrinth of a fossil member
of Balaenopteridae (Mysticeti).
The bony labyrinth of the extinct balaenopterid (Figures 5.28-5.29) is larger than
that of Tursiops (Figures 5.30-5.31). The anterior-posterior length of the mysticete inner
ear is 19.67 mm (Tursiops equals 10.01 mm), and the gross volume of the inner ear
cavities of the mysticete is 1075.51 mm
3
(Tursiops equals 167.98 mm
3
). The greater
dimensions of the mysticete labyrinth likely reflects body size differences between
balaenopterids and Tursiops, with the average body mass of most mysticete species being
several orders of magnitude greater than that of the bottlenose dolphin (Silva and
Downing, 1995). Average body mass for Tursiops is 179.5 kg, whereas the smallest
extant balaenopterid is 4,000 kg (Balaenoptera acutorostrata; Silva and Downing, 1995).
The cochlea of the balaenopterid is larger than Tursiops in all dimensions,
including volume (973.91 mm
3
versus 157.11 mm
3
), cochlear canal length (53.02 mm
versus 24.01 mm), degree of coiling (886° versus 661°), and even aspect ratio, although
to a lesser extent (0.48 versus 0.47). The volumetric contribution of the cochlea of
towards the total inner ear volume is greater for Tursiops (93.5%) than for the
balaenopterid (90.6%), although the value for the balaenopterid is exceptionally high.
The significant contribution of the total volume by the cochlea is higher for the two
cetacean taxa than any other mammal investigated here, including the afrotherians
Chrysochloris, Macroscelides, and Trichechus (each with a cochlear contribution around
71-72%). The other extreme is the cochlea of the elephantoid, which only contributes
30.6% of the total volume of the bony labyrinth.
249
FIGURE 5.28. Bony labyrinth of fossil Balaenopteridae. A, stereopair and labeled line
drawing of digital endocast in anterior view; B, stereopair and labeled line drawing of
digital endocast in dorsal view; C, stereopair and labeled line drawing of digital endocast
in lateral view; D, line drawing of cochlea viewed down axis of rotation to display degree
of coiling; E, line drawing of cochlea in profile. Abbreviations: aa, anterior ampulla; ac,
anterior semicircular canal; ant, anterior direction; cc, canaliculus cochleae for aqueduct
of cochlea; co, cochlea; cr, common crus; dor, dorsal direction; fc, fenestra cochleae; fv,
fenestra vestibuli; la, lateral ampulla; lc, lateral semicircular canal; med, medial
direction; pa, posterior ampulla; pc, posterior semicircular canal; pl, primary bony
lamina; pos, posterior direction; ps, outpocketing for perilymphatic sac; sl, secondary
bony lamina; sr, spherical recess of vestibule.
250
A
B
C
D E
co
5 mm
sl
fc
fv
lc
pc
cc
av
aa
ac
sr
co
sl
pl
cc
fc
av
cr
pa
pc
la
lc
aa
ac
co
sl
pa
pc
av
cr
ac
aa
lc
fv
fc
pl
ps
5 mm
5 mm
fc
sl
co
co
sl
fc
dor
med
med
pos
dor
ant
251
FIGURE 5.29. CT slices through ear region of fossil Balaenopteridae. Abbreviations: aa,
anterior ampulla; ac, anterior semicircular canal; av, bony channel for aqueduct of
vestibule; cc, canaliculus cochleae; cf, foramina within cribriform plate; cn, canal for
cranial nerve VIII; co, cochlea; dor, dorsal direction; fc, fenestra cochleae; fn, canal for
cranial nerve VII; la, lateral ampulla; lc, lateral semicircular canal; med, medial
direction; pa, posterior ampulla; pc, posterior semicircular canal; pl, primary bony
lamina; pos, posterior direction; sg, canal for spiral ganglion within primary bony lamina;
sl, secondary bony lamina; st, stapes within fenestra vestibuli; vb, vestibule; vn, canal for
vestibular branch of cranial nerve VIII.
252
pos
dor
648
679
710
741
772
803
834
865
896
958
927
10 mm
med
dor
1 mm
fn
fn
fn
fn
fn
st
fn
av
av
cn
cn
cn
cn
co
co
fc
fc
pa
fc
cc
cn
vn
vn
fn
co
sl
sl
sl
sg
cf
cf
co
cn
pl
pl
vb
aa
ac
ac
lc
648 679
741710 772
834803 865
927896 958
sg
sg
sg
cf
vn
vb
pl
pl
pl
sl
sl
sl
sl
lc
lc
pc
pc
co
co
co
st
st
st
253
FIGURE 5.30. Bony labyrinth of Tursiops truncatus. A, stereopair and labeled line
drawing of digital endocast in anterior view; B, stereopair and labeled line drawing of
digital endocast in dorsal view; C, stereopair and labeled line drawing of digital endocast
in lateral view; D, line drawing of cochlea viewed down axis of rotation to display degree
of coiling; E, line drawing of cochlea in profile. Abbreviations: aa, anterior ampulla; ac,
anterior semicircular canal; ant, anterior direction; cc, canaliculus cochleae for aqueduct
of cochlea; co, cochlea; cr, common crus; dor, dorsal direction; fc, fenestra cochleae; fv,
fenestra vestibuli; la, lateral ampulla; lc, lateral semicircular canal; med, medial
direction; pa, posterior ampulla; pc, posterior semicircular canal; pl, primary bony
lamina; pos, posterior direction; ps, outpocketing for perilymphatic sac; sl, secondary
bony lamina.
254
A
B
C
D E
co
sl
pl
fc
ps
cc
lc
pc
la
av
aa
ac
sl
co
cr
cc
ps
av
aa
la
5 mm
lc
pc
pa
fc
fc
cc
pl
sl
co
av
ac
pc
lc
fv
pa
5 mm
5 mm
co
sl
pl
cc
fc
co
sl
cc
fc
dor
med
med
pos
dor
ant
255
FIGURE 5.31. CT slices through ear region of Tursiops truncatus. Abbreviations: ac,
anterior semicircular canal; ant, anterior direction; av, bony channel for aqueduct of
vestibule; cf, foramina within cribriform plate; cn, canal for cranial nerve VIII; co,
cochlea; dor, dorsal direction; fn, canal for cranial nerve VII; iam, internal auditory
meatus; la, lateral ampulla; lat, lateral direction; lc, lateral semicircular canal; med,
medial direction; pl, primary bony lamina; sg, canal for spiral ganglion within primary
bony lamina; sl, secondary bony lamina; st, stapes within fenestra vestibuli; vb, vestibule.
256
ant
co
cf
sg
sg
pl
pl
sl
130 140
160150 170
190180 200
220210 230
dor
130
140
150
160
170
180
190
200
210
230
220
5 mm
1 mm
lat
dor
iam
iam
iam
iam
iam
fn
fn
fn
fn
fn
fn
cn
cn
cn
cn
fc
cn
cn
cf
co
co
co
co
co
co
co
st
vb
cf
st
sl
sl
sg
sg
fn
fn
av
av
av av
lc
la
vb
ac
ac
sl
sg
sl
sg
sl
257
As is evident with the aspect ratios reported above, the cochleae of both cetaceans
are planispiral, and the apical turns of the cochlea are smaller in diameter than the basal
whorl. A well-developed secondary bony lamina is present in both labyrinths (Figures
5.29, 5.31), extending from a point anterior to the fenestra cochleae for a considerable
distance along the radial wall of the cochlear canal. The secondary lamina, which is
expressed as a distinct groove on the endocast, is present for around two thirds of the
basal turn (238°) in the balaenopterid, but it persists for a short distance beyond the basal
turn in Tursiops (396°).
An anteriorly oriented excavation of the cochlea (expressed as a flange on the
endocast) is immediately basal to the apical terminus of the secondary bony lamina in the
balaenopterid. A similar structure is observed in the reconstruction of the inner ear of the
extinct mysticete Herpetocetus (Geisler and Luo, 1996), and it might be characteristic of
the Mysticeti. Such an extension of the cochlea is not observed in Tursiops or any other
mammal studied here. The anterior excavation is distinct, but it is not deep enough to
give the cochlea an elliptical outline when viewed vestibularly.
The bony canaliculus for the aqueduct of the cochlea is significantly longer in
Tursiops (6.47 mm) than in the balaenopterid (3.83 mm). The canaliculus of Tursiops is
roughly triangular in cross-section, whereas the bony passage is flattened. The scala
tympani is not inflated near the proximal end of the canaliculus cochleae posterior to the
fenestra cochleae in Tursiops, but a groove (expressed as a ridge on the endocast) extends
from the canaliculus cochleae for a short distance on the tympanal surface of the cochlea
in the balaenopterid. The angle between the planes of the basal turn of the cochlea and
lateral semicircular canal for the balaenopterid (23.2°) and Tursiops (21.3°) are similar to
the angles measured for the terrestrial cetartiodactyls Bathygenys (26.8°) and Sus (23.8°).
258
The fenestra vestibuli is separated from the fenestra cochleae by a great distance
in both cetacean taxa. This likely is the result of a flexure near the junction of the cochlea
and vestibule. This cochlear hook is a common feature in the inner ear of cetaceans,
primarily odontocetes (Fleischer, 1976; Luo and Eastman, 1995; Luo and Marsh, 1996).
The stapedial ratios, as calculated from the fenestra vestibuli, are more circular in both
cetacean taxa (1.52 for the balaenopterid; 1.44 for Tursiops) than the ratios calculated for
other placental mammals.
The entire vestibular apparatus of the cetacean bony labyrinth is significantly
smaller than that of other taxa. This is expressed not only in the imagery of the endocasts,
but also in the volumetric measurements (cochlea of Tursiops contributes around 93% of
the volume, so vestibular apparatus contributes 7%). The vestibule of Tursiops is bowed
medially with the fenestra vestibuli opening through an anterior excavation of the cavity.
The vestibule is not curved in the balaenopterid, and the spherical recess is small and is
distinguished from the elliptical recess by a gentle constriction behind the fenestra
vestibuli. A similar constriction is observed in Tursiops, but the connection between the
anterior aspect of the spherical recess and the cochlea is not distinguishable.
An extension of the elliptical recess adjacent to the constriction between the
vestibular compartments leads to the lateral and posterior ampulla. The posterior limb of
the lateral semicircular canals of both cetacean taxa do not have separate openings into
the vestibule. Rather, the lateral canal empties into the posterior ampulla just above the
vestibular aperture of the ampulla (a secondary common crus is not present in either
taxon). The bony channel for the aqueduct of the vestibule exits the bony labyrinth near
the medial edge of the vestibular aperture of the common crus in the balaenopterid. The
passage in this taxon is expressed as fine thread that extends 3.65 mm to the endocranial
aperture of the aqueduct. The vestibular aperture for the channel for the aqueduct of the
259
vestibule of Tursiops is separated from the common crus by a relatively greater distance
than it is in the balaenopterid. In medial view, the massive cochlea shields the channel in
Tursiops, but is best observed when the bony labyrinth is oriented dorsally.
The angle between the planes of the anterior and posterior semicircular canals is
obtuse in both Tursiops (97.6°) and the balaenopterid (105°), but the remaining canal
plane angles are acute. The lowest angles for each cetacean were measured between the
anterior and lateral canals (balaenopterid equals 71.6°; Tursiops equals 75°). Overall, the
semicircular canals of Tursiops fit onto single planes (the anterior and posterior canals
are planar). The only canal of Tursiops that does not fit onto a single plane, the lateral
semicircular canal, deviates from its plane by a total of 8.86°, although this degree is not
significant (ratio of total linear deviation over cross-sectional diameter is 0.85). The
angular deviations of the anterior (9.0°) and posterior (15.9°) semicircular canals of the
balaenopterid are significant (ratio of anterior is 1.27; posterior is 1.56), but the lateral
semicircular canal of the balaenopterid deviates from its plane by 5.4°, which is not
significant (ratio is 0.39).
An identical pattern is observed between the balaenopterid and Tursiops when the
semicircular canal arc radii are compared. The radius of the anterior canal arc is the
largest in both taxa (2.45 mm in the balaenopterid; 1.19 mm in Tursiops), and the radius
of the posterior arc is the smallest (1.92 mm in the balaenopterid; 0.84 mm in Tursiops).
The radius of curvature of the lateral semicircular canal arc is 2.11 mm in the
balaenopterid and 1.36 mm in Tursiops. A common pattern is not observed in any of the
other dimensions of the semicircular canals. The slender canal length of the anterior
semicircular canal of the balaenopterid is the greatest (10.65 mm; lateral equals 8.54 mm;
posterior equals 9.46 mm), although the lateral canal is the longest in Tursiops (4.61 mm;
anterior equals 4.14 mm; posterior equals 4.35 mm). The diameter of the lumen is
260
greatest in the lateral semicircular canal in the balaenopterid (0.51 mm; anterior equals
0.32 mm; posterior equals 0.34 mm), whereas the diameters of the anterior and posterior
canals are equal (0.27 mm) and larger than the value taken for the lateral canal (0.25 mm)
in Tursiops.
The aspect ratio of the posterior semicircular canal is greatest in both Tursiops
(1.60) and the balaenopterid (1.21). The aspect ratio of the lateral semicircular canal is
the smallest in the balaenopterid (0.39; anterior equals 0.91), and the ratio of the of the
anterior canal is the smallest in Tursiops (0.95), although it is not much different than that
of the lateral canal (0.96). The largest ratio between the length of any slender canal and
arc radius of curvature among the cetaceans examined was calculated for the posterior
semicircular canal of Tursiops (5.17). The ratio calculated for the posterior canal of the
balaenopterid was the largest among the three canals in this specimen as well (4.94).
Although the ratio calculated for the posterior canal of Tursiops was larger than that of
the balaenopterid, the ratios for the anterior (4.19) and lateral (4.05) semicircular canals
were larger than those calculated for Tursiops (3.47 and 3.38 respectively).
The bony labyrinth of the ancestral cetacean lacked a secondary common crus
formed between the lateral and posterior semicircular canal, as did the ancestor of
Cetartiodactyla, but the cetacean labyrinth was derived from the ancestral cetartiodactyl
condition in that the lateral canal opened into the posterior ampulla, rather than into the
vestibule directly. Although this state is a synapomorphy for Cetacea within
Cetartiodactyla, the lateral canal also opens into the posterior ampulla in Perissodactyla,
Scandentia, some Carnivora, and some Chiroptera (see below). A second otic
synapomorphy that separates Cetacea from the terrestrial cetartiodactyls is a low position
of the plane of the lateral semicircular canal compared to the ampullar entrance of the
261
posterior semicircular canal. The state is derived with respect to the ancestral
cetartiodactyl condition, and it is a reversal to the ancestral therian state.
Additional states reconstructed for the ancestor of Cetacea include the anterior
semicircular canal arc as the greatest among the three arcs, and a low aspect ratio for the
cochlear spiral in profile. The coiling of the cochlea of the ancestral cetacean (853.4°)
was retained from the ungulate ancestor (857.4°), but the contribution of the ancestral
cetacean cochlea to the total labyrinthine volume was greater than that calculated for
Ungulata (84.0% versus 55.0%). The high contribution of the cochlea to the total volume
distinguishes Cetacea from other members of Cetartiodactyla.
Perissodactyla
The odd-toed ungulates that make up extant Perissodactyla are divided into
Equidae (horses), Tapiridae (tapirs), and Rhinocerotidae (rhinoceroses). Monophyly of
Perissodactyla is well supported (Gregory, 1910; Simpson, 1945; McKenna and Bell,
1997; Cao et al., 1999; Norman and Ashley, 2000; Bininda-Emonds et al., 2007), as is a
sister taxon relationship between Tapiridae and Rhinocerotidae within the group
(Radinsky, 1964; Cao et al., 1999; Norman and Ashley, 2000; Murphy et al., 2001a, b;
Hooker, 2005; Bininda-Emonds et al., 2007). Only the modern horse, Equus caballus,
was available for examination. Imagery of the inner ear and an endocast of the bony
labyrinth is presented in Figures 5.32-5.33.
The total volume of the inner ear cavities of Equus (165.16 mm
3
) is similar to that
of the dolphin, Tursiops truncatus (167.98 mm
3
). This is also reflected in the length of
the skull, where the skull of Equus (530.00 mm) is slightly shorter than that of Tursiops
(543.02 mm). However, the total length of the inner ear is larger in Equus (16.52 mm)
262
FIGURE 5.32. Bony labyrinth of Equus caballus. A, stereopair and labeled line drawing
of digital endocast in anterior view; B, stereopair and labeled line drawing of digital
endocast in dorsal view; C, stereopair and labeled line drawing of digital endocast in
lateral view; D, line drawing of cochlea viewed down axis of rotation to display degree of
coiling; E, line drawing of cochlea in profile. Abbreviations: aa, anterior ampulla; ac,
anterior semicircular canal; ant, anterior direction; av, bony channel for aqueduct of
vestibule; cc, canaliculus cochleae for aqueduct of cochlea; co, cochlea; cr, common
crus; cv, canal for cochlear vein; dor, dorsal direction; er, elliptical recess of vestibule;
fc, fenestra cochleae; fv, fenestra vestibuli; la, lateral ampulla; lc, lateral semicircular
canal; med, medial direction; pa, posterior ampulla; pc, posterior semicircular canal; pl,
primary bony lamina; pos, posterior direction; ps, outpocketing for perilymphatic sac; sr,
spherical recess of vestibule.
263
A
B
C
D E
co
5 mm
av
cr
aa
ac
pc
lc
pa
la
sr
er
fv
fc
ps
pl
co
aa
la
er
ps
fc
lc
pa
pc
av
cc
ac
cv
cv
cc
av
pa
la
fv
fc
ps
co
lc
aa
ac
cr
pc
5 mm
5 mm
co
cv
cc
fc
ps
co
ps
fc
cccv
dor
med
med
pos
dor
ant
264
FIGURE 5.33. CT slices through ear region of Equus caballus. Abbreviations: ac,
anterior semicircular canal; av, bony channel for aqueduct of vestibule; cc, canaliculus
cochleae for aqueduct of cochlea; cn, canal for cranial nerve VIII; co, cochlea; cr,
common crus; dor, dorsal direction; er, elliptical recess of vestibule; fc, fenestra
cochleae; fn, canal for cranial nerve VII; fv, fenestra vestibuli; iam, internal auditory
meatus; la, lateral ampulla; lc, lateral semicircular canal; med, medial direction; pa,
posterior ampulla; pc, posterior semicircular canal; pl, primary bony lamina; pos,
posterior direction; ps, outpocketing for perilymphatic sac; st, stapes that has fallen into
vestibule; vb, vestibule; vn, canal for vestibular branch of cranial nerve VIII.
265
dor
med
co
cc cc
pa
lc
pc
pa
218
233
248
263
278
293
308
323
338
368
353
5 mm
pos
med
218 233
pl
pl
st
cn
cn
fn
fn
fn
vn
vb
er
vb
fc
fv
co
la
av
ps
iam
cv
263248 278
308293 323
353338 368
1 mm
av
pc
av
av
av
av
cc
cr
cr
ac
ac
ac
pc
ac
iam
co
pc
cv
266
than it is in Tursiops, and the horse has a greater average body mass (258.3 kg versus
179.5 kg; Silva and Downing, 1995).
Although the total volume of the bony labyrinth of Equus is similar to that in
Tursiops, the cochlea of the horse is only half the volume of the cochlea of the dolphin
(84.33 mm
3
versus 157.11 mm
3
). Because of this, the cochlea’s total volumetric
contribution of Equus (51.1%) is significantly less than that of Tursiops (93.5%), but
rather is more in line with the percentage calculated for the terrestrial cetartiodactyl
Bathygenys reevesi (54.2%).
Each turn of the cochlea, of which there are two and a half (900°), is separated
from the adjacent whorls (Figure 5.32). The cochlea of Equus has an aspect ratio of 0.41,
which is similar to that calculated for the elephantoid (0.42) and Orycteropus afer (0.45).
The total length of the cochlear canal of Equus is 22.08 mm. This value is similar
to that measured in the cochlea of Sus (22.89 mm), and even Tursiops (24.01 mm),
despite the fact that the volume of the cochlea of the horse falls in between the dolphin
(157.11 mm
3
) and pig (36.25 mm
3
).
A bony secondary lamina, which extends around two fifths of the basal turn
(153°), is observed in Equus (Figure 5.33). The scala tympani is expanded into a wedge
shaped excavation leading to the straight canaliculus cochleae for the aqueduct of the
cochlea posterior to the fenestra cochleae. The canaliculus narrows towards its terminus,
and flattens into a fissure near its external aperture, although it retains robusticity along
its course. The bony passage is 11.33 mm in length, which is the longest canaliculus
measured for any taxon described here (Table 5.3). A short and delicate secondary
passage, likely for a vein, exits the medial side of the canaliculus near the midpoint of the
bony channel for the aqueduct.
267
The angle between the planes of the basal turn of the cochlea and the lateral
semicircular canal in Equus is 37.9°, and the stapedial ratio, measured from dimensions
of the fenestra vestibuli, is 1.7. The spherical and elliptical recesses are separated by a
constriction of the vestibule ventral to the ampullae of the semicircular canals. The
constriction forms a bony ring (expressed as a wide sulcus on the endocast) surrounding
the vestibule. The bony ring sits on a plane that nearly is parallel with that of the lateral
semicircular canal. The elliptical recess is elongate, and it slightly bows medially, away
from the arc of the lateral semicircular canal.
The anterior and lateral ampullae open into a slight excavation at the anterior end
of the elliptical recess. At the posterior end of the recess, the posterior limb of the lateral
semicircular canal opens into the posterior ampulla, immediately anterolateral to the
vestibular aperture of the ampulla. Because of this, the lateral semicircular canal does not
have its own opening into the vestibule (also observed in both cetaceans, Hemicentetes,
and other taxa, including the tree shrew, some carnivorans, and some bats, as described
below). Additionally, the plane of the canal is high relative to other vestibular
constituents. The elevated lateral semicircular canal divides the space enclosed by the
posterior semicircular canal arc when the endocast of the bony labyrinth is viewed
posteriorly, and the sagittal labyrinthine index of Equus is 10.50. This is the smallest
index calculated for any mammal described thus far in the paper, in which the plane of
the lateral canal takes a high position.
A groove (expressed on the endocast as a low ridge) extends from the
dorsomedial edge of the spherical recess to the vestibular aperture of the bony channel for
the aqueduct of the vestibule, which is situated ventral and medial to the vestibular
aperture of the common crus. The channel is developed as a very delicate thread for half
of its length in Equus, extending posteriorly and crossing the base of the common crus
268
when the endocast of the bony labyrinth is viewed medially. The distal half of the
channel is broad and flattened, indicating that the aqueduct enters a fissure before exiting
on the endocranial surface of the petrosal. The total length of the bony channel for the
aqueduct of the vestibule is 11.66 mm, which is a bit larger than the length of the
canaliculus cochleae (11.33 mm).
The planes of the posterior and lateral semicircular canals of Equus roughly form
a right angle with one another (90.1°), and the angle between the planes of the posterior
and anterior canals only is slightly obtuse (93.3°). The angle between the anterior and
lateral semicircular canal planes is 84.7°. The semicircular canals themselves are fairly
planar, especially the anterior canal, which deviates from an average plane by a total of
2.22° (the total angular deviation of the lateral and posterior canals for their planes are
4.68° and 5.74° respectively). The degree of deviation is not significant for any canal,
where the ratios of total linear deviation over cross-sectional diameter of the canal for the
anterior, lateral, and posterior canals are 0.27, 0.64, and 0.70 respectively.
Both the slender canal length (18.63 mm) and radius of canal arc (3.83 mm) of the
posterior semicircular canal are the greatest among these dimensions measured for the
three canals of Equus. The lateral semicircular canal exhibited the smallest value of these
two dimensions, where the slender canal length is 14.28 mm (versus 17.35 mm for the
anterior) and the arc radius is 3.55 mm (versus 3.62 mm for the anterior). The largest
canal lumen diameter in cross-section was measured for the anterior semicircular canal
(0.51 mm; lateral equals 0.45 mm; posterior 0.50 mm), and the largest volume was
measured for the lateral semicircular canal (2.57 mm
3
; anterior equals 2.19 mm
3
;
posterior equals 2.32 mm
3
).
The aspect ratios of the anterior and lateral semicircular canal arcs are similar
(0.95 and 0.96 respectively), and both are lower than the aspect ratio of the posterior
269
canal (1.60). The high aspect ratio of the posterior semicircular canal indicates that the
height of the canal arc is greater than the width. The greatest ratio of slender canal length
to arc radius was calculated for the posterior semicircular canal (5.17), and the ratio for
the anterior (3.46) and lateral (3.38) semicircular canals are not significantly different.
The cochlear spiral of Equus possesses the ancestral ungulate state of a low aspect
ratio, giving the cochlea a flattened appearance. The pattern of semicircular canal arc
radii in Equus, with the largest radius being the anterior, is inherited from its
boreoeutherian ancestor, although the entry of the lateral semicircular canal into the
posterior ampulla is derived and shared with Cetacea. The high position of the lateral
semicircular canal with respect to the ampullar opening of the posterior canal is retained
from the ancestor of Boreoeutheria.
Carnivora
Extant carnivorans belong to two phylogenetically distinct clades, Caniformia
(dogs, bears, raccoons, weasels, and pinnipeds) and Feliformia (cats, hyenas, mongooses,
and viverrids). Monophyly of Pinnipedia (within Caniformia) has been questioned in the
past (McLaren, 1960; Tedford, 1976), although most recent data are in support of a single
origin for seals, sealions, and walruses (Wyss, 1987; Bininda-Emonds et al., 1999; Flynn
and Wesley-Hunt, 2005; Bininda-Emonds et al., 2007). Most, if not all, carnivoran
classifications include pinnipeds with dogs and dog-like animals (Simpson, 1945;
McKenna and Bell, 1997; Flynn and Wesley-Hunt, 2005).
The carnivorans examined here include two common terrestrial species (Canis
familiaris and Felis catus), as well as the aquatic Stellar sea lion, Eumetopias jubatus
(Pinnipedia). The sea lion is a much larger animal, with an average body mass of 735 kg
(Loughlin et al., 1987), than either the cat (3.4 kg; Silva and Downing, 1995) or dog
270
(upwards of 31 kg; Galvão, 1947). The number of CT slices obtained through the ear
regions of Felis (627 slices) and Eumetopias (498 slices) are is significantly greater than
the number obtained for Canis (92 slices). Because of this, the imagery through the ear
region of Canis is of a lower resolution than that of Felis and Eumetopias, and minute
features of the inner ear of the dog are not discernable (such as the bony channel for the
aqueduct of the cochlea). The inner ear of Canis is imaged in Figures 5.34-5.35, the inner
ear of Eumetopias is depicted in Figures 5.36-5.37, and that of Felis in Figures 5.38-5.39.
The total volume of the inner ear cavities of the Canis specimen used in this study
(31.36 mm
3
) is less than that computed for both the Felis (45.78 mm
3
) and Eumetopias
specimens (138.60 mm
3
). In fact, the volume of the cochlea of the cat is 31.12 mm
3
,
which nearly is the volume of the bony labyrinth of the dog as a whole. The specimen of
Canis used is from a small dog (a Chihuahua), which might explain the size difference.
Even so, the percent of the total inner ear volume that is the cochlea is similar between
the two terrestrial carnivorans, as the cochlea of Felis contributes 68.0% of the volume,
and the cochlea of Canis contributes 66.1% (volume of cochlea equals 20.72 mm
3
). The
cochlea of Eumetopias contributes less to the labyrinth (53.5%; cochlear volume equals
74.17 mm
3
). Similarly, the length of the inner ear cavity of the dog (8.10 mm) is not
much different than that measured for the cat (8.91 mm) and the length of the bony
labyrinth of Eumetopias is substantially greater (13.71 mm) than either of the other
species.
The length of the cochlear canal in Eumetopias (19.25 mm) is greater than either
the dog (13.85 mm) or the cat (16.76 mm), and the cochlea of the dog completes a greater
degree of coiling (1155.6°) than the cat (1091.8°) and especially the sea lion (795.4°). A
secondary bony lamina is observed extending along the radial wall of the cochlea in all
three taxa, although the lamina persists for a greater relative distance in Eumetopias
271
FIGURE 5.34. Bony labyrinth of Canis familiaris. A, stereopair and labeled line drawing
of digital endocast in anterior view; B, stereopair and labeled line drawing of digital
endocast in dorsal view; C, stereopair and labeled line drawing of digital endocast in
lateral view; D, line drawing of cochlea viewed down axis of rotation to display degree of
coiling; E, line drawing of cochlea in profile. Abbreviations: aa, anterior ampulla; ac,
anterior semicircular canal; ant, anterior direction; cc, canaliculus cochleae for aqueduct
of cochlea; co, cochlea; cr, common crus; dor, dorsal direction; er, elliptical recess of
vestibule; fc, fenestra cochleae; fv, fenestra vestibuli; la, lateral ampulla; lc, lateral
semicircular canal; med, medial direction; pa, posterior ampulla; pc, posterior
semicircular canal; pl, primary bony lamina; pos, posterior direction; scr, secondary
common crus; sl, secondary bony labyrinth; sr, spherical recess of vestibule.
272
A
B
C
D E
co
sl
pl
1 mm
fc
fv
sr
la
aa
cr
ac
pc
lc
scr
co
pl
cc
pa
scr
pc
fc
lc
la
ac
aa
er
co
aa
la
cr
ac
lc
pc
scr
pa
cc
fv
fc
fc
ps
co
co
cc
fc
1 mm
1 mm
dor
med
med
pos
dor
ant
273
FIGURE 5.35. CT slices through ear region of Canis familiaris. Abbreviations: aa,
anterior ampulla; ac, anterior semicircular canal; cc, canaliculus cochleae for aqueduct of
cochlea; cn, canal for cranial nerve VIII; co, cochlea; cr, common crus; dor, dorsal
direction; fc, fenestra cochleae; fn, canal for cranial nerve VII; fv, fenestra vestibuli; iam,
internal auditory meatus; la, lateral ampulla; lat, lateral direction; lc, lateral semicircular
canal; pa, posterior ampulla; pc, posterior semicircular canal; pl, primary bony lamina;
pos, posterior direction; scr, secondary common crus; sg, canal for spiral ganglion within
primary bony lamina; st, stapes within fenestra vestibuli; vb, vestibule.
274
pos
lat
11
17
23
29
35
41
47
53
59
71
65
5 mm
dor
co
1 mm
lat
11 17
2923 35
4741 53
6559 71
co
co
co
fv
st
aa
ac
lc
lc
lc
cc
cr
pa
iam
fc
fc
la
cn
cn
cn
fn
fn
cn
vb
vb
vb
pl
pl
pl
cr
scr
cr
pc
pc
ac
ac
ac
275
FIGURE 5.36. Bony labyrinth of Eumetopias jubatus. A, stereopair and labeled line
drawing of digital endocast in anterior view; B, stereopair and labeled line drawing of
digital endocast in dorsal view; C, stereopair and labeled line drawing of digital endocast
in lateral view; D, line drawing of cochlea viewed down axis of rotation to display degree
of coiling; E, line drawing of cochlea in profile. Abbreviations: aa, anterior ampulla; ac,
anterior semicircular canal; ant, anterior direction; av, bony channel for aqueduct of
vestibule; cc, canaliculus cochleae for aqueduct of cochlea; co, cochlea; cr, common
crus; dor, dorsal direction; fc, fenestra cochleae; fv, fenestra vestibuli; la, lateral ampulla;
lc, lateral semicircular canal; med, medial direction; pa, posterior ampulla; pc, posterior
semicircular canal; pl, primary bony lamina; pos, posterior direction; ps, outpocketing for
perilymphatic sac; sl, secondary bony labyrinth; sr, spherical recess of vestibule.
276
A
B
C
D E
co
sl
sr
cr
aa
la
fv
fc
lc
pc
ac
cc
co
sl
pl
av
cr
aa
ac
la
lc
pa
pc
co
fv
sr
cr
er
lc
aa
pa
av
pc
ac
fc
cc
la
5 mm
5 mm
fc
co
co
ps
fc
cc
sl
dor
med
med
pos
dor
ant
277
FIGURE 5.37. CT slices through ear region of Eumetopias jubatus. Abbreviations: aa,
anterior ampulla; ac, anterior semicircular canal; av, bony channel for aqueduct of
vestibule; cf, foramina in cribriform plate; cn, canal for cranial nerve VIII; co, cochlea;
cr, common crus; dor, dorsal direction; er, elliptical recess; fc, fenestra cochleae; fn,
canal for cranial nerve VII; fv, fenestra vestibuli; la, lateral ampulla; lat, lateral direction;
lc, lateral semicircular canal; pa, posterior ampulla; pc, posterior semicircular canal; pl,
primary bony lamina; pos, posterior direction; sg, canal for spiral ganglion within
primary bony lamina; sl, secondary bony lamina; sr, spherical recess of vestibule; st,
stapes fallen into vestibule; vn, canal for vestibular branch of cranial nerve VIII.
278
pos
lat
95
129
163
197
231
265
299
333
367
435
401
1 mm
dor
co
co
co
fc
fc
fv
sl
sr
sr
sr
pl
cn
cn
vn
vn
fn
la
lc
er
aa
ac
ac
pa
pc
pc
pc
pc
cr
ac
sa
pl
sg
sg
cf
lat
1 mm
95 129
197163 231
299265 333
401367 435
sg
sg
st
av
av
lc
lc
pl
fv
279
FIGURE 5.38. Bony labyrinth of Felis catus. A, stereopair and labeled line drawing of
digital endocast in anterior view; B, stereopair and labeled line drawing of digital
endocast in dorsal view; C, stereopair and labeled line drawing of digital endocast in
lateral view; D, line drawing of cochlea viewed down axis of rotation to display degree of
coiling; E, line drawing of cochlea in profile. Abbreviations: aa, anterior ampulla; ac,
anterior semicircular canal; ant, anterior direction; av, bony channel for aqueduct of
vestibule; cc, canaliculus cochleae for aqueduct of cochlea; co, cochlea; cr, common
crus; dor, dorsal direction; er, elliptical recess of vestibule; fc, fenestra cochleae; fv,
fenestra vestibuli; la, lateral ampulla; lc, lateral semicircular canal; med, medial
direction; pa, posterior ampulla; pc, posterior semicircular canal; pl, primary bony
lamina; pos, posterior direction; ps, outpocketing for perilymphatic sac; sl, secondary
bony labyrinth; sr, spherical recess of vestibule.
280
A
B
C
D E
co
sl
fv
fc
la
aa
ac
cr
av
pc
pa
lc
ps
sr
er
co
cc
av
pc
pa
la
aa
cr
pl
sl
fc
ac
lc
1 mm
fc
fv
co
aa
la
lc
cr
ac
pc
av
pa
cc
ps
1 mm
1 mm
ps
fc
cc
co
co
sl
ps
fc
cc
dor
med
med
pos
dor
ant
281
FIGURE 5.39. CT slices through ear region of Felis catus. Abbreviations: aa, anterior
ampulla; ac, anterior semicircular canal; av, bony channel for aqueduct of vestibule; cc,
canaliculus cochleae for aqueduct of cochlea; cf, foramina in cribriform plate; cn, canal
for cranial nerve VIII; co, cochlea; cr, common crus; dor, dorsal direction; fc, fenestra
cochleae; fn, canal for cranial nerve VII; fv, fenestra vestibuli; hf, hiatus Fallopii for exit
of greater petrosal nerve; la, lateral ampulla; lat, lateral direction; lc, lateral semicircular
canal; pa, posterior ampulla; pc, posterior semicircular canal; pl, primary bony lamina;
pos, posterior direction; sg, canal for spiral ganglion within primary bony lamina; sl,
secondary bony lamina; sr, spherical recess of vestibule; vb, vestibule.
282
pos
dor
84
115
146
177
208
239
270
301
332
394
363
2 mm
lat
dor
fn
fn
fn
fn
fn
fn
cn
cn
hf
co
co
co
co
cn
sg
sl
sl
cn
pl
cr
av
ac
ac
sg
pl
cf
pl
pl
fv
sr
sr
av
av
cr
aa
aa
la
lc
lc
pa
pc
pc
pc
pc
pc
vb
cc
fc
sl
pl
fn
1 mm
84 115
177146 208
270239 301
363332 394
283
(248.6°) and Felis (242.6°) than what was measured for Canis (104.0°). The aspect ratios
of the cochleae of carnivorans are high relative to other species described above. The
ratio of the cat (0.69) is higher than that of the dog (0.64), and the ratio calculated for the
sea lion is intermediate (0.68). Other species with high aspect ratios include Sus and
Macroscelides (ratios around 0.71 for both taxa).
The basal whorl of each carnivoran cochlea is separated from the apical turns,
where the apical turns fit within the arc created by the basal turn when the cochlea is in
vestibular view. The apical turns sit upon one another in all three taxa, but each
successive whorl is smaller than the turn immediately basal to it, forming a pointed cone
beyond the basal whorl. The scala tympani is expanded posterior to the fenestra cochleae
in all carnivoran taxa. The expansion of the scala tympani leads to the bony canaliculus
cochleae for the aqueduct of the cochlea. The canaliculus is longer and more robust in
Felis (3.60 mm) than in Canis (2.08 mm), although the longest canaliculus was measured
for Eumetopias (4.16 mm).
The angle between the planes of the basal turn of the cochlea and the lateral
semicircular canal in Canis (20.8°) is similar to that measured for the cetaceans (21.3° in
Tursiops; 23.2° in the balaenopterid) and Sus (23.8°). However, the angle is greater in
Felis (45.8°), being more similar to the hyrax, Procavia capensis (45.4°), than to the dog.
The angle formed between the cochlea and lateral canal of Eumetopias (31.6°) is
intermediate between the dog and cat, most closely resembling Orycteropus (31.9°). The
scalae tympani and vestibuli bend around the dorsal border of the fenestra cochleae in the
carnivorans. The fenestra vestibuli of Felis is elliptical (stapedial ratio equals 1.9),
although the fenestra of Canis is distinctly more circular (ratio equals 1.3). The fenestra
vestibuli of Eumetopias is intermediate between the other carnivorans (1.7).
284
The spherical recess of the vestibule is separated from the elliptical recess by a
constriction of the vestibule in the carnivoran species. The elliptical recess is gently
curved, forming a secondary excavation at its anterior end for the vestibular apertures of
the anterior and lateral ampullae in Canis and Felis, but not in Eumetopias. Rather, the
elliptical recess of Eumetopias is concave laterally. The anterior excavation of the
elliptical recess of the vestibule in both Canis and Felis is expressed as a pedestal for the
ampullae in the digital endocasts. The anterior ampulla of Eumetopias forms a teardrop-
shaped structure, although the lateral ampulla is deflated and dorsoventrally compressed
in this taxon.
The bony channel for the aqueduct of the vestibule was not observed in Canis,
owing to the inadequate resolution of the CT data. The structure is observed in both
Eumetopias and Felis, in which the bony channel opens ventral to the medial edge of the
vestibular aperture of the common crus in both taxa, and bends laterally along its course.
As is the case in many of the species described here, the channel for the aqueduct ends in
a flattened fissure. The length of the channel is longer in Felis (3.77 mm) than that
measured for Eumetopias (2.26 mm).
The posterior limb of the lateral semicircular canal opens into the posterior
ampulla, rather than the vestibule itself, dorsal to the anterior edge of the vestibular
aperture of the posterior ampulla in Felis. The position of the lateral semicircular canal in
Felis is high relative to the other vestibular components. When the endocast of the bony
labyrinth of Felis is viewed anteriorly, the lateral canal crosses the space enclosed by the
posterior semicircular canal (sagittal labyrinthine index equals 13.1). The lateral
semicircular canal of Canis is situated in a lower position than in Felis, and the plane of
the canal does not cross the space enclosed by the posterior canal in anterior view. The
lateral semicircular canal of Eumetopias empties into the posterior ampulla, as in the cat,
285
but the plane of the canal does not take a high position in the sea lion as it does in the cat.
In this manner, the vestibular apparatus of Eumetopias appears more similar to Canis
among the carnivorans examined here.
As in the cat and sea lion, the posterior limb of the lateral semicircular canal does
not open into the vestibule in Canis. Unlike the other taxa, the lateral canal of Canis does
not open into the posterior ampulla directly either, but rather the lateral canal is fused
with the posterior semicircular canal to form a secondary common crus. A secondary crus
is developed in the aardvark (Orycteropus afer) and also in non-placental mammals.
The greatest angle between the planes of any two semicircular canals in
Eumetopias was measured between the anterior and posterior semicircular canals (105°).
The angle between the canals is also obtuse in Canis (101°). The angle between the
anterior and posterior canals in Felis is closer to a right angle (91.4°), although the
posterior and lateral semicircular canal planes nearly are perpendicular in Canis (89.1°)
and Eumetopias (90.6°). The angle between the planes of the posterior and lateral
semicircular canals of Felis is 96.7°, which is the greatest angle between any two canals
in the cat. The angle between the anterior and lateral semicircular canals is 80.4° in
Canis, 79.7° in Eumetopias, and 76.8° in Felis.
The lateral semicircular canal is the least planar of the three canals in Eumetopias,
with a total angular deviation of 16.4°. In fact, the lateral canal of Eumetopias is the least
planar of any semicircular canal measured for any carnivoran specimen examined here
(4.4° for Felis; 5.1° for Canis). The posterior semicircular canal of Canis is the least
planar of its three canals, with a total angular deviation from the plane of the canal of
10.8°. However, the posterior canal of Felis does not deviate from its plane in any
significant manner, but the total angular deviation calculated for the posterior canal of
Eumetopias is 9.5°. The deviation of the anterior canal is smaller in F. catus (4.5°) than
286
in Canis (6.0°), but not by much. The anterior semicircular canal of Eumetopias deviates
by a miniscule amount (0.8°). None of the canals in either of the terrestrial species
deviate significantly from their respective planes. The ratios of the total linear deviation
over cross-sectional diameter of the anterior, lateral, and posterior semicircular canals of
Canis are 0.59, 0.40, and 0.94 respectively, and the ratios for the same canals in Felis are
0.57, 0.51, and 0.00 (planar). The anterior canal of Eumetopias does not deviate from its
plane significantly (linear deviation to lumen diameter ratio equals 0.11), although the
lateral and posterior semicircular canals of the sea lion deviate by a substantial amount
(ratio of lateral equals 1.70; posterior equals 1.15).
The semicircular canals of both terrestrial taxa form graceful curves along their
courses. As with most dimensions within the bony labyrinth, the semicircular canals of
Felis are larger than the canals of Canis, although a common pattern across the canal arc
radii is observed in both taxa. Namely, the radius of the arc of the anterior semicircular
canal is the greatest for both Canis (1.73 mm; lateral equals 1.57 mm; posterior equals
1.43 mm) and Felis (1.92 mm; lateral equals 1.68 mm; posterior equals 1.91 mm).
However, the lateral semicircular canal is the largest in terms of radius in Eumetopias
(3.13 mm; anterior equals 3.00 mm; posterior equals 2.86 mm).
The posterior semicircular canal of the cat is the longest of all of the canals in this
species (9.39 mm), and the lateral canal is the shortest (7.48 mm; posterior equals 8.78
mm). Likewise, the lateral semicircular canal of Canis is the shortest of its canals (7.08
mm), but its anterior canal is the longest (8.57 mm), rather than its posterior canal (7.34
mm) as was observed in Felis. Unlike the terrestrial carnivorans, the lateral canal of
Eumetopias is the longest (14.70 mm; anterior equals 12.99 mm; posterior equals 14.08
mm). The lateral semicircular canal of Canis may be the shortest of the three canals in
this species, but the lumen of the lateral canal has the greatest cross-sectional diameter in
287
the dog (0.35 mm; anterior equals 0.31 mm; posterior equals 0.29mm). Similarly, the
lateral canal has the greatest diameter in Eumetopias. All of the canals of Felis were
equal in cross-sectional diameter (0.26 mm).
The aspect ratio of the arc of the lateral semicircular canal is highest in both Canis
(1.01) and Felis (1.04), and the aspect ratio is the smallest for the anterior canal arc for
both species (0.82 and 0.77 respectively). The aspect ratio of the posterior semicircular
canal arc is 0.98 for Canis and 1.01 for Felis. The aspect ratios of all three canal arcs are
larger in Eumetopias than those measured for the terrestrial carnivorans (anterior equals
0.96; lateral equals 1.24; posterior equals 1.18). The ratio of the slender canal length over
arc radius of the posterior semicircular canal was the greatest among the three canals in
Canis (5.14; anterior equals 4.97; lateral equals 4.45), Eumetopias (4.92; anterior equals
4.33; lateral equals 4.72), and Felis (4.93; anterior equals 4.57; lateral equals 4.45).
Two labyrinthine characters are synapomorphies for Carnivora within Ferae. The
first is the higher aspect ratio of the carnivoran cochlea in profile that gives the cochlear
spiral the “sharp-pointed” condition described by Gray (1907, 1908). The second
synapomorphy is the entrance of the lateral canal into the posterior ampulla, which is
observed in Eumetopias and Felis. The secondary common crus observed in Canis is an
apomorphic reversal to the ancestral therian condition, and it also is observed in
Orycteropus among crown placentals. The ancestral coiling of the cochlea of Carnivora is
over a quarter of a turn greater than that reconstructed for the ancestor of Ferae (986.5°
versus 888.2°), and the carnivoran cochlea contributes 5% more to the total labyrinthine
volume than that of the feran ancestor (61% versus 56%).
A single character from the bony labyrinth, reversal to a low position of the lateral
semicircular canal in relation to the ampullar entrance of the posterior canal, is optimized
as a synapomorphy for Caniformia. The lateral canal of Felis is positioned high, which is
288
derived from the ancestral eutherian condition, but is plesiomorphic for Carnivora as a
whole. The low position of the lateral canal is a reversal for Caniformia. The lateral
semicircular canal enters the posterior ampulla in the ancestral caniform (even though a
secondary common crus is present in Canis), and the arc of the anterior semicircular
canal is the largest of the three canal arcs (even though the lateral canal arc is the largest
in Eumetopias). The ancestral labyrinth of Caniformia possesses a cochlea with a high
aspect ratio that coiled 979.3° and contributed 60% of the total labyrinthine volume.
Pholidota
Although extant species of pangolins are known only from Africa and Asia,
fossils of Pholidota have been recovered from Tertiary deposits of Europe and North
America (McKenna and Bell, 1997; Rose et al., 2005). Pangolins have not contributed
greatly to the mammalian biota throughout time (Rose et al., 2005), nor is Pholidota a
taxonomically diverse group at present, with only eight species recognized within the
single genus Manis (Reeder et al., 2007).
The gross volume of the inner ear cavities of the pangolin, Manis tricuspis,
examined in this study (28.53 mm
3
) is similar to that of Bathygenys reevesi (29.83 mm
3
),
Dasypus novemcinctus (26.48 mm
3
), and Canis familiaris (31.36 mm
3
). Likewise, the
lengths of the bony labyrinth are not vastly different (6.66 mm for Manis; 7.40 mm for
Bathygenys; 8.06 mm for Dasypus; 8.10 mm for Canis). The overall body size of Manis
is similar to Dasypus (4.5 versus 4,8 k; Silva and Downing, 1995), which is reflected in
the dimensions of the inner ear. The volume of the cochlea in Manis is 14.00 mm
3
, which
is 49.1% of the total volume of the bony labyrinth. The cochlear contribution measured in
Manis is closer to that calculated for Bathygenys (54.2%) than Canis (66.1%).
289
The cochlea completes over two and one third turns (863°), and the length of the
canal is 9.64 mm. The aspect ratio of the cochlear spiral is 0.54, and the apical whorls of
the cochlea sit upon the basal turn, rather than fit within the basal turn (Figure 5.40) as is
observed in cetaceans. A secondary bony lamina is not developed within the cochlea of
Manis (Figure 5.41), and the angle between the plane of the basal turn of the cochlea and
the lateral semicircular canal is 20.3°. The secondary lamina also is absent in the
terrestrial cetartiodactyls Bathygenys and Sus.
The scala tympani of the cochlea is expanded internal to the fenestra cochleae.
The excavation of the scala tympani leads to a robust canaliculus cochleae for the
aqueduct of the cochlea. The canaliculus is oriented dorsally and takes a straight course
as it extends to the external surface of the petrosal (2.85 mm in length). The external
aperture for the aqueduct of the cochlea is shared by a second canal, which empties into
the lateral aspect of the canaliculus cochleae proximal to the midpoint of the canaliculus.
The second canal is not straight. The canal opens lateral to the canaliculus cochlea and
hooks dorsally to join the canaliculus. A curved canal that fuses to the canaliculus
cochleae is not observed in any other mammal.
A slight constriction of the vestibule dorsal to the fenestra vestibuli (stapedial
ratio of 1.7 calculated from fenestra) is the only separation between the spherical and
elliptical recesses. Otherwise, the vestibule is a single and undivided unit. The bony
channel for the aqueduct of the vestibule exits the vestibule medial to the vestibular
opening of the short and stout common crus. The 2.45 mm long channel terminates in a
spatulate shaped fissure.
The posterior limb of the lateral semicircular canal opens into the vestibule
anterodorsal to the vestibular aperture of the posterior ampulla. Because the lateral canal
empties directly into the vestibule at its posterior end, there is not a secondary common
290
FIGURE 5.40. Bony labyrinth of Manis tricuspis. A, stereopair and labeled line drawing
of digital endocast in anterior view; B, stereopair and labeled line drawing of digital
endocast in dorsal view; C, stereopair and labeled line drawing of digital endocast in
lateral view; D, line drawing of cochlea viewed down axis of rotation to display degree of
coiling; E, line drawing of cochlea in profile. Abbreviations: aa, anterior ampulla; ac,
anterior semicircular canal; ant, anterior direction; av, bony channel for aqueduct of
vestibule; cc, canaliculus cochleae for aqueduct of cochlea; co, cochlea; cr, common
crus; dor, dorsal direction; er, elliptical recess of vestibule; fc, fenestra cochleae; fv,
fenestra vestibuli; la, lateral ampulla; lc, lateral semicircular canal; med, medial
direction; pa, posterior ampulla; pc, posterior semicircular canal; pl, primary bony
lamina; pos, posterior direction; ps, outpocketing for perilymphatic sac; sr, spherical
recess of vestibule.
291
A
B
C
D E
co
la
aa
sr
er
cr
fv fc
pa
lc
ac
av
pc
co
pl
cc
av
cr
er
aa
lc
1 mm
ac
pc
co
fv
fc
cc
av
cr
pc
aa
la
ps
lc
er
pa
sr
ac
1 mm
1 mm
co
fc
co
ps
ps
cc
fc
dor
med
med
pos
dor
ant
292
FIGURE 5.41. CT slices through ear region of Manis tricuspis. Abbreviations: aa,
anterior ampulla; ac, anterior semicircular canal; av, bony channel for aqueduct of
vestibule; cc, canaliculus cochleae for aqueduct of cochlea; cn, canal for cranial nerve
VIII; co, cochlea; dor, dorsal direction; fc, fenestra cochleae; fn, canal for cranial nerve
VII; fv, fenestra vestibuli; iam, internal auditory meatus; la, lateral ampulla; lat, lateral
direction; lc, lateral semicircular canal; med, medial direction; pa, posterior ampulla; pc,
posterior semicircular canal; pl, primary bony lamina; pos, posterior direction; sr,
spherical recess of vestibule; vb, vestibule; vn, canal for vestibular branch of cranial
nerve VIII.
293
pos
dor
17
23
29
35
41
47
53
59
65
77
71
2 mm
med
co
co
co
co
dor
1 mm
17 23
3529 41
5347 59
7165 77
pl
pl
fn
sg
cn
cn
vn
iam
vb
vb
cc
pc
cc
cc
ac
vb
av
pl
aa
cc
ac
lc
ac
sr
fc
aa
la
vb
fv
cn
lc
pa
pc
pc
pc
ac
294
crus between the lateral and posterior semicircular canals. The position of the plane of the
lateral semicircular canal, intersecting points along the lumen of the canal, is high relative
to the rest of the vestibular elements, and the sagittal labyrinthine index of Manis is 20.5.
The three semicircular canals are especially thick relative to the size of the labyrinth,
including the lateral canal, which makes the high position of the canal plane difficult to
observe.
The angle between the planes of the posterior and lateral semicircular canals
approach a right angle (88.6°), and the other angles between canal planes are acute
(anterior-lateral equals 77.0°; anterior-posterior canal equals 84.8°). The lateral
semicircular canal of Manis does not deviate from its average plane by any significant
degree, and the total angular deviations of the anterior (6.7°) and posterior (7.3°) are not
great. The degrees of deviation measured for both the anterior and posterior semicircular
canals are not significant (ratios of total linear deviation over cross-sectional diameter are
0.31 and 0.40 respectively).
As stated above, the semicircular canals of Manis are thick relative to the total
size of the bony labyrinth. As a comparison, the total length of the bony labyrinth of
Manis (6.66 mm) is not much less than that of the oreodont Bathygenys (7.40), although
the cross-sectional diameter of the anterior, lateral, and posterior semicircular canals of
the pangolin (0.55 mm, 0.62 mm, and 0.52 mm respectively) are substantially greater
than that measured for the oreodont (0.44 mm, 0.33 mm, and 0.34 mm for the same
canals). However, a similar pattern is not observed in the slender semicircular canal
lengths and radii of the semicircular canal arcs, which are smaller in Manis than
Bathygenys. The radii of the anterior, lateral, and posterior semicircular canals of Manis
are 1.46 mm, 1.06 mm, and 1.66 mm respectively (radii for Bathygenys are 1.91 mm,
1.52 mm, and 1.79 mm), and the lengths of the canals are 6.59 mm, 3.71 mm, and 7.03
295
mm for the anterior, lateral, and posterior canals (lengths for Bathygenys are 9.72 mm,
7.11 mm, and 10.01 mm).
The lowest aspect ratio of a semicircular canal arc was calculated for the anterior
canal (0.76). The ratio for the lateral semicircular canal arc is 0.82, and a ratio of 0.93
was computed for the posterior arc. The ratio of the slender canal length to the arc radius
is largest for the anterior semicircular canal (4.52). The ratio for the posterior canal is
4.23, and the ratio for the lateral canal in Manis is 3.49.
The bony labyrinth of Manis inherited a direct entry of the lateral semicircular
canal into the vestibule from the ancestral placental, and the high position of the plane of
the lateral semicircular canal compared to the posterior canal is retained from the
ancestor of Boreoeutheria. The low aspect ratio of the cochlea observed in Manis is the
same as that reconstructed for the ancestor of Eutheria. Because the state of the cochlea
could not be reconstructed for the ancestor of Placentalia, the condition in Manis is either
a primitive retention or a secondary reversal. The arc of the posterior semicircular canal
of Manis is the largest among the three arcs, which is derived relative to both the ancestor
of Boreoeutheria, as well as the most recent common ancestor of Pholidota and Carnivora
(for which the anterior arc is the largest).
There are no unambiguous synapomorphies within the inner ear that support an
exclusive Carnivora plus Pholidota clade (Ferae). The ancestor of the clade retained
features that were present in the ancestor of Placentalia, including entry of the lateral
semicircular canal into the vestibule directly, and an anterior semicircular canal arc that
was the largest among the three arcs (also present in the ancestor of Theria). The plane of
the lateral semicircular canal of the ancestor of Ferae was high compared to the ampullar
entrance of the posterior canal, which was the state reconstructed for the ancestor of
296
Boreoeutheria. The state of the aspect ratio of the cochlea was equivocal as reconstructed
for the feran ancestor.
Megachiroptera
Chiroptera (bats) is the only group of truly volant mammals, and with over 1,000
species, it forms the second most speciose group of mammals (second only to rodents).
Bats traditionally are separated into Megachiroptera, which includes non-echolocating
bats and potentially one group of echolocators (see below), and Microchiroptera, which
only includes echolocating bats. The results of several recent molecular studies support a
closer relationship between Pteropidae (of which Pteropus lyelli is used as a
representative; Figures 5.42-5.43) and the echolocating Rhinolophidae (of which
Rhinolophus ferrumequinum was examined here) than between Rhinolophidae and other
echolocating bats, which are represented here by the Nycteridae species Nycteris grandis
and the Molossidae species Tadardia brasiliensis (Teeling et al., 2000, 2005; Simmons et
al., 2008). However, the results of Bininda-Emonds et al. (2007) separates Pteropidae as
the sister taxon to all other bats. Because the morphological descriptions of the bony
labyrinth are organized based on the relationships recovered by Bininda-Emonds et al.
(2007) in the present study, description of the inner ear of Rhinolophus is included with
Nycteris and Tadarida.
Pteropus includes the bats with the largest body sizes (Silva and Downing, 1995),
and the average body mass of Pteropus lyelli (435 g) is an order of magnitude larger than
that of the microchiroperan species examined (see Table 5.2). The total length of the
bony labyrinth (Figure 5.42) of Pteropus is 6.19 mm, and the gross volume of the inner
ear cavities is 7.01 mm
3
. The cochlea contributes 58.9% of the total labyrinthine volume
(4.13 mm
3
). The cochlear canal is 7.66 mm in length, and completes over one and three
297
FIGURE 5.42. Bony labyrinth of Pteropus lyelli. A, stereopair and labeled line drawing
of digital endocast in anterior view; B, stereopair and labeled line drawing of digital
endocast in dorsal view; C, stereopair and labeled line drawing of digital endocast in
lateral view; D, line drawing of cochlea viewed down axis of rotation to display degree of
coiling; E, line drawing of cochlea in profile. Abbreviations: aa, anterior ampulla; ac,
anterior semicircular canal; ant, anterior direction; av, bony channel for aqueduct of
vestibule; cc, canaliculus cochleae for aqueduct of cochlea; co, cochlea; cr, common
crus; dor, dorsal direction; er, elliptical recess of vestibule; fc, fenestra cochleae; fv,
fenestra vestibuli; la, lateral ampulla; lc, lateral semicircular canal; med, medial
direction; pa, posterior ampulla; pc, posterior semicircular canal; pl, primary bony
lamina; pos, posterior direction; ps, outpocketing for perilymphatic sac; sa, subarcuate
fossa; sg, canal for spiral ganglion within primary lamina; sl, secondary bony lamina; sr,
spherical recess of vestibule; st, stapes within fenestra vestibuli.
298
A
B
C
D E
sl
co
1 mm
pl
fc
fv
sr
la
aa
av
pa
lc
cr
pc
ac
co
sl
pl
ps
cc
av
cr
er
fc
aa
la
ac
pc
co
sl
cc
fc
fv
ps
aa
la
er
lc
lc
pa
pc
av
cr
ac
1 mm
1 mm
co
cc
fc
ps
cc
fc
co
sl
dor
med
med
pos
dor
ant
299
FIGURE 5.43. CT slices through ear region of Pteropus lyelli. Abbreviations: aa, anterior
ampulla; ac, anterior semicircular canal; av, bony channel for aqueduct of vestibule; cn,
canal for cranial nerve VIII; co, cochlea; cr, common crus; dor, dorsal direction; er,
elliptical recess; fc, fenestra cochleae; fn, canal for cranial nerve VII; la, lateral ampulla;
lc, lateral semicircular canal; med, medial direction; pa, posterior ampulla; pc, posterior
semicircular canal; pl, primary bony lamina; pos, posterior direction; sr, spherical recess
of vestibule.
300
pos
dor
22
36
50
64
78
92
106
120
134
162
148
1 mm
med
dor
co
co
pl
1 mm
22 36
6450 78
10692 120
148134 162
cn
sg
fn
av
st
av
st
cn
vn
sa
sa
sa
sa
sa
sa
sr
er
er
er
ac
ac
ac
ac
ac
cr
cr
lc
lc
lc
la
la
fc
fc
pc
pc
pc
pc
pa
aa
aa
pl
co
301
quarters turns (656°). The secondary bony lamina is present (Figure 5.43) and persists for
nearly the entire basal turn (335°).
The aspect ratio of the cochlea in Pteropus is 0.61, and the cochlear canal itself
appears narrow compared to the labyrinth as a whole. The apical turnings of the cochlea
fit within the basal whorl when the cochlea is viewed down its axis of rotation. The angle
between the planes of the basal turn of the cochlea and the lateral semicircular canal is
36.2°, and the stout and straight canaliculus cochleae for the aqueduct of the cochlea is
1.62 mm in length.
The stapedial ratio of Pteropus, as measured from the fenestra vestibuli, is 1.8.
The spherical and elliptical recesses are undivided, although an excavation is present at
the anterior end of the vestibule, which is expressed as a pedestal for the anterior and
lateral ampullae on the endocast. The common crus and semicircular canals are delicate
and form graceful curves in Pteropus. The bony channel for the aqueduct of the vestibule
exits the inner ear cavities medial to the vestibular aperture of the common crus. The
channel, which is 0.73 mm in length, extends posteriorly before terminating in a
triangular-shaped fissure. The posterior limb of the lateral semicircular canal enters the
vestibule dorsal to the vestibular aperture of the posterior ampulla, giving the plane of the
lateral semicircular canal a relatively high position. The sagittal labyrinthine index of
Pteropus is 29.7.
The planes of the posterior and lateral semicircular canals essentially form a right
angle (90.4°), while the angle between the anterior and lateral canals is acute (84.9°) and
the angle between the anterior and posterior canal is obtuse (98.3°). None of the
semicircular canals fit on a single plane, although the total angular deviation of the
posterior canal (4.7°) is not significant (ratio of total linear deviation over cross-sectional
diameter is 0.52). On the other hand, the anterior (10.3°) and lateral (14.3°) semicircular
302
canals of Pteropus deviate from their average planes by a significant amount (ratios of
linear deviation over canal diameter are 1.63 and 1.35 respectively).
The radius of the arc of the anterior semicircular canal (1.57 mm) is greater than
that measured for the lateral (1.28 mm) and posterior canals (1.35 mm). This differs from
the other dimensions of the semicircular canals in Pteropus. For example, the slender
canal length of the posterior semicircular canal (7.03 mm) is greater than either the
anterior (6.86 mm) or lateral canal (5.86 mm). Although the lateral semicircular canal is
the smallest of the three in terms of slender canal length and arc radius, the cross-
sectional diameter of the lateral canal (0.24 mm) is greater than that measured for either
the anterior (0.17 mm) or posterior semicircular canal (0.21 mm).
The aspect ratio of the lateral semicircular canal arc of Pteropus is the highest
among the three canals (0.97; anterior equals 0.94; posterior equals 0.85). The high ratio
of the lateral semicircular canal arc indicates that the height and width of the arc are
nearly identical. The ratio of the slender canal length to the canal arc radius of the
posterior semicircular canal is the greatest (5.20; ratio of anterior equals 4.37; ratio of
lateral equals 4.56).
The ancestor of Chiroptera retained ancestral placental features, including a
lateral semicircular canal that was positioned high compared to the posterior canal and
that opened into the vestibule directly, as well as the anterior semicircular canal arc as the
largest among the three arcs. The ancestral chiropteran cochlea had a high aspect ratio (a
condition shared with Carnivora), coiled 763.8°, and contributed 61% to the overall
volume of the inner ear cavities (also shared with Carnivora). The labyrinth of Pteropus
retains all discrete character states from its chiropteran ancestor, but the cochlea of
Pteropus coils to a lesser degree (656.0°).
303
Microchiroptera
Among the microchiropteran species examined in the present study, Nycteris
grandis (Figures 5.44-5.45), Rhinolophus ferrumequinum (Figures 5.46-5.47), and
Tadarida brasiliensis (Figures 5.48-5.49), the species with the largest body mass (as
reported in Silva and Downing, 1995) is Nycteris (29.3 g; mass of Rhinolophus equals
17.2 g; mass of Tadarida equals 12.1 g). However, the bony labyrinth of Rhinolophus is
the largest, both in terms of gross volume of the inner ear cavities (5.90 mm
3
; volume of
Nycteris equals 2.13 mm
3
; volume of Tadarida equals 3.86 mm
3
), as well as the total
length of the labyrinth (3.76 mm; length of Nycteris equals 3.39 mm; length of Tadarida
equals 3.22 mm). Likewise, the volume of the cochlea of Rhinolophus (2.80 mm
3
) is
greater than that measured for both Nycteris (1.42 mm
3
) and Tadarida (2.80 mm
3
).
The cochleae of the three species comprise over half of the total inner ear volume.
The cochlea of Nycteris contributes 66.5% of the total volume, which is similar to the
percentage calculated in Canis (66.1%) and Felis (68.0%). The cochlea comprises 72.5%
of the labyrinthine volume in Tadarida, which is similar to the percentage calculated in
the afrotherians Chrysochloris (71.3%), Macroscelides (71.7%), and Trichechus (71.1%).
The largest volumetric contribution among the bats examined was calculated for
Rhinolophus (88.8%). The only other mammals that have a larger contribution than
Rhinolophus are the cetaceans (contribution in Tursiops equals 93.5%; contribution in the
balaenopterid equals 90.6%).
The cochlea of Rhinolophus (Figure 5.46) completes just over three complete
turns (1114°), whereas the cochlea of Nycteris (Figure 5.44) and Tadarida (Figure 5.48)
complete around two and one quarter turns (795° and 752° respectively). Likewise, the
length of the cochlear canal of Rhinolophus (11.57 mm) is greater than that measured in
either Nycteris (6.66 mm) or Tadarida (6.95 mm). A secondary bony lamina is present in
304
FIGURE 5.44. Bony labyrinth of Nycteris grandis. A, stereopair and labeled line drawing
of digital endocast in anterior view; B, stereopair and labeled line drawing of digital
endocast in dorsal view; C, stereopair and labeled line drawing of digital endocast in
lateral view; D, line drawing of cochlea viewed down axis of rotation to display degree of
coiling; E, line drawing of cochlea in profile. Abbreviations: aa, anterior ampulla; ac,
anterior semicircular canal; ant, anterior direction; cc, canaliculus cochleae for aqueduct
of cochlea; co, cochlea; cr, common crus; dor, dorsal direction; fc, fenestra cochleae; fv,
fenestra vestibuli; la, lateral ampulla; lc, lateral semicircular canal; med, medial
direction; pa, posterior ampulla; pc, posterior semicircular canal; pos, posterior direction;
ps, outpocketing for perilymphatic sac; sl, secondary bony lamina.
305
A
B
C
D E
co
1 mm
1 mm
sl
pl
fv
fc
cc
aa
cr
ac
pc
lc
la
co
sl
pa
fc
fv
la
aa
ac
cr
lc
pc
co
fv
la
ac
aa
lc
cr
fc
sl
cc
pc
pa
co
fc
cc
sl
co
sl
cc
fc
ps
dor
med
med
pos
dor
ant
306
FIGURE 5.45. CT slices through ear region of Nycteris grandis. Abbreviations: aa,
anterior ampulla; ac, anterior semicircular canal; cc, canaliculus cochleae for aqueduct of
cochlea; cn, canal for cranial nerve VIII; co, cochlea; fc, fenestra cochleae; fv, fenestra
vestibuli; la, lateral ampulla; lc, lateral semicircular canal; med, medial direction; pa,
posterior ampulla; pc, posterior semicircular canal; pl, primary bony lamina; pos,
posterior direction; sa, subarcuate fossa; sl, secondary lamina; vb, vestibule; ven, ventral
direction.
307
pos
med
10
14
18
22
26
30
34
38
42
50
46
1 mm
ven
ac
ac
med
2218 26
3430 38
4642 50
10 14
ac
ac
co
co
co
co
co
cc
sl
co
pl
vb
vb
vb
vb
vb
pa
ac
ac
pc
pc
pc
pc
aa
aa
vn
cn
cn
sa
sa
sa
sa
sa
sa
sa
lc
lc
co
lc
lc
lc
lc
fv
fc
fc
fv
vb
la
lc
1 mm
ac
308
FIGURE 5.46. Bony labyrinth of Rhinolophus ferrumequinum. A, stereopair and labeled
line drawing of digital endocast in anterior view; B, stereopair and labeled line drawing
of digital endocast in dorsal view; C, stereopair and labeled line drawing of digital
endocast in lateral view; D, line drawing of cochlea viewed down axis of rotation to
display degree of coiling; E, line drawing of cochlea in profile. Abbreviations: aa,
anterior ampulla; ac, anterior semicircular canal; ant, anterior direction; av, bony channel
for aqueduct of vestibule; cc, canaliculus cochleae for aqueduct of cochlea; co, cochlea;
cr, common crus; dor, dorsal direction; er, elliptical recess of vestibule; fc, fenestra
cochleae; fv, fenestra vestibuli; la, lateral ampulla; lc, lateral semicircular canal; med,
medial direction; pa, posterior ampulla; pc, posterior semicircular canal; pl, primary
bony lamina; pos, posterior direction; sl, secondary bony lamina.
309
A
B
C
D E
co
sl
pl
fv
fc
lc
la
cr
pc
ac
av
aa
co
pl
pc
lc
er
aa
ac
la
1 mm
co
fc
fc
cc
pc
cr
sl
av
ac
aa
la
lc
1 mm
1 mm
sl
co
co
sl
cc
pl
dor
med
med
pos
dor
ant
310
FIGURE 5.47. CT slices through ear region of Rhinolophus ferrumequinum.
Abbreviations: aa, anterior ampulla; ac, anterior semicircular canal; av, bony channel for
aqueduct of vestibule; cn, canal for cranial nerve VIII; co, cochlea; cr, common crus;
dor, dorsal direction; fc, fenestra cochleae; fn, canal for cranial nerve VII; fv, fenestra
vestibuli; la, lateral ampulla; lc, lateral semicircular canal; med, medial direction; pa,
posterior ampulla; pc, posterior semicircular canal; pl, primary bony lamina; pos,
posterior direction; sa, subarcuate fossa; sg, canal for spiral ganglion within primary
lamina; sl, secondary lamina; vb, vestibule; ven, ventral direction; vn, canal for
vestibular branch of cranial nerve VIII.
311
pos
dor
2
6
10
14
18
22
26
30
34
42
38
1 mm
med
dor
co
co
co
co
co
co co
co
sa
sa
sa
sa
pa
cn
cn
vn
vb vb
vb
vb
aa
ac
lc
ac
fv
1 mm
2 6
1410 18
2622 30
3834 42
coco
sl
sl
sl
sl
sl
sl
sg
sg
pl
pl
pl
pl
fn
lc
lc
lc
fc
pc
lc
ac
ac
cr
cr
av
pl
sl
sl
312
FIGURE 5.48. Bony labyrinth of Tadarida brasiliensis. A, stereopair and labeled line
drawing of digital endocast in anterior view; B, stereopair and labeled line drawing of
digital endocast in dorsal view; C, stereopair and labeled line drawing of digital endocast
in lateral view; D, line drawing of cochlea viewed down axis of rotation to display degree
of coiling; E, line drawing of cochlea in profile. Abbreviations: aa, anterior ampulla; ac,
anterior semicircular canal; ant, anterior direction; av, bony channel for aqueduct of
vestibule; cc, canaliculus cochleae for aqueduct of cochlea; co, cochlea; cr, common
crus; dor, dorsal direction; er, elliptical recess of vestibule; fc, fenestra cochleae; fv,
fenestra vestibuli; la, lateral ampulla; lc, lateral semicircular canal; med, medial
direction; pa, posterior ampulla; pc, posterior semicircular canal; pl, primary bony
lamina; pos, posterior direction; ps, outpocketing for the perilymphatic sac; sl, secondary
bony lamina; sr, spherical recess of vestibule.
313
A
B
C
D E
co
sl
aa
la
fv
sr
av
cr
ac
lc
pc
sl
pl
av
cr
pa
fc
aa
la
er
pc
lc
1 mm
co
co
sl
fc
fv
la
av
aa
ac
lc
cr
pc
ps
pa
1 mm
1 mm
co
fc
co
sl
cc
fc
ps
dor
med
med
pos
dor
ant
314
FIGURE 5.49. CT slices through ear region of Tadarida brasiliensis. Abbreviations: aa,
anterior ampulla; ac, anterior semicircular canal; ant, anterior direction; av, bony channel
for aqueduct of vestibule; cf, foramina within cribriform plate; cn, canal for cranial nerve
VIII; co, cochlea; cr, common crus; er, elliptical recess of vestibule; fc, fenestra
cochleae; fn, canal for cranial nerve VII; fv, fenestra vestibuli; la, lateral ampulla; lat,
lateral direction; lc, lateral semicircular canal; med, medial direction; pa, posterior
ampulla; pc, posterior semicircular canal; pl, primary bony lamina; pos, posterior
direction; sa, subarcuate fossa; sg, canal for spiral ganglion within primary lamina; sl,
secondary lamina; sr, spherical recess of vestibule; ven, ventral direction; vn, canal for
vestibular branch of cranial nerve VIII.
315
ven
ant
17
52
87
122
157
192
227
262
297
367
332
1 mm
lat
pc
cr
cr
av
av
av
av
ac
lc
lc
ac
sa
sa
er
er
sr
pc
pc
ant
0.5 mm
17 52
12287 157
227192 262
332297 367
cr cr
cf
co
fn
vn
pc
fc
sg
cn
cn
cn
co
co
co
co
cn
fc
fv
la
pa
plpl
sl
sl
sl
ac
aa
sg
sg
sg
cf
pl
pl
316
each of the chiropteran taxa examined. The secondary lamina of Rhinolophus (Figure
5.47), which is expressed as a groove on the endocast, persists for the greatest relative
distance along the radial wall of the cochlear canal (935°), and the least in Nycteris (316°;
Figure 5.45). The secondary lamina of Tadarida (Figure 5.49) extends for 659°, which is
intermediate between the other two microchiropterans. The secondary laminae of
Rhinolophus and Tadarida are the only laminae examined so far to extend beyond the
basal turn, with the exception of the lamina in Tursiops (396°). The lamina of
Rhinolophus completes more turns than that of any other mammal examined in this study
(see also Stan!k, 1933).
The aspect ratio of the cochlear spiral in profile is the smallest for Tadarida (0.52;
0.61 for Nycteris; 0.63 for Rhinolophus). The apical whorls of the cochlea of Tadarida sit
upon the basal turn, whereas the apical whorls fit within the more basal turns in both
Rhinolophus and Nycteris. A groove, which is expressed as a ridge on the endocast,
situated on the axial wall of the cochlear canal opposite the fenestra cochleae, leads to the
short canaliculus cochleae for the aqueduct of the cochlea in Tadarida (0.12 mm in
length). A canaliculus is observed in both Rhinolophus (0.59 mm in length) and Nycteris
(0.66 mm in length), although the groove opposite the fenestra cochleae is not observed
in these latter species. The canaliculus is very straight in Rhinolophus, and it is oriented
nearly perpendicular to the plane of the basal turn of the cochlea.
The plane of the basal turn of the cochlea of Rhinolophus forms an angle of 5.5°
with the plane of the lateral semicircular canal, which is the smallest angle measured
between the structures of any mammal described here (see Table 5.3). The angle in
Nycteris (47.2°) is similar to that measured in Felis (45.8°) and the elephantoid (48.5°),
and the angle in Tadarida (29.2°) is not much different from that observed in Eumetopias
(31.6°) and Orycteropus (31.9°).
317
The fenestra vestibuli is elliptical in Tadarida (stapedial ratio equals 2.0),
although the ratio calculated for Nycteris (1.0) indicates a circular fenestra. The stapedial
ratio, as measured from the fenestra vestibuli, of Rhinolophus is 1.4. The spherical recess
of Tadarida is separated from the elliptical recess by a mild constriction of the vestibule.
The elliptical recess is elongated, with extensions at its anterior and posterior ends. The
recesses are not distinguishable within the vestibule of either Rhinolophus or Nycteris,
but as in Tadarida, the vestibule of the other two species possess an anterior excavation
for the anterior and posterior ampullae, as well as a posterior excavation for the posterior
ampulla and common crus. The posterior excavation is best developed in Rhinolophus.
The common crura of all three bats are tall and especially slender in Rhinolophus
and Nycteris. The bony channel for the aqueduct of the vestibule leaves the inner ear
medial and anterior to the vestibular aperture of the common crus in Tadarida and
Rhinolophus, rather than directly medial to the crus as in Pteropus. The channel is
straight and 1.40 mm in length in Rhinolophus, opening on the surface of the petrosal
near the apex of the common crus, but the channel of Tadarida (1.42 mm in length)
curves gently towards the posterior end of the labyrinth. The presence of a bony channel
for the aqueduct of the vestibule cannot be determined for Nycteris, because the data
were not adequate to resolve the structure (only 70 slices with an interpixel spacing of
0.0654 mm in Nycteris versus 380 slices with and interpixel spacing of 0.0097 mm in
Tadarida; see Table 1). However, the CT dataset of the ear region of Rhinolophus
contains fewer slices (45) than that of Nycteris, yet the channel for the aqueduct of the
vestibule is observed, perhaps because the slices were of a higher resolution interpixel
spacing of 0.043 mm).
The posterior limb of the lateral semicircular canal opens into the posterior
ampulla in Nycteris and Rhinolophus, but the canal opens directly into the vestibule
318
anterior to the posterior ampulla in Tadarida. The lateral canals of Tadarida and
Rhinolophus are positioned high with respect to the other vestibular constituents (sagittal
labyrinthine indices 22.1 and 38.3 respectively), but the canal is comparatively lower in
Nycteris, in which the lateral canal does not cross the space enclosed by the posterior
semicircular canal.
The angle between the planes of the posterior and anterior semicircular canals is
the largest measured in each of the microchiropteran taxa, especially in Nycteris (112°;
angle equals 104° in Rhinolophus; angle equals 98.4° in Tadarida). The anterior and
lateral semicircular canals express the smallest angle in all of the species, especially
Tadarida (74.7°; angle equals 79.9° in Rhinolophus; angle equals 85.9° in Nycteris). The
angle between the posterior and lateral semicircular canals of Nycteris (86.3°) is smaller
than that measured for both Rhinolophus (87.9°) and Tadarida (98.4°).
The semicircular canals of Tadarida are the most planar among the
microchiropterans, and none of them deviate significantly from their planes (ratios of
total linear deviation over cross-sectional diameter of anterior, lateral, and posterior
canals are 0.21, 0.36, and 0.00 respectively). The posterior canal does not deviate
significantly from its plane in Tadarida, although the canal is the least planar in both
Rhinolophus (total deviation equals 13.9°; ratio equals 2.09) and Nycteris (22.7°; ratio
equals 2.77). The greatest total angular deviation in Tadarida was measured for the
lateral semicircular canal (4.7°; deviation equals 4.1° in Rhinolophus with ratio of 0.54;
deviation equals 6.6 in Nycteris with ratio of 0.74). The total angular deviation of the
anterior semicircular canal is 8.3° in Rhinolophus, 4.1° in Nycteris, and 2.0° in Tadarida.
The degree of deviation of the anterior semicircular canal is significant in Rhinolophus
(ratio equals 1.66), but the canal does not deviate significantly in Nycteris (ratio equals
0.58).
319
The radius of the arc of the anterior semicircular canal is greater than the other
canal radii in the three microchiropteran taxa examined here. Among the three species,
the anterior arc radius is largest for Nycteris (0.97 mm; 0.83 mm for Rhinolophus; 0.85
mm for Tadarida). The smallest arc radius was measured for the lateral semicircular
canal arc in Rhinolophus (0.69 mm; 0.87 mm for Nycteris; 0.74 mm for Tadarida). The
radii of the arc of the posterior semicircular canal for Rhinolophus, Nycteris, and
Tadarida are 0.74 mm, 0.79 mm, and 0.73 mm respectively.
The semicircular canals themselves are longer for Nycteris (4.35 mm for anterior;
3.40 mm for lateral; 4.36 mm for posterior) than either Rhinolophus (3.52 mm for
anterior; 3.21 mm for lateral; 3.90 mm for posterior) or Tadarida (3.90 mm for anterior;
3.26 mm for lateral; 3.59 mm for posterior). However, the canals of Tadarida are larger
in terms of cross-sectional diameter. The diameters of the anterior, lateral, and posterior
canals, respectively, are 0.15 mm, 0.17 mm, and 0.14 mm for Tadarida, 0.07 mm, 0.09
mm, and 0.09 mm for Rhinolophus, and 0.12 mm, 0.14 mm, and 0.11 for Nycteris.
The aspect ratio of the arc of the lateral semicircular canal is the lowest among the
three canals for the microchiropterans, particularly for Rhinolophus (0.46; 0.71 for
Nycteris; 0.58 for Tadarida). Only the aspect ratio of the lateral canal in the balaenopterid
(0.39) is smaller than that calculated for Rhinolophus. The highest aspect ratio among the
microchiropteran canal arcs was measured for the posterior canal of Rhinolophus (0.98;
0.95 for Nycteris; 0.91 for Tadarida). The ratio of the slender canal length to arc radius
for the posterior semicircular canal was the greatest ratio in Nycteris (5.51; anterior ratio
equals 4.48; lateral ratio equals 3.91), Rhinolophus (5.25; anterior ratio equals 4.25;
lateral equals 4.64), and Tadarida (4.88; anterior equals 4.62; lateral equals 4.45).
There are no unambiguous synapomorphies within the bony labyrinth uniting
Chiroptera as a whole, nor is there evidence from the inner ear that Rhinolophus shares a
320
more recent common ancestor with Pteropus than the definitive microchiropterans
Nycteris and Tadarida. However, the lateral semicircular canals of both Nycteris and
Rhinolophus empty into the posterior ampulla, whereas the lateral canals of Pteropus and
Tadarida open into the vestibule directly.
A secondary common crus is not observed in any of the bats examined. In this
regard, the bony labyrinth of Chiroptera is derived from that of the ancestral eutherian,
but retains this morphology from the ancestral placental. Most of the bats are derived
from the ancestral eutherian condition in the position of the lateral semicircular canal in
relation to the ampullar opening of the posterior canal, although Nycteris retains the
ancestral condition. Because of this, the state in the ancestor of Microchiroptera is
equivocal as reconstructed. Tadarida retains the ancestral therian state of a flattened
cochlea, whereas the cochleae of all other bats have a high aspect ratio. Nonetheless, the
ancestral microchiropteran condition is a cochlea with a high aspect ratio, which is
retained from the ancestor to all of Chiroptera. The largest semicircular canal arc is
observed in the anterior canal in all of the bats, although this feature is plesiomorphic and
shared with most therian taxa. The cochlea of the microchiropteran ancestor coils 820.1°
and contributes 68.0% of the total labyrinthine volume, both of which are greater values
than those reconstructed for the ancestor of Chiroptera (763.7°; 61.0%).
The most recent common ancestor of Rhinolophus and Tadarida possessed a
cochlea with a high aspect ratio, a lateral semicircular canal positioned high compared to
the posterior canal, and an anterior semicircular canal arc that was the largest among the
three arcs. All of these states also are present in the ancestor of Chiroptera. Because the
lateral semicircular canal opened into the vestibule in Tadarida and into the posterior
ampulla in Rhinolophus, the state in the most recent common ancestor of these taxa was
reconstructed as equivocal. The cochlea of this ancestor coiled 895.6°, and contributed
321
76.0% of the entire labyrinthine volume. Although the cochlea contributed a great
amount of the labyrinthine volume, it was not as great as that reconstructed for Cetacea
(84.0%).
Eulipotyphla
The sister taxon to the Ungulata, Ferae, and Chiroptera polytomy is Eulipotyphla.
The constituents of Eulipotyphla are Erinaceidae (hedgehogs), Soricidae (shrews),
Talpidae (moles), Solenodon, and the extinct genus Nesophontes (following Asher,
2005). These taxa traditionally were grouped with the afrosoricid Tenrecidae and
Chrysochloridae within the paraphyletic Lipotyphla (sensu McKenna and Bell, 1997),
which in turn was a subset of Insectivora (sensu Simpson, 1945), that also included
Macroscelidea. Although most recent phylogenetic analyses fail to support either
insectivoran or lipotyphlan monophyly, a monophyletic Eulipotyphla often is recovered
(e.g., Stanhope et al., 1998; Murphy et al., 2001a, b; Asher et al., 2003; Grenyer and
Purvis, 2003; Nikaido et al., 2003; Bininda-Emonds et al., 2007). However, eulipotyphlan
monophyly is not always found, even among molecular studies that group the
afrosoricids with other afrotherians (e.g., Emerson et al., 1999; Mouchaty et al., 2000).
The eulipotyphlans form the third most speciose clade of placental mammals
(Wilson and Reeder, 1993; Reeder et al., 2007), most of which belong to the subclade
Soricomorpha (shrews, moles, solenodons, and nesophontids). The sister taxon to the
soricomorphs are the hedgehogs, which belong to the group Erinaceomorpha. Both major
subclades of eulipotyphlan are represented - the hedgehog Atelerix albiventris (Figures
5.50-5.51) and the shrew Sorex monticolus (Figures 5.52-5.53).
Atelerix is significantly larger than Sorex, with a body mass of 866 grams versus
6.1 grams for the shrew (Silva and Downing, 1995). Likewise, the bony labyrinth of
322
FIGURE 5.50. Bony labyrinth of Atelerix albiventris. A, stereopair and labeled line
drawing of digital endocast in anterior view; B, stereopair and labeled line drawing of
digital endocast in dorsal view; C, stereopair and labeled line drawing of digital endocast
in lateral view; D, line drawing of cochlea viewed down axis of rotation to display degree
of coiling; E, line drawing of cochlea in profile. Abbreviations: aa, anterior ampulla; ac,
anterior semicircular canal; ant, anterior direction; cc, canaliculus cochleae for aqueduct
of cochlea; co, cochlea; cr, common crus; dor, dorsal direction; er, elliptical recess of
vestibule; fc, fenestra cochleae; fv, fenestra vestibuli; la, lateral ampulla; lc, lateral
semicircular canal; med, medial direction; pa, posterior ampulla; pc, posterior
semicircular canal; pl, primary bony lamina; pos, posterior direction; sl, secondary bony
lamina; sr, spherical recess of vestibule.
323
A
B
C
D E
co
sl
pl
sr
la
aa
fv
fc
1 mm
1 mm
1 mm
cc
cr
ac
pc
pa
lc
co
pl
cc
cr
fc
pa
pc
la
aa
ac
lc
er
co
fc
cc
pc
ac
cr
aa
la
lc
sr
er
fc
co
co
sl
fc
cc
dor
med
med
pos
dor
ant
324
FIGURE 5.51. CT slices through ear region of Atelerix albiventris. Abbreviations: aa,
anterior ampulla; ac, anterior semicircular canal; cn, canal for cranial nerve VIII; co,
cochlea; cr, common crus; dor, dorsal direction; fc, fenestra cochleae; fn, canal for
cranial nerve VII; fv, fenestra vestibuli; la, lateral ampulla; lc, lateral semicircular canal;
med, medial direction; pa, posterior ampulla; pc, posterior semicircular canal; pl,
primary bony lamina; pos, posterior direction; vb, vestibule; vn, canal for vestibular
branch of cranial nerve VIII.
325
pos
dor
3
8
13
18
23
28
33
38
43
53
48
1 mm
med
dor
ac
1 mm
3 8
1813 23
3328 38
4843 53
pc
ac
aa
la
ac
ac
fv
fc
cr
ac
fn vn
fn
cn
co
co
pl
vb
vb
vb vb
vb
vb
pc
pc
pa
lc
lc lc
lc
la
ac
la
co
co
co
fc
fc
326
FIGURE 5.52. Bony labyrinth of Sorex monticolus. A, stereopair and labeled line
drawing of digital endocast in anterior view; B, stereopair and labeled line drawing of
digital endocast in dorsal view; C, stereopair and labeled line drawing of digital endocast
in lateral view; D, line drawing of cochlea viewed down axis of rotation to display degree
of coiling; E, line drawing of cochlea in profile. Abbreviations: aa, anterior ampulla; ac,
anterior semicircular canal; ant, anterior direction; av, bony channel for aqueduct of
vestibule; cc, canaliculus cochleae for aqueduct of cochlea; co, cochlea; cr, common
crus; dor, dorsal direction; er, elliptical recess of vestibule; fc, fenestra cochleae; fv,
fenestra vestibuli; la, lateral ampulla; lc, lateral semicircular canal; med, medial
direction; pa, posterior ampulla; pc, posterior semicircular canal; pl, primary bony
lamina; pos, posterior direction; sl, secondary bony lamina; sr, spherical recess of
vestibule.
327
A
C
B
D E
co
aa
la
cr
ac
av
lc
pa
pc
fv
fc
sr
er
co
sl
cc
pa
av
pc
er
sr
aa
la
ac
lc
1 mm
cr
pl
pa
la
aa
ac
lc
fc
pc
cr
cc
av
co
1 mm
1 mm
fc
co
co
sl
cc
fc
dor
med
med
pos
dor
ant
328
FIGURE 5.53. CT slices through ear region of Sorex monticolus. Abbreviations: aa,
anterior ampulla; ac, anterior semicircular canal; av, bony channel for aqueduct of
vestibule; cn, canal for cranial nerve VIII; co, cochlea; cr, common crus; dor, dorsal
direction; er, elliptical recess; fc, fenestra cochleae; la, lateral ampulla; lat, lateral
direction; lc, lateral semicircular canal; pc, posterior semicircular canal; pl, primary bony
lamina; pos, posterior direction; sr, spherical recess of vestibule; st, stapes within
fenestra vestibuli; vb, vestibule.
329
pos
dor
21
31
41
51
61
71
81
91
101
121
111
1 mm
dor
lat
co
co
co
co
co
cn
21 31
5141 61
8171 91
111101 121
co
pl
st
fc
pl
pl
sr
sr
er
er
vb
vb
aa
aa
ac
ac
ac
ac
ac
ac
la
la
lc
lc
pc
lc
av
av
av
av
cr
cr
cr
cr
pc
pc
pc
1 mm
330
Atelerix is larger than that of Sorex (Figure 5.17), both in terms of length (5.46 mm
versus 2.81 mm respectively) and volume (4.58 mm versus 0.81 mm
3
respectively). The
volume of the cochlear canal of Atelerix (2.28 mm
3
) is greater than that measured for
Sorex (0.37 mm
3
), although the respective contributions of the cochlea to total inner ear
volume (49.7% versus 45.5%) are similar.
Not only is the length of the cochlear canal of Atelerix (4.99 mm) longer than that
of Sorex (2.52 mm), the cochlea of the hedgehog completes a greater degree of coiling
(624°) than the shrew (493°). The secondary bony lamina of Sorex persists for the first
half of the basal turn of the cochlea (179°), whereas the lamina persists for two thirds of
the basal turn in Atelerix (240°). The aspect ratio of the spiral of the cochlea in profile of
Atelerix (0.69) is greater than the ratio calculated for Sorex (0.47).
The plane of the basal turn of the cochlea forms an angle of 53.8° with the plane
for the lateral semicircular canal in Atelerix, but an angle of only 9.4° was measured for
Sorex. The canaliculus cochleae for the aqueduct of the cochlea is 0.77 mm in length in
Atelerix versus 0.23 mm in Sorex. The scala tympani of the cochlea is expanded internal
to the fenestra cochleae, leading to the canaliculus in both eulipotyphlan taxa. The
expansion in Atelerix is elongated, and curves ventrally, forming a hook on the endocast
that terminates with the fenestra cochleae. Elongation of the expansion is not observed in
Sorex.
The fenestra vestibuli of both eulipotyphlan taxa are elliptical, with stapedial
ratios (calculated from dimensions of the fenestra) of 1.8 for Atelerix and 1.7 for Sorex.
The vestibules of both taxa are constricted, separating the spherical and elliptical
recesses, the latter with anterior and posterior excavations in both Atelerix and Sorex. The
bony channel for the aqueduct of the vestibule opens medial to the vestibular aperture of
the common crus in Atelerix, but the channel exits the inner ear cavities anterior to the
331
crus in Sorex. The relative lengths of the channels are different between the
eulipotyphlans. The channel is very short (0.38 mm in length) and straight in Atelerix,
although the channel extends for a greater distance in Sorex, where the delicate canal
parallels the common crus for much of its length before turning posteriorly to open on the
external surface of the petrosal. The absolute length of the channel for the aqueduct of the
vestibule is 1.58 mm in Sorex.
The posterior limb of the lateral semicircular canal opens directly into the
vestibule in both taxa, although the vestibular aperture of the canal of Atelerix is further
separated from the base of the posterior ampulla than the canal in Sorex. The lateral canal
is positioned high relative to the posterior semicircular canal in both the hedgehog and
shrew, particularly in Atelerix (sagittal labyrinthine index equals 26.4; index equals 11.9
in Sorex).
The angle between the planes of the posterior and lateral semicircular canals
(92.1°) is the greatest between canals in Atelerix, although the angle between the anterior
and posterior canals in the hedgehog labyrinth is not much different (91.7°). The angle
between the anterior and lateral canals is significantly more acute (82.2°). The greatest
angle in Sorex was measured between the anterior and posterior canals (89.6°), but as was
measured for Atelerix, the angle between the posterior and lateral canals is similar
(89.3°). The angle between the anterior and lateral semicircular canals in Sorex is 75.3°.
Among the semicircular canals of both eulipotyphlan taxa, the posterior canal
Sorex deviates the most from its plane (total angular deviation equals 21.2°; deviation in
Atelerix equals 14.6°). The least planar canal of Atelerix is the lateral semicircular canal
(18.9°), which does not deviate significantly from its plane in the shrew. The total
angular deviation of the anterior canal is 10.6° in Atelerix and 7.9° in Sorex. The degrees
of deviation for the anterior, lateral, and posterior semicircular canals are significant for
332
Atelerix (ratios of total linear deviation over cross-sectional diameter equal 1.40, 2.00,
and 2.22 respectively), but only the posterior canal of Sorex exhibits significant deviation
(ratio is 1.87; anterior is 0.78; lateral is 0.00).
The arc radius of curvature of the anterior semicircular canal is largest in both
Atelerix (1.24 mm) and Sorex (0.65 mm), although the radius of the posterior arc is
similar to that of the anterior in both taxa (1.22 mm for Atelerix; 0.63 mm for Sorex). The
radius of the arc of the lateral semicircular canal is 0.88 mm for Atelerix and 0.48 mm in
Sorex. The anterior semicircular canal of Atelerix (5.88 mm) is greater than either the
lateral (3.67 mm) or posterior canals (5.80 mm), although the longest canal in Sorex is the
posterior semicircular canal (3.42 mm; anterior equals 3.20 mm; lateral equals 1.63 mm).
The anterior and posterior semicircular canal volumes are the same in Atelerix (0.04 mm
3
each), which is a greater value than the lateral canal (0.03 mm
3
). The volumes of all of
the canals in Sorex are identical (0.02 mm
3
). The cross-sectional diameters of the
anterior and posterior semicircular canals are the same in Sorex (0.12 mm), which is a
smaller value than that measured for the lateral canal (0.14 mm). The diameter of the
anterior canal is largest in Atelerix (0.16 mm; lateral and posterior both equal 0.14 mm).
The lateral and posterior semicircular canal arcs of Atelerix approach perfect
circles (height and width nearly identical) with respective aspect ratios of 0.99 and 0.97
(versus 0.88 and 0.72 in Sorex). The ratio of slender canal length to arc radius of the
posterior semicircular canal is the greatest in Sorex (5.44), and the ratio of the lateral
canal is the smallest in the shrew (3.38). The ratio of the lateral canal in Atelerix is the
smallest (4.15), and the ratio is identical for the anterior and posterior canals (4.74).
No features of the bony labyrinth support monophyly of Eulipotyphla, nor are
there any unambiguous characters that unite the eulipotyphlans with the afrosoricids
(Chrysochloris and Hemicentetes). Both Sorex and Atelerix are derived from the ancestral
333
eutherian condition in that the lateral semicircular canal enters the vestibule directly
rather than forming a secondary common crus with the posterior canal, as well as a high
position of the plane of the lateral canal in relation to the ampullar opening of the
posterior semicircular canal. Vestibular entry of the lateral canal is inherited from the
ancestor of Placentalia. The cochlea of Atelerix is derived from the ancestral eutherian in
that the aspect ratio of the spiral is high, whereas the cochlea of Sorex retains the
primitive flattened condition reconstructed for the ancestor of Theria.
The ancestral states of Eulipotyphla include a lateral semicircular canal that opens
into the vestibule directly (retained from placental ancestor) and is positioned high
compared to the ampullar entrance of the posterior canal (retained from boreoeutherian
ancestor), and an anterior semicircular canal arc with the largest radius among the three
arcs (retained from therian ancestor). The state of the aspect ratio is reconstructed as
equivocal for the ancestor of Eulipotyphla. The cochlea of the most recent common
ancestor of Atelerix and Sorex contributes 50.0% of labyrinthine volume and the cochlear
canal coils 622.6°. The low degree of coiling of the ancestral eulipotyphlan either is a
retention from its therian ancestor, or else a reversal to a more ancestral morphology.
Euarchontoglires
Euarchontoglires contains the remaining placental mammal clades. Among these
are the highly speciose Rodentia, Primates, Lagomorpha, Dermoptera, and Scandentia.
Gross dimensions of the bony labyrinths of Euarchontoglires are provided in Table 5.2.
Dimensions of the cochlea are provided in Table 5.3, and dimensions and orientations of
the semicircular canals are reported in Tables 5.4-5.6.
334
The states reconstructed for the bony labyrinth of the most recent common
ancestor of Euarchontoglires are the same as those for Boreoeutheria. That is, the lateral
semicircular canal opens into the vestibule directly in the absence of a secondary
common crus, the lateral semicircular canal is positioned high compared to the posterior
semicircular canal, and the anterior canal arc is the largest in terms of radius of curvature
among the three arcs. The cochlea of the ancestral euarchontoglire coils 956.9°, which is
over a quarter of a turn greater than that reconstructed for the ancestral boreoeutherian
(815.4°), and the cochlea of Euarchontoglires contributes 53.0% of the total inner ear
volume (retained from the ancestor of Boreoeutheria, which had a cochlea contributing
55.0%). An unequivocal state of the aspect ratio of the cochlea could not be
reconstructed.
Recognition of a close relationship between rodents and lagomorphs can be traced
back to the seminal classification of Linnaeus (1758), in which he also included the
rhinoceros (although before 1758, he restricted Glires to rodents and lagomorphs;
Linnaeus, 1748). Monophyly of Glires persisted in several later classifications, most
notably those of Flower (1883), Gregory (1910), Simpson (1945), and McKenna and Bell
(1997). That is not to say that a monophyletic Glires has been free from controversy (see
Wood, 1957; Meng and Wyss, 2005). Both morphological (Gidley, 1912; Wood, 1957,
1962; Van Valen, 1971; McKenna, 1975) and molecular investigations (Moody et al.,
1949; Easteal, 1990; Stanhope et al., 1992; Porter et al., 1996; Arnason et al., 2002;
Misawa and Janke, 2003) have either allied Rodentia or Lagomorpha with various other
placental mammal taxa, or else render the groups within Glires paraphyletic with varying
levels of robusticity. Despite the ambiguity of rodent and lagomorph affinites in earlier
studies, a unified Glires is supported by many recent phylogenetic analyses (Liu et al.,
2001; Meng and Wyss, 2001; Murphy et al., 2001a, b; Lin et al., 2002; de Jong et al.,
335
2003; Meng et al., 2003; Springer et al., 2003; Douzery and Huchon, 2004; Meng, 2004;
Asher et al., 2005; Asher, 2007; Bininda-Emonds, 2007; Wible et al., 2007).
The most recent common ancestor of Rodentia and Lagomorpha (Glires) retained
a lateral semicircular canal that opened into the vestibule directly in absence of a
secondary common crus from Placentalia, a position of the lateral canal high compared to
the posterior canal from Boreoeutheria, and the highest arc radius of curvature measured
for the anterior semicircular canal arc from Theria. Although the euarchontoglire
ancestral state of the aspect ratio of the cochlea was equivocal, the ancestral glire
possessed a cochlea with a high aspect ratio, which was shared with Scandentia among
the members of Euarchontoglires. The cochlea of the ancestral Glires coiled 923.6°, and
the cochlea contributed 55% of the total labyrinthine volume, which was inherited from
the ancestral boreoeutherian.
Primates, dermopterans, and scandentians together form the clade Euarchonta
(following Beard, 1993; Waddell et al., 1999). However, the results of Bininda-Emonds
do not recover a monophyletic Euarchonta. Rather, Scandentia is placed in a polytomy
with Glires and a Dermoptera plus Primates clade, which is referred to as Primatomorpha
(Figure 5.2). The ancestral labyrinth of Primatomorpha retained an anterior semicircular
canal with the greatest arc radius of curvature among the three canal arcs from the
ancestral therian, a high position of the lateral semicircular canal from the ancestral
boreoeutherian, and a direct vestibular entrance of the lateral semicircular canal from the
ancestral placental. The aspect ratio of the cochlea was low for Primatomorpha, which is
a unique state within Euarchontoglires.
336
Rodentia
Rodents make up the most speciose clade of mammals, contributing over 40% of
all named extant mammal species (Reeder et al., 2007). The supertree of Bininda-
Emonds et al. (2007) depicts Rodentia as a natural clade, although rodent monophyly has
been questioned. The results of some phylogenetic analyses based on molecular sequence
data support the hypothesis that guinea pigs do not share a common ancestry with other
rodent taxa, but rather with Primates and ungulates (Graur et al., 1991, 1992; Li et al.,
1992; Ma, 1993; D’Erchia et al., 1996). Despite these analyses, the majority of data, both
morphological (Luckett and Hartenberger, 1993; Asher et al., 2003; Meng et al., 2003)
and molecular (Cao et al., 1994; Frye and Hedges, 1995; Cao et al., 1997; Huchon et al.,
1999; Adkins et al., 2001; Murphy et al., 2001a, b; Huchon et al., 2002), support a
monophyletic Rodentia.
The rodents examined in this study are the mouse Mus musculus (Figures 5.54-
5.55) and the guinea pig Cavia porcellus (Figures 5.56-5.57). The bony labyrinth of
Cavia is larger than that of Mus, both in terms of labyrinthine length (22.22 mm versus
1.47 mm) and volume (7.13 mm
3
versus 2.71 mm
3
). The dimensions of the inner ear
cavities are mirrored by the average body masses of the two species (728 g for Cavia and
15.5 g for Mus; Silva and Downing, 1995). The volume of the cochlea of Cavia (12.26
mm
3
) is significantly greater than that measured for Mus (0.86 mm
3
), although the
cochlea forms a greater proportion of the bony labyrinth in the mouse (58.6%) than in the
guinea pig (55.2%).
The most noticeable aspect of the bony labyrinth of Cavia is the sharp cone
formed by the cochlea (Figure 5.56). Not only is the aspect ratio of the cochlea (1.29)
twice what is observed in Mus (0.62), the ratio of the guinea pig is greater than that
calculated for any other mammal discussed here (closest is Macroscelides with a ratio of
337
FIGURE 5.54. Bony labyrinth of Mus musculus. A, stereopair and labeled line drawing
of digital endocast in anterior view; B, stereopair and labeled line drawing of digital
endocast in dorsal view; C, stereopair and labeled line drawing of digital endocast in
lateral view; D, line drawing of cochlea viewed down axis of rotation to display degree of
coiling; E, line drawing of cochlea in profile. Abbreviations: aa, anterior ampulla; ac,
anterior semicircular canal; ant, anterior direction; av, bony channel for aqueduct of
vestibule; cc, canaliculus cochleae for aqueduct of cochlea; co, cochlea; cr, common
crus; dor, dorsal direction; er, elliptical recess of vestibule; fc, fenestra cochleae; fv,
fenestra vestibuli; la, lateral ampulla; lc, lateral semicircular canal; med, medial
direction; pa, posterior ampulla; pc, posterior semicircular canal; pl, primary bony
lamina; pos, posterior direction; ps, outpocketing for the perilymphatic sac; sl, secondary
bony lamina; sr, spherical recess of vestibule.
338
A
B
C
D E
co
sl
1 mm
1 mm
1 mm
sr
er
aa
la
fc
pa
fv
av
ac
cr
lc
pc
co
sl
pl
av
cr
pc
pa
aa
la
lc
ac
er
sr
co
ps
fc
fv
pa
pc
cr
av
aa
lc
la
ac
cc
co
fc
co
sl
fc
ps
dor
med
med
pos
dor
ant
339
FIGURE 5.55. CT slices through ear region of Mus musculus. Abbreviations: aa, anterior
ampulla; ac, anterior semicircular canal; av, bony channel for aqueduct of vestibule; cn,
canal for cranial nerve VIII; co, cochlea; cr, common crus; fc, fenestra cochleae; fn,
canal for cranial nerve VII; fv, fenestra vestibuli; la, lateral ampulla; lat, lateral direction;
lc, lateral semicircular canal; med, medial direction; pa, posterior ampulla; pc, posterior
semicircular canal; pos, posterior direction; sa, subarcuate fossa; sg, canal for spiral
ganglion within primary bony lamina; sl, secondary bony lamina; vb, vestibule; vn, canal
for vestibular branch of cranial nerve VIII.
340
pos
lat
4
11
18
25
32
39
46
53
60
73
67
1 mm
med
co
co
co
co
co
cn
cn
cn
aa
aa
la
la
sa
sa
sa
sa
sa
lc
lc
ac
ac
ac
fn
vb
vb
vb
pa
pc
cr
av
av
cr
vn
fv
fc
sg
sg
sl
sg
lat
4 11
2518 32
4639 53
6760 73
lc
ac
ac
av
av
ac
pc
pc
pc
1 mm
341
FIGURE 5.56. Bony labyrinth of Cavia porcellus. A, stereopair and labeled line drawing
of digital endocast in anterior view; B, stereopair and labeled line drawing of digital
endocast in dorsal view; C, stereopair and labeled line drawing of digital endocast in
lateral view; D, line drawing of cochlea viewed down axis of rotation to display degree of
coiling; E, line drawing of cochlea in profile. Abbreviations: aa, anterior ampulla; ac,
anterior semicircular canal; ant, anterior direction; av, bony channel for aqueduct of
vestibule; cc, canaliculus cochleae for aqueduct of cochlea; co, cochlea; cr, common
crus; dor, dorsal direction; er, elliptical recess of vestibule; fc, fenestra cochleae; fv,
fenestra vestibuli; la, lateral ampulla; lc, lateral semicircular canal; med, medial
direction; pa, posterior ampulla; pc, posterior semicircular canal; pl, primary bony
lamina; pos, posterior direction; ps, outpocketing for the perilymphatic sac; sl, secondary
bony lamina; sr, spherical recess of vestibule.
342
A
B
C
D E
sl
co
1 mm
fc
fv
sr
er
av
cr
ac
aa
la
pa
lc
pc
ps
sr
co
pl
cc
av
pc
er
sr
aa
la
ac
lc
cr
co
sl
fv
fc
ps
cc
pa
la
aa
lc
cr
ac
av
pc
er
1 mm
1 mm
co
fc
cc
co
sl
fc
cc
dor
med
med
pos
dor
ant
343
FIGURE 5.57. CT slices through ear region of Cavia porcellus. Abbreviations: aa,
anterior ampulla; ac, anterior semicircular canal; av, bony channel for aqueduct of
vestibule; cc, canaliculus cochleae for aqueduct of cochlea; cn, canal for cranial nerve
VIII; co, cochlea; cr, common crus; dor, dorsal direction; fc, fenestra cochleae; fn, canal
for cranial nerve VII; la, lateral ampulla; lc, lateral semicircular canal; med, medial
direction; pa, posterior ampulla; pc, posterior semicircular canal; pl, primary bony
lamina; pos, posterior direction; sa, subarcuate fossa; sg, canal for spiral ganglion within
primary bony lamina; sr, spherical recess of vestibule; st, stapes within fenestra vestibuli;
vn, canal for vestibular branch of cranial nerve VIII.
344
pos
dor
244
270
296
322
348
374
400
426
452
504
478
5 mm
med
dor
1 mm
244 270
322296 348
400374 426
478452 504
co co
co
cn
cn
cn
cn
cn
cn
fn
fn
fc
fn
vn
vb
sr
sr
la
aa
lc
lc
lc
ac
ac
ac
sa
sa
av
av av
av
cc
cr
cr
pa
pc
pc
pc
pc
er
pl
pl
sg
sg
sg
st
st
345
0.80). The cochlea of Cavia coils to a much greater degree than any other mammal
studied here (1457.1°), completing over four whorls. Even the highly coiled cochlea of
Procavia capensis only coils three and three quarters turns (see Figure 5.16). The cochlea
of Mus coils to 627.7°, and the length of the cochlear canal is 3.87 mm (13.43 mm for
Cavia).
The scala tympani is expanded interior to the fenestra cochleae in both taxa. The
expansion leads to the canaliculus cochleae for the aqueduct of the cochlea. The
canaliculus is a short and straight tube in Mus (0.17 mm in length), but the slender canal
of Cavia (2.52 mm in length) is gently curved and ends in a triangular shaped fissure. The
plane of the basal turn of the cochlea deviates from the plane of the lateral semicircular
canal to a greater degree in Cavia (35.1°) than in Mus (10.8°).
The bony vestibule of Mus is not divided into the spherical and elliptical recesses,
although an excavation at the anterior end of the vestibule, which is expressed as a
bulbous pedestal in the endocast, for the posterior ampulla, common crus, and posterior
limb of the lateral semicircular canal is present in the mouse. The vestibule is subdivided
into the spherical and elliptical recesses by a constriction interior to the fenestra vestibuli
in Cavia. The fenestra vestibuli is elliptical in the guinea pig, with a stapedial ratio of 2.9
(ratio in Mus equals 1.9), which is expressed as an oblong depression in the spherical
recess on the endocast. Both the spherical and elliptical recesses are elongate cavities.
Unlike Mus, the common crus and posterior ampulla of Cavia do not empty into a
posterior extension of the vestibule. The bony channel for the aqueduct of the vestibule
exits the inner ear anterior to the medial edge of the vestibular aperture of the common
crus in both taxa (length equals 3.82 mm in Cavia; 1.28 mm in Mus). The posterior limb
of the lateral semicircular canal opens directly into the vestibule anterior to the posterior
ampulla in both rodents. The plane of the lateral canal is positioned dorsal relative to the
346
posterior semicircular canal, in both Cavia and Mus (sagittal labyrinthine indices equal
25.3 and 25.8 respectively).
The planes of the anterior and posterior semicircular canals in Cavia form an
angle of 105° (angle equals 94.4° in Mus), which is the greatest angle measured between
any two canals measured for both rodent species. The planes of the anterior and lateral
semicircular canals in Cavia form an angle of 77.2° (88.8° for Mus), which is the smallest
angle measured between any two canals measured for both rodents. The angle measured
between the posterior and lateral semicircular canals is 85.5° for Cavia and 95.6° for
Mus. The semicircular canals of Cavia are less planar than the canals of Mus, especially
the posterior canal, which deviates from its plane by a total of 30.7° in Cavia (ratio of
total linear deviation over cross-sectional diameter is 3.13; angular deviation is 3.4° and
ratio is 1.16 in Mus). The lateral semicircular canal of Mus is the most planar canal in
either taxon, with a total angular deviation of 1.9° (ratio is 0.13; angular deviation is
15.8° and ratio is 1.49 for Cavia). The total angular deviation of the anterior semicircular
canal from its plane is 19.1° for Cavia (ratio is 0.31) and 13.3° for Mus (ratio is 3.10).
The anterior semicircular canal is the largest in terms of arc radius of curvature
and slender canal length for both rodents (1.88 mm and 9.01 mm respectively for Cavia;
0.78 mm and 3.86 mm for Mus), and the lateral canal is the smallest for the same
dimensions (1.57 mm and 6.49 mm for Cavia; 0.60 mm and 2.48 mm for Mus). The arc
radius and length of the posterior semicircular canal are 1.63 mm and 8.18 mm
respectively for Cavia, and 0.67 mm and 3.60 mm for Mus. The cross-sectional diameter
of the anterior semicircular canal of Mus (0.16 mm) is greater than the lateral (0.15 mm)
and posterior canals (0.13 mm), but the diameter of the lateral canal of Cavia (0.29 mm)
is greater than the anterior (0.21 mm) and posterior semicircular canals (0.28 mm).
347
Both the largest and smallest semicircular canal arc aspect ratios were measured
for the arcs of Cavia. The largest semicircular canal arc aspect ratio in Cavia is observed
in the posterior canal (0.99; ratio equals 0.75 in Mus), and the smallest ratio is observed
in the lateral semicircular canal for the guinea pig (0.49; ratio equals 0.92 in Mus). The
aspect ratio of the anterior semicircular canal is 0.75 in Cavia and 0.67 in Mus. The ratio
of the slender canal length to arc radius of the posterior semicircular canal is the greatest
for both species (5.02 for Cavia; 5.39 for Mus), and the ratio is the smallest in the lateral
canal (4.13 for Cavia; 4.12 for Mus). The canal length to arc radius ratio of the anterior
semicircular canal is 4.79 for Cavia and 4.98 for Mus.
The labyrinths of Cavia and Mus retain the ancestral condition reconstructed for
Theria in that the largest semicircular arc radius is observed in the anterior canal. Further,
the labyrinth of the ancestor of Rodentia retained the ancestral placental entry of the
lateral canal (into the vestibule directly), the ancestral boreoeutherian position of the
lateral semicircular canal (high compared to the posterior canal), and the ancestral glire
cochlear aspect ratio (high). The cochlea of the rodent ancestor coiled 1002.8° (close to
1013.1° reconstructed for the most recent common ancestor of Cetacea plus Sus) and
contributed 56.0% of the total labyrinthine volume (close to 55.0% contribution of the
cochlea of Boreoeutheria).
Lagomorpha
Lagomorphs (rabbits and pikas) are classically allied with rodents (proposed as
far back as Linnaeus, 1748) and the majority of recent phylogenetic analyses support this
hypothesis (e.g., Liu et al., 2001; Murphy et al., 2001a, b; Meng et al., 2003; Wible et al.,
2007). However, Lagomorpha has been united with ungulates (Gidley, 1912; Moody et
al., 1949; Wood, 1957) and other placental clades (Misawa and Janke, 2003). Given the
348
predominance of data supporting a clade exclusive to rodents and lagomorphs (Glires),
such a relationship is accepted here.
Two lagomorph species examined here were Lepus californicus (Figures 5.58-
5.59) and Sylvilagus floridanus (Figures 5.60-5.61). The bony labyrinth of Lepus is the
most voluminous (24.27 mm
3
versus 11.32 mm
3
) and longest (7.39 mm versus 5.82 mm).
The black-tailed jackrabbit (Lepus) is a larger species overall than the eastern cottontail
(Sylvilagus), with an average body mass of 2.35 kg for the species, versus 1.16 kg for
Sylvilagus (Silva and Downing, 1995). The volume of the cochlea of Lepus (13.07 mm
3
)
is over twice that measured for Sylvilagus (6.26 mm
3
), although the relative contribution
that the cochlea of each species to total labyrinthine volume is comparable between the
species (53.9% for Lepus; 55.3% for Sylvilagus).
The length of the cochlear canal of Lepus (8.80 mm) is slightly larger than that of
Sylvilagus (8.75 mm), although the cochlea of the cottontail coils to a greater degree than
the jackrabbit (817° versus 693°). Likewise, the secondary bony lamina extends a greater
relative distance along the radial wall of the of the cochlea in Sylvilagus (200°) than in
Lepus (147°), and the aspect ratio of the cochlea of Sylvilagus (0.71) is greater than that
calculated for Lepus (0.64). The apical turns of the cochleae of both lagomorphs sit upon
the basal whorl, as is observed in Mus musculus and Cavia porcellus, and the plane of the
basal turn of the cochlea forms a similar angle with the plane of the lateral semicircular
canal in both Lepus (40.6°) and Sylvilagus (40.3°). The scala tympani of the cochlea is
expanded internal to the fenestra cochleae, which leads to a robust canaliculus cochleae
in each species. The canaliculus of Lepus (4.80 mm in length) forms a straight tube that is
subcircular in cross-section, but the bony canal is flattened and curves ventrally in
Sylvilagus.
349
FIGURE 5.58. Bony labyrinth of Lepus californicus. A, stereopair and labeled line
drawing of digital endocast in anterior view; B, stereopair and labeled line drawing of
digital endocast in dorsal view; C, stereopair and labeled line drawing of digital endocast
in lateral view; D, line drawing of cochlea viewed down axis of rotation to display degree
of coiling; E, line drawing of cochlea in profile. Abbreviations: aa, anterior ampulla; ac,
anterior semicircular canal; ant, anterior direction; av, bony channel for aqueduct of
vestibule; cc, canaliculus cochleae for aqueduct of cochlea; co, cochlea; cr, common
crus; dor, dorsal direction; er, elliptical recess of vestibule; fc, fenestra cochleae; fv,
fenestra vestibuli; la, lateral ampulla; lc, lateral semicircular canal; med, medial
direction; pa, posterior ampulla; pc, posterior semicircular canal; pl, primary bony
lamina; pos, posterior direction; ps, outpocketing for the perilymphatic sac; sl, secondary
bony lamina; sr, spherical recess of vestibule.
350
A
B
C
D E
co
1 mm
sl
sr
fv
ps
fc
cc
pc
la
pa
aa
er
cr
lc
av
ac
co
pl
av
aa
ac
la
er
lc
fc
pa
pc
cr
ps
aa
la
co
ps
fv
fc
cc
pa
pc
av
lc
cr
ac
1 mm
1 mm
co
fc
cc
sl
co
sl
fc
cc
ps
dor
med
med
pos
dor
ant
351
FIGURE 5.59. CT slices through ear region of Lepus californicus. Abbreviations: aa,
anterior ampulla; ac, anterior semicircular canal; av, bony channel for aqueduct of
vestibule; cc, canaliculus cochleae for aqueduct of cochlea; cn, canal for cranial nerve
VIII; co, cochlea; cr, common crus; dor, dorsal direction; fc, fenestra cochleae; fn, canal
for cranial nerve VII; la, lateral ampulla; lc, lateral semicircular canal; med, medial
direction; pa, posterior ampulla; pc, posterior semicircular canal; pl, primary bony
lamina; pos, posterior direction; sa, subarcuate fossa; sl, secondary bony lamina; st,
stapes within fenestra vestibuli; vb, vestibule.
352
pos
dor
44
50
56
62
68
74
80
86
92
104
98
5 mm
med
dor
1 mm
co
co
cn
cn
cn
cc
ac
ac
aa
la
lc
lc
fc
vb
vb
fn
fn
pa
pa
sa
sa
sa
sa
sa
cr
cr
pc
pl
pl
sl
st
44 50
6256 68
8074 86
9892 104
ac
av
av
av
ac
ac
pc
pc
pc
ac
353
FIGURE 5.60. Bony labyrinth of Sylvilagus floridanus. A, stereopair and labeled line
drawing of digital endocast in anterior view; B, stereopair and labeled line drawing of
digital endocast in dorsal view; C, stereopair and labeled line drawing of digital endocast
in lateral view; D, line drawing of cochlea viewed down axis of rotation to display degree
of coiling; E, line drawing of cochlea in profile. Abbreviations: aa, anterior ampulla; ac,
anterior semicircular canal; ant, anterior direction; av, bony channel for aqueduct of
vestibule; cc, canaliculus cochleae for aqueduct of cochlea; co, cochlea; cr, common
crus; dor, dorsal direction; er, elliptical recess of vestibule; fc, fenestra cochleae; fv,
fenestra vestibuli; la, lateral ampulla; lc, lateral semicircular canal; med, medial
direction; pa, posterior ampulla; pc, posterior semicircular canal; pl, primary bony
lamina; pos, posterior direction; ps, outpocketing for the perilymphatic sac; sl, secondary
bony lamina; sr, spherical recess of vestibule.
354
A
B
C
D E
1 mm
1 mm
1 mm
sr
er
aa
la
fc
fv
lc
cr
pc
ac
co
pl
sl
er
aa
ac
la
cr
av
cc
pc
fc
lc
aa
la
fv
fc
sl
co
cc
ps
pa
pc
av
er
cr
ac
lc
pa
fc
cc
co
co
sl
cc
ps
fc
dor
med
med
pos
dor
ant
355
FIGURE 5.61. CT slices through ear region of Sylvilagus floridanus. Abbreviations: aa,
anterior ampulla; ac, anterior semicircular canal; av, bony channel for aqueduct of
vestibule; cc, canaliculus cochleae for aqueduct of cochlea; cn, canal for cranial nerve
VIII; co, cochlea; cr, common crus; dor, dorsal direction; fc, fenestra cochleae; fn, canal
for cranial nerve VII; fv, fenestra vestibuli; la, lateral ampulla; lc, lateral semicircular
canal; med, medial direction; pa, posterior ampulla; pc, posterior semicircular canal; pl,
primary bony lamina; pos, posterior direction; sa, subarcuate fossa; sg, canal for spiral
ganglion within primary bony lamina; sl, secondary bony lamina; sr, spherical recess of
vestibule; st, stapes within fenestra vestibuli; vb, vestibule; vn, canal for vestibular
branch of cranial nerve VIII.
356
pos
dor
117
128
142
156
170
184
198
212
226
254
240
1 mm
med
dor
co
co
co
cn
cn
cn
co
sl
sl
sg
sg
ac
fn
fv
aa
ac
sa
sa
sa
vn
vn
vb
1 mm
117 128
156142 170
198184 212
240226 254
ac
ac
av
av
av
ac
ac
lc
lc
lc
lc
lc
lc
lc
cr
cr
cc
vb
vb
la
la
sr
fv
fc
pl
pl
ps
pc
pa
pa
pc
pc
pcpc
357
The fenestrae vestibuli are less elliptical in the lagomorphs (stapedial ratio equals
1.7 for Lepus; ratio equals 1.5 for Sylvilagus) than for the rodents (Mus equals 1.9; Cavia
equals 2.9), but they are not as round as the fenestra of the microchiropteran bat Nycteris
grandis (1.0). A gentle constriction of the vestibule divides the spherical and elliptical
recesses in both Lepus and Sylvilagus, where the elliptical recesses of the lagomorphs are
elongated with distinct excavations at the anterior and posterior ends (expressed as
pedestals for the ampullae of the semicircular canals on the endocasts).
The bony channel for the aqueduct of the vestibule exits the inner ear cavities
medial to the vestibular aperture of the common crus. The channel is a delicate passage in
Sylvilagus, and it does not end as a flattened fissure as in most other mammals, including
Lepus. The channel is longer in Lepus than it is in Sylvilagus, both in terms of absolute
length (3.71 mm versus 2.08 mm) and length relative to the common crus (channel
terminates ventral to the apex of the common crus in Sylvilagus, but dorsal to the top of
the crus in Lepus).
The lateral semicircular canal opens directly into the vestibule dorsal to the
posterior ampulla in both lagomorphs, giving the plane of the lateral canal a dorsal
position in relation to the posterior semicircular canal (sagittal labyrinthine index equals
32.4 in Lepus; index equals 33.9 in Sylvilagus). The level of the lateral semicircular canal
compared to the posterior canal in the lagomorphs is similar to that observed in
Macroscelides proboscideus (index equals 32.7), where the lateral canal divides the arc
of the posterior canal when the labyrinth is viewed anteriorly.
The planes of the anterior and posterior semicircular canals form the greatest
angle between canals in both Lepus (94.0°) and Sylvilagus (97.5°). The smallest angle
between canals in Lepus was measured between the anterior and lateral semicircular
canals (84.2°; angle equals 92.7° in Sylvilagus), and the smallest angle measured within
358
the labyrinth of Sylvilagus is between the posterior and lateral canals (77.9°; angle equals
88.6° in Lepus). The posterior semicircular canal is the least planar canal in both taxa,
where the canal of Sylvilagus deviates from its plane to a greater degree than that of
Lepus (25.3° versus 10.9°). The posterior canal of Sylvilagus deviates to a significant
degree (ratio of linear deviation over cross-sectional diameter is 3.79), as does the
posterior canal of Lepus (ratio is 1.09). The lateral semicircular canal of Lepus is the most
planar among all of the canals between the two species, with a total angular deviation of
only 2.1° (canal deviates by 5.3° in Sylvilagus). The deviation is not significant for either
species (ratio of 0.24 for Lepus; 0.51 for Sylvilagus). The anterior semicircular canal
deviates from its average plane by a total of 3.9° in Lepus and 5.0° in Sylvilagus.
However, only the anterior canal of Sylvilagus deviates to a significant degree (ratio is
1.28; 0.59 for Lepus).
The arc of the anterior semicircular canal not only has the greatest radius of
curvature in both Lepus (2.34 mm; radius of lateral equals 1.66 mm; radius of posterior
equals 1.69 mm) and Sylvilagus (1.86 mm; radius of lateral equals 1.29; radius of
posterior equals 1.44 mm), but the slender canal length of the anterior semicircular canal
in both Lepus (11.45 mm) and Sylvilagus (8.98 mm) is greater than either the lateral (6.86
mm and 5.65 mm respectively) or posterior canal (8.10 mm and 7.38 mm respectively).
Likewise, the volume of the anterior semicircular canal of Lepus (0.32 mm
3
) is greater
than either the lateral (0.25 mm
3
) or posterior canals (0.24 mm
3
), although the most
voluminous canal within the labyrinth of Sylvilagus is the lateral semicircular canal (0.19
mm
3
versus 0.16 mm
3
for the anterior and 0.14 mm
3
for the posterior canals). The cross-
sectional diameter of the posterior semicircular canal of Lepus (0.30 mm; diameter equals
0.16 mm in Sylvilagus) is greater than either the anterior (0.27 mm; 0.13 for Sylvilagus)
or lateral semicircular canal (0.26 mm; 0.24 mm for Sylvilagus).
359
The aspect ratios of the anterior and posterior canals are greater in Sylvilagus
(0.97 and 0.94 respectively) than in Lepus (0.86 and 0.81 respectively), but the ratio
calculated for the arc of the lateral canal is greater in Lepus than Sylvilagus (0.87 versus
0.84). As in the majority of mammals described so far, the ratio of the slender canal
length to arc radius for the posterior semicircular canal of Sylvilagus (5.13) is greater than
that computed for the anterior (4.84) and lateral semicircular canals (4.38). However, the
greatest ratio among the canals of Lepus was calculated for the anterior semicircular
canal (4.89; 4.13 for the lateral canal; 4.80 for the posterior canal).
There are no unambiguous synapomorphies that support monophyly of
Lagomorpha within Glires or Euarchontoglires. The states reconstructed for the ancestor
of Lagomorpha are the same as those for both Rodentia and Glires, as the lagomorphs
retain the ancestral therian condition of the largest radius of curvature measured for the
anterior semicircular canal arc, the placental condition of the direct vestibular entrance of
the lateral semicircular canal, the ancestral boreoeutherian condition of the high position
of the lateral semicircular canal compared to the ampullar opening of the posterior canal,
and the glire condition of the high aspect ratio of the cochlea. The cochlea of the most
recent common ancestor of lagomorphs coils 751.0° and contributes 53.0% to the total
volume of the inner ear cavities.
Primates
Primates consists of two major groups, Strepsirhini which includes the lemurs and
lorises, and Haplorhini which includes monkeys and apes. The haplorhines are divided
further into three groups, which are Tarsiidae (tarsiers), Platyrhini (New World
monkeys), and Catarhini (Old World monkeys and apes). Monophyly of all of these
clades is supported by numerous phylogenetic analyses (e.g., Shoshani et al., 1996; Kay
360
et al., 1997; Goodman et al., 1998; Poux and Douzery, 2004; Bininda-Emonds, 2007;
Jane"ka et al., 2007).
The two primate species examined here are the rhesus monkey, Macaca mulatta
(Figures 5.62-5.63), and the human being, Homo sapiens (Figures 5.64-5.65). The
average body mass of adult humans (74-86 kg; Ogden et al., 2004) is significantly greater
than that of rhesus monkeys (4.7 kg; Silva and Downing, 1995), and this pattern is
mirrored by the volume (164.73 mm
3
versus 41.64 mm
3
) and length (16.31 mm versus
11.23 mm) of the bony labyrinth. The human cochlea is larger than that of Macaca in
absolute volume (71.49 mm
3
versus 20.96 mm
3
) and canal length (22.49 mm versus
16.94 mm), but the cochlea of Homo contributes a lesser amount to the entire bony
labyrinth than does the cochlear cavity of Macaca (50.3% and 43.4% respectively). Only
the cochlea of the elephantoid proboscidean contributes less (30.6%) to the bony
labyrinth among the mammal species discussed so far.
The cochlea of Macaca completes a greater degree of coiling than the cochlea of
Homo (1088° versus 889°; see Figures 5.62 and 5.64), and the secondary bony lamina
persists to a greater relative distance in the rhesus monkey than the human (81° versus
22°; see Figures 5.63 and 5.65). The aspect ratios of the cochlea in profile for Macaca
and Homo are 0.48 and 0.36 respectively. The scala tympani is expanded internal to the
fenestra cochleae. The canaliculus cochleae for the aqueduct of the cochlea exits the inner
ear from this excavation, and the canaliculus forms a long tunnel (10.86 mm in length)
ending in a triangular cavity in Homo. The canaliculus forms a flattened, ever-expanded
passage in Macaca (3.53 mm in length).
The apical turns of the cochlea are separated from the basal whorl (fitting inside
of the basal whorl when the cochlea is viewed down its axis of rotation), and the apical
turns sit on top of one another. A difference between the angle formed by the planes of
361
FIGURE 5.62. Bony labyrinth of Macaca mulatta. A, stereopair and labeled line drawing
of digital endocast in anterior view; B, stereopair and labeled line drawing of digital
endocast in dorsal view; C, stereopair and labeled line drawing of digital endocast in
lateral view; D, line drawing of cochlea viewed down axis of rotation to display degree of
coiling; E, line drawing of cochlea in profile. Abbreviations: aa, anterior ampulla; ac,
anterior semicircular canal; ant, anterior direction; av, bony channel for aqueduct of
vestibule; cc, canaliculus cochleae for aqueduct of cochlea; co, cochlea; cr, common
crus; dor, dorsal direction; er, elliptical recess of vestibule; fc, fenestra cochleae; fv,
fenestra vestibuli; la, lateral ampulla; lc, lateral semicircular canal; med, medial
direction; pa, posterior ampulla; pc, posterior semicircular canal; pl, primary bony
lamina; pos, posterior direction; sl, secondary bony lamina; sr, spherical recess of
vestibule.
362
A
B
C
D E
co
pl
5 mm
fc
fv
sr
er
aa
la
av
ac
cr
pc
pa
lc
co
pl
cc
av
cr
pc
er
aa
la
pa
ac
lc
co
fv
lc
aa
la
ac
cr
cc
fc
sl
pa
av
pc
5 mm
5 mm
fc
co
pl
sl
co
cc
fc
dor
med
med
pos
dor
ant
363
FIGURE 5.63. CT slices through ear region of Macaca mulatta. Abbreviations: aa,
anterior ampulla; ac, anterior semicircular canal; ant, anterior direction; av, bony channel
for aqueduct of vestibule; cc, canaliculus cochleae for aqueduct of cochlea; cf, foramina
within cribriform plate; cn, canal for cranial nerve VIII; co, cochlea; cr, common crus;
dor, dorsal direction; er, elliptical recess of vestibule; fc, fenestra cochleae; fn, canal for
cranial nerve VII; fv, fenestra vestibuli; in, incus; la, lateral ampulla; lat, lateral
direction; lc, lateral semicircular canal; ma, malleus; med, medial direction; pa, posterior
ampulla; pc, posterior semicircular canal; pl, primary bony lamina; sa, subarcuate fossa;
sl, secondary bony lamina; st, stapes within fenestra vestibuli; vb, vestibule; vn, canal for
vestibular branch of cranial nerve VIII.
364
dor
ant
320
354
388
422
456
490
524
558
592
660
620
5 mm
lat
ant
co
co
co
cn
cn
fn
fn
cc
cf
cf
st
in
ma
ma
ma
lc
co
vb
vn
vb
er
cr
cr
la
lc
lc
aa
ac
ac
sa
sa
pc
pl
pl
sl
pl
pc
pa
av
fn
1 mm
320 354
422388 456
524490 558
620592 660
pc pc
pc
pc
cc
fc
av
av
fc
fv
365
FIGURE 5.64. Bony labyrinth of Homo sapiens. A, stereopair and labeled line drawing
of digital endocast in anterior view; B, stereopair and labeled line drawing of digital
endocast in dorsal view; C, stereopair and labeled line drawing of digital endocast in
lateral view; D, line drawing of cochlea viewed down axis of rotation to display degree of
coiling; E, line drawing of cochlea in profile. Abbreviations: aa, anterior ampulla; ac,
anterior semicircular canal; ant, anterior direction; av, bony channel for aqueduct of
vestibule; cc, canaliculus cochleae for aqueduct of cochlea; co, cochlea; cr, common
crus; dor, dorsal direction; er, elliptical recess of vestibule; fc, fenestra cochleae; fv,
fenestra vestibuli; la, lateral ampulla; lc, lateral semicircular canal; med, medial
direction; pa, posterior ampulla; pc, posterior semicircular canal; pl, primary bony
lamina; pos, posterior direction; sr, spherical recess of vestibule.
366
A
B
C
D E
co
pl
fc
fv
sr
er
aa
la
cc
cc
cc
pc
pa
lc
ac
co
pl
er
aa
5 mm
ac
la
av
av
cr
pc
lc
sr
5 mm
5 mm
co
la
aa
ac
cr
lc
fv
fc
pa
pc
av
co
fc
co
fc
dor
med
med
pos
dor
ant
367
FIGURE 5.65. CT slices through ear region of Homo sapiens. Abbreviations: aa, anterior
ampulla; ac, anterior semicircular canal; ant, anterior direction; av, bony channel for
aqueduct of vestibule; cc, canaliculus cochleae for aqueduct of cochlea; cf, foramina
within cribriform plate; cn, canal for cranial nerve VIII; co, cochlea; cr, common crus; fc,
fenestra cochleae; fn, canal for cranial nerve VII; fv, fenestra vestibuli; la, lateral
ampulla; lat, lateral direction; lc, lateral semicircular canal; pa, posterior ampulla; pc,
posterior semicircular canal; pl, primary bony lamina; sg, spiral ganglion within primary
lamina; sr, spherical recess of vestibule; st, stapes within fenestra vestibuli; vb, vestibule;
ven, ventral direction; vn, canal for vestibular branch of cranial nerve VIII.
368
lat
ant
25
86
147
208
269
330
391
452
513
635
574
5 mm
ven
co
co
cn
cn
cn
sr
fc
pa
ac
lc
lc
pc
pc
ac
fv
vb
vb
pa
aa
la
av
fn
pl
ant
ant
1 mm
25
25
86
86
208
208
147
147
269
269
391
391
330
330
452
452
574
574
513
513
635
635
pl
sg
sg
pl
cf
cf
lc
pl
vn
pc
pc
av
fv
cr
369
the basal turn of the cochlea and lateral semicircular canal was measured between the
primate species, where the angle was much larger in Homo (62.4°) than measured in
Macaca (47.8°). The angle between the cochlea and lateral canal of Macaca is similar to
that observed in the elephantoid proboscidean (48.5°), Procavia (45.4°), and Felis
(45.8°), but the angle in Homo is greater than that in any other mammal.
The fenestrae vestibuli of the primates are among the most elliptical fenestrae
among the mammals examined here (stapedial ratio equals 2.5 in Macaca; ratio equals
3.0 in Homo), similar to Cavia porcellus (ratio equals 2.9). The vestibule is constricted
internal to the fenestra vestibuli, thereby defining the border between the spherical and
elliptical recesses. The bony channel for the aqueduct of the vestibule leaves the inner ear
dorsal to the medial edge of the common crus. The channel, which is robust in Homo
(5.47 mm in length; length equals 3.76 mm in Macaca), terminates as a fissure in both
species,. The semicircular canals of Homo are relatively thick compared to the canals of
Macaca, and the common crus of Homo is short and stout, similar to the crus in Manis
tricuspis.
The greatest angle between the planes of any two semicircular canals in primates
was measured between the anterior and posterior semicircular canals, totaling 100° for
both Macaca and Homo. The planes of the anterior and lateral semicircular canals of
Homo form a similar angle (98.9°; 83.1° for Macaca), and the angle between the
posterior and lateral semicircular canals is 89.8° for Homo and 89.0° for Macaca.
The anterior semicircular canal is the least planar canal in each primate, with a
total angular deviation of 26.4° in Macaca and 19.5° in Homo. The total deviation of the
lateral semicircular canal in Macaca (7.7°) is less than that measured for the posterior
canal (11.8°), and the posterior semicircular canal is the most planar within the labyrinth
of Homo (total deviation of 6.3° versus 7.1° calculated for the lateral canal). The degree
370
of deviation of the anterior canal is significant for both primates (ratio of total linear
deviation to cross-sectional diameter is 1.08 for Homo; 3.75 for Macaca), but only the
posterior canal of Macaca deviates significantly (ratio is 1.11; 0.6 for Homo). The degree
of deviation of the lateral canal is not significant in either species (ratio is 0.34 for Homo;
0.66 for Macaca).
The posterior limb of the lateral semicircular canal opens directly into the
vestibule, nearly equidistant between the vestibular apertures of the common crus and
posterior ampulla, in both primate species. The sagittal labyrinthine indices of Macaca
(50.1) and Homo (55.8) are greater than that calculated for any other mammal discussed
here. The closest non-primate to approach this index is Procavia (44.9).
The posterior semicircular canal of Homo is the largest in all dimensions explored
in this study, including the arc radius (3.10 mm; 2.94 mm for anterior; 2.35 mm for
lateral), for which the anterior canal has the greatest value in most mammals. In fact, the
radius of the anterior semicircular canal (2.70 mm) is greater than either the lateral (2.47
mm) or posterior canal (2.54 mm) in Macaca. The slender canal lengths of the posterior
semicircular canals of both Homo and Macaca (14.73 mm and 13.05 mm respectively)
are greater than either the anterior (13.55 mm and 12.79 mm respectively) or lateral
canals (10.31 mm and 10.57 mm). The posterior semicircular canal of Homo has a cross-
sectional diameter of 1.12 mm (0.92 mm for anterior; 0.86 mm), which is over twice as
large as the diameter measured in Macaca (0.47 mm; 0.33 mm for anterior; 0.50 mm for
lateral).
The aspect ratios of the arcs of the semicircular canals are similar between the two
primate taxa. The ratios of the anterior semicircular canals are 0.87 and 0.86 for Macaca
and Homo respectively, while the ratios for the lateral canals are 0.89 and 0.85
respectively. The highest aspect ratio in each species was calculated for the posterior
371
canal arc (0.98 in Macaca; 1.08 in Homo), where the heights and widths of the canal arcs
are nearly identical. The ratio of the slender canal length to arc radius of the posterior
semicircular canal is larger than the other canals in both Macaca (5.13; 4.74 for anterior;
4.29 for lateral) and Homo (4.76; 4.61 for anterior; 4.39 for lateral).
There are no unambiguous synapomorphies in the bony labyrinth to support
monophyly of Primates, and the clade retains the ancestral primatomorphan morphology
of the cochlear spiral in that the cochlea has a low aspect ratio in profile. The anterior
semicircular canal arc has the largest radius of curvature, which is retained from the
ancestor to Theria, although the greatest radius in Homo was measured for the posterior
canal arc. The arc of the posterior semicircular canal of no other euarchontoglire is the
largest in terms of radius of curvature, and the only mammals for which the posterior
canal arc is the greatest are Manis (only member of Laurasiatheria with the posterior
canal the greatest), Dasypus (the distribution within Xenarthra beyond this taxon is
unknown), and Orycteropus and Procavia among afrotherians.
The ancestral primate retained the ancestral placental condition of the direct
vestibular entrance of lateral semicircular canals in the absence of a secondary common
crus, and the plane of the lateral canal was high relative to the ampullar entrance of the
posterior canal, which was retained from the ancestor of Boreoeutheria, if not earlier
(state is equivocal for Placentalia). The cochlea of the ancestor of Primates coiled 980.2°,
and the cochlea contributed 48.0% of the total labyrinthine volume.
Dermoptera
The colugos are gliding mammals divided into two extant species, Cynocephalus
volans and Galeopterus variegatus (Wilson and Reeder, 2005), and the bony labyrinth of
Cynocephalus is used as a representative of Dermoptera. Phylogenetic analyses based on
372
molecular data reconstruct a close relationship between Primates and Dermoptera
(Schmitz and Zischler, 2003; Bininda-Emonds et al., 2007; Jane"ka et al., 2007), with the
occasional result of Dermoptera nested within Primates (Murphy et al., 2001a; Schmitz et
al., 2002)
Although the average body mass of Cynocephalus volans is less than Sylvilagus
floridanus (1.0 versus 1.2 kg; Silva and Downing, 1995), the dimensions of the inner ear
of the colugo are greater than that for the rabbit. The gross volume of the bony labyrinth
of Cynocephalus is 20.32 mm
3
(versus 11.32 mm
3
for Sylvilagus), and the anterior-
posterior length of the labyrinth is 7.17 mm (versus 5.82 mm for Sylvilagus). The
cochlear canal of Cynocephalus contributes 48.4% of the total labyrinthine volume (9.83
mm
3
), which is similar to the contribution calculated for Homo sapiens (50.3%). The
cochlear spiral completes nearly two and two thirds whorls (954°), and the secondary
bony lamina persists for around one fifth of the basal turn of the cochlea (65.4°), as is
illustrated in Figures 5.66-5.67.
The cochlear canal is 12.20 mm in length, and the apical turns of the cochlea fit
within the basal coils when the cochlea is viewed down its axis of rotation. The bony
canaliculus cochleae (0.90 mm in length) is developed as a delicate tube that curves along
its course. A second channel, which likely carried a blood vessel in life, extends away
from the bony labyrinth alongside the canaliculus cochleae. The planes of the basal turn
of the cochlea and lateral semicircular canal form an angle of 34.6°.
The fenestra vestibuli is elliptical, with a stapedial ratio of 2.0 (the same value
calculated for Tadarida brasiliensis), and a constriction of the vestibule internal to the
fenestra vestibuli can be used to distinguish between the spherical and elliptical recesses.
The ampullae are very round in Cynocephalus, and the posterior limb of the lateral
semicircular canal opens directly into the vestibule immediately dorsal to the vestibular
373
FIGURE 5.66. Bony labyrinth of Cynocephalus volans. A, stereopair and labeled line
drawing of digital endocast in anterior view; B, stereopair and labeled line drawing of
digital endocast in dorsal view; C, stereopair and labeled line drawing of digital endocast
in lateral view; D, line drawing of cochlea viewed down axis of rotation to display degree
of coiling; E, line drawing of cochlea in profile. Abbreviations: aa, anterior ampulla; ac,
anterior semicircular canal; ant, anterior direction; av, bony channel for aqueduct of
vestibule; cc, canaliculus cochleae for aqueduct of cochlea; co, cochlea; cr, common
crus; er, elliptical recess of vestibule; dor, dorsal direction; fc, fenestra cochleae; fv,
fenestra vestibuli; la, lateral ampulla; lc, lateral semicircular canal; med, medial
direction; pa, posterior ampulla; pc, posterior semicircular canal; pl, primary bony
lamina; pos, posterior direction; sl, secondary bony lamina; sr, spherical recess of
vestibule.
374
co
sr
er
aa
la
fc
pa
av
cr
ac
pc
lc
fv
A
B
C
D E
co
pl
av
cr
er
pc
fc
sr
aa
ac
la
lc
1 mm
co
fv
fc
ac
aa
la
lc
cr
cc
av
pa
pc
1 mm
1 mm
fc
co
co
sl
fc
cc
dor
med
med
pos
dor
ant
375
FIGURE 5.67. CT slices through ear region of Cynocephalus volans. Abbreviations: aa,
anterior ampulla; ac, anterior semicircular canal; av, bony channel for aqueduct of
vestibule; cn, canal for cranial nerve VIII; co, cochlea; cr, common crus; dor, dorsal
direction; fc, fenestra cochleae; fn, canal for cranial nerve VII; fv, fenestra vestibuli; in,
incus; la, lateral ampulla; lc, lateral semicircular canal; ma, malleus; med, medial
direction; pa, posterior ampulla; pc, posterior semicircular canal; pl, primary bony
lamina; pos, posterior direction; sa, subarcuate fossa; sg, spiral ganglion within primary
lamina; sl, spiral bony lamina; sr, spherical recess of vestibule; st, stapes fallen into
vestibule; vb, vestibule; vn, canal for vestibular branch of cranial nerve VIII.
376
pos
dor
38
64
90
116
142
168
194
220
246
298
272
2 mm
med
co
co
cn
cn
cn
cn
sr
fn
fn
vn
vb
vb
sg
pl
pl
pl
dor
1 mm
38 64
11690 142
194168 220
272246 298
ma
ma
in
lc
in
st
st
ac
ac
sa
sa
sa
sa
ac
ac
ac
cr
aa
la
la
pa
lc
lc
lc
pc
pc
lc
pc
pc
pa
av
cr
av
av
sl
fv
ac
fv
fc
377
aperture of the posterior ampulla (sagittal labyrinthine index equals 30.8). The bony
channel for the aqueduct of the vestibule (1.80 mm in length) is a straight tube that exits
the inner ear cavities medial to the vestibular aperture of the common crus.
The planes of the anterior and posterior semicircular canals of Cynocephalus form
a 90° angle with each other, and the other angles between semicircular canals are not far
off. The angle between the anterior and lateral canals is 92.2°, and an angle of 91.8° was
measured between the posterior and lateral semicircular canals. The anterior canal is the
least planar of the three semicircular canals with a total angular deviation from its plane
of 13.5°, compared to 3.5° and 3.0° calculated for the lateral and posterior semicircular
canals respectively. Likewise, the anterior canal deviates to a significant degree (ratio of
total linear deviation over cross-sectional canal diameter is 1.64), but the deviations of
the lateral and posterior canals are not significant (ratios are 0.24 and 0.28 respectively).
The anterior semicircular canal is the largest in terms of arc radius (1.93 mm
versus 1.47 and 1.70 mm for the lateral and posterior canals respectively) and slender
canal length (9.93 mm; 6.99 mm for lateral; 8.38 mm for posterior). This pattern is
observed in most of the mammals considered for this study. However, the lateral
semicircular canal is the largest in terms of cross-sectional diameter (0.37 mm; 0.27 mm
for anterior; 0.32 for posterior).
The arcs of the semicircular canals are graceful and circular, particularly the arc
of the posterior canal, which has an aspect ratio of 1.05 (similar to the ratio of 1.08
calculated for Homo). The aspect ratios of the anterior and lateral semicircular canals are
0.82 and 0.85 respectively. The ratio of the slender canal length to arc radius of the
anterior semicircular canal of Cynocephalus (5.15) is greatest among the canals (ratio of
lateral equals 4.75; ratio of posterior equals 4.94), which is different than the condition in
378
most of the mammals examined here, where the greatest ratio is observed in the posterior
semicircular canal.
The bony labyrinth of Cynocephalus retains all states reconstructed in the most
recent common ancestor of Primatomorpha (Primates plus Dermoptera). The aspect ratio
of the cochlea is low (retained from Primatomorpha), the lateral semicircular canal is
high compared to the ampullar opening of the posterior semicircular canal (retained from
Boreoeutheria), the lateral canal opens into the vestibule directly in the absence of a
secondary common crus (retained from Placentalia), and the greatest arc radius of
curvature was measured for the anterior semicircular canal (retained from Theria). The
contribution of the cochlea calculated for Cynocephalus (48.0%) is retained from the
ancestor of Primatomorpha (50.0%), and the coiling of the cochlea (953.7°) is similar to
that reconstructed for the ancestor of Euarchontoglires (956.9°).
Scandentia
The final species to be considered here is the tree shrew, Tupaia glis (Figures
5.68-5.69). Scandentians were considered to have “insectivoran” affinities in early
classifications (e.g., Flower, 1883), as well as close associations with Macroscelidea
(Gregory, 1910). The results of later studies have been used to remove tree shrews from
Lipotyphla and to postulate a closer relationship between Scandentia and Primates, at
times with tree shrews included within Primates (Carlsson, 1922; Simpson, 1945). Most
mammalian systematists today agree that Scandentia is a clade exclusive of Primates
(Van Valen, 1965; Campbell, 1966; Butler, 1972), and the majority of anatomical and
molecular evidence supports the monophyly of Euarchonta (Primates, Dermoptera,
Scandentia), even if the relationships within the clade remain unresolved (Adkins and
379
FIGURE 5.68. Bony labyrinth of Tupaia glis. A, stereopair and labeled line drawing of
digital endocast in anterior view; B, stereopair and labeled line drawing of digital
endocast in dorsal view; C, stereopair and labeled line drawing of digital endocast in
lateral view; D, line drawing of cochlea viewed down axis of rotation to display degree of
coiling; E, line drawing of cochlea in profile. Abbreviations: aa, anterior ampulla; ac,
anterior semicircular canal; ant, anterior direction; av, bony channel for aqueduct of
vestibule; cc, canaliculus cochleae for aqueduct of cochlea; co, cochlea; cr, common
crus; dor, dorsal direction; er, elliptical recess of vestibule; fc, fenestra cochleae; fv,
fenestra vestibuli; la, lateral ampulla; lc, lateral semicircular canal; med, medial
direction; pa, posterior ampulla; pc, posterior semicircular canal; pl, primary bony
lamina; pos, posterior direction; sl, secondary bony lamina; sr, spherical recess of
vestibule.
380
A
B
C
D E
co
fv
fc
sr
er
aa
la
pa
cr
ac
ac
pc
lc
sl
co
er
av
pa
la
aa
fc
ac
ac
lc
1 mm
pc
pl
aa
la
cr
av
pa
lc
sl
fv
fcco
pc
cc
1 mm
1 mm
co
fc
sl
co
sl
cc
fc
dor
med
med
pos
dor
ant
381
FIGURE 5.69. CT slices through ear region of Tupaia glis. Abbreviations: aa, anterior
ampulla; ac, anterior semicircular canal; av, bony channel for aqueduct of vestibule; cn,
canal for cranial nerve VIII; co, cochlea; cr, common crus; la, lateral ampulla; lat, lateral
direction; lc, lateral semicircular canal; pa, posterior ampulla; pc, posterior semicircular
canal; pl, primary bony lamina; pos, posterior direction; sa, subarcuate fossa; st, stapes
within fenestra vestibuli; vb, vestibule; ven, ventral direction.
382
pos
ven
282
300
318
336
354
372
390
408
426
462
444
1 mm
lat
co
co
co
vb
vb
vb
vb
st
lc
lc
aa
sa
sa
sa
sa
sa
aa
av
ac
ac
ac
la
la
cn
ven
1 mm
282 300
354336318
408390372
462444426
ac
ac
cr
cr
lc lc
lc
lc
lc
av
av
pc
pc
pc
pc
pa
cn
cn
cn
pl
st
383
Honeycutt, 1991; Murphy et al., 2001a, b; Silcox et al., 2005; Bininda-Emonds et al.,
2007).
The gross volume of the inner ear cavities of Tupaia is 9.83 mm
3
, and the
anterior-posterior length of the bony labyrinth is 6.67 mm. The length of the labyrinth of
Tupaia is similar to that measured for the dermopteran Cynocephalus volans (7.17 mm),
despite a body mass of Cynocephalus (1 kg; Silva and Downing, 1995) that is one order
of magnitude larger than that of Tupaia (131 g; Silva and Downing, 1995). The volume
of the cochlea of Tupaia is 5.43 mm
3
, which is 55.2% of the total inner ear volume.
The cochlea of the tree shrew completes over three turns (1125°), and the
secondary bony lamina extends beyond half of the basal coil (220°). The length of the
cochlear canal is 10.51 mm, and the aspect ratio of the cochlear spiral in profile is 0.66,
which is among the highest calculated among the mammal sample. The scala tympani of
the cochlea is expanded internal to the fenestra cochleae, from which the canaliculus
cochleae exits the cochlea. The canaliculus is a straight tube that extends posterodorsally
from the scala tympani. The canaliculus is 0.66 mm in length. The planes of the basal
turn of the cochlea and lateral semicircular canal form an angle of 29°.
The shape of the fenestra vestibuli of Tupaia is elliptical, with a stapedial (aspect)
ratio of 2.6, similar to the rhesus monkey, Macaca mulatta (2.53). A slight constriction of
the vestibule internal to the fenestra vestibuli separates the spherical and elliptical
recesses. The bony channel for the aqueduct of the vestibule opens immediately anterior
to the medial edge of the vestibular aperture of the common crus. The channel, which is
2.61 mm in length, curves posteriorly and terminates in a triangular fissure. The posterior
limb of the lateral semicircular canal does not open directly into the vestibule, but rather
into the anterior aspect of the posterior ampulla. A groove extends from the entry point of
the lateral canal to the vestibule across the anterior wall of the posterior ampulla (the
384
groove is expressed as a rounded ridge on the endocast). The lateral semicircular canal
does not join with the posterior canal, and a secondary common crus is not formed.
The planes of the anterior and posterior semicircular canals form an angle of 106°,
and the planes of the posterior and lateral canals form an angle of 102°. The angle
between the planes of the anterior and lateral semicircular canals is significantly smaller
(82°). The anterior semicircular canal is the least planar among the three canals with a
total deviation of 23.1°. The lateral and posterior semicircular canals deviate from their
planes by a total of 8.4° and 5.4° respectively. Both the anterior and posterior
semicircular canals deviate from their planes by a significant degree (ratios of total linear
deviation over cross-sectional diameter are 3.76 and 1.54 respectively). The lateral
semicircular canal does not deviate from its plane significantly (ratio is 0.97), although
nearly so.
The semicircular canals form delicate arcs, and the canals themselves are slender
compared to the rest of bony labyrinth (as opposed to the thick canals observed in Homo).
The radius of the arc of the anterior semicircular canal of Tupaia (1.73 mm) is greater
than either the lateral (1.44 mm) or posterior canals (1.50 mm). Similarly, the slender
canal length of the anterior semicircular canal is the greatest (9.24 mm; 7.84 mm for
lateral; 8.07 mm for posterior). The lateral semicircular canal is largest in terms of cross-
sectional diameter (0.22 mm; anterior and posterior canals have a diameter of 0.22 mm
each).
The arc of the anterior semicircular canal appears more circular than the lateral
and posterior canal arcs (Figure 5.63), although the aspect ratio of the posterior canal arc
(0.96) is higher than that calculated for the anterior canal (0.85). This is a result of the
method employed to measure the height and width of the posterior semicircular canal arc,
which does not reflect the shape of the arc in this situation. The aspect ratio of the lateral
385
semicircular canal is 0.71, which more accurately represents the ellipse formed by the
lateral canal arc. The ratio of the slender canal length to arc radius of the anterior
semicircular canal (5.35) is not the greatest, as is observed in most of the mammals
examined here, but rather the smallest. The ratio for the lateral canal (5.46) is the largest,
and the ratio for the posterior canal falls in between (5.40).
The bony labyrinth of Tupaia is derived from the ancestral eutherian condition in
that the plane of the lateral semicircular canal is high in relation to the ampullar entrance
of the posterior canal, although this state was inherited from the most recent common
ancestor of boreoeutherians. The lateral semicircular canal opens into the posterior
ampulla separate from the posterior canal (a secondary common crus is not formed), a
condition that is unique to Tupaia among euarchontoglires, but shared by Hemicentetes,
Cetacea, Equus, Carnivora (except Canis), and the bats Nycteris and Rhinolophus. The
greatest arc radius of curvature was measured for the anterior semicircular canal in
Tupaia, which is consistent for most of the therian mammals considered here.
The high aspect ratio of the cochlea of Tupaia is derived from the ancestral
eutherian condition, which the taxon shares with Glires within Euarchontoglires. The
shape of the cochlear spiral may be a synapomorphy supporting a Tupaia plus Glires
clade, although the ancestral state of Euarchontoglires is equivocal with respect to this
character. The cochlea coils to a greater degree (1125.0°) than that reconstructed for the
ancestor of Euarchontoglires (956.9°), but not more than a quarter turn. The cochlea of
Scandentia contributes 55.0% of the total labyrinthine volume, which is he same
percentage calculated for the cochlea of Boreoeutheria.
386
RESULTS – DIMENSION COMPARISONS
Large-bodied animals tend towards large inner ears. For example, the labyrinths
of large-bodied Homo sapiens and Equus caballus are among the most voluminous, while
the inner ears of Mus musculus and Sorex monticolus are the smallest. In order to test if
there is a correlation between body size and inner ear dimensions, the coefficient of
correlation was calculated between specific measurements and body mass (Table 5.7).
The total size of the bony labyrinth, both in terms of the total volume of the cavities and
length of the inner ear, are related strongly to body mass across the sample when the data
are transformed using the natural logarithm (Figure 5.70). A coefficient of correlation (r;
not to be confused by the slender canal radius of Jones and Spells, 1963) of 0.94 was
calculated between labyrinth length and body mass, and a coefficient of 0.95 between
total labyrinthine volume and body mass.
Because the size of the bony labyrinth is closely correlated to the overall size of
the animal, the dimensions of the bony labyrinth can be used to estimate the body mass of
fossil species. The length of the bony labyrinth rather than its volume is used here to
make this estimate, because it is less prone to error. Volumes of the inner ear constituents
are calculated from the segmented endocast, where boundaries between the middle and
inner ear cavities, or else between the cochlea and vestibule, can be ambiguous for some
species. Consistent boundaries are maintained as much as possible, but the longitudinal
measure of the length of the bony labyrinth is less ambiguous.
The equation for the regression line between the length of the bony labyrinth and
body mass is y=0.151x+0.8212, where “x” equals the body mass and “y” equals the
length of the bony labyrinth. The accuracy of the equation can be tested by estimating the
body mass was of Canis familiaris specimen, which was not incorporated in the
correlations with body mass (see discussion in the materials and methods section). Using
387
TABLE 5.7. Coefficients of correlation (r) calculates for dimensions over body mass
a
Measurement
Coefficient of Correlation - r
Labyrinth
Volume
0.95
Length
0.94
Cochlea
Volume
0.93
Percent of Total Volume
0.13
Canal Length
0.84
Aqueduct Length
0.91
Coiling
0.02
2° Lamina Extension
0.36
Angle with Lateral Canal
0.36
Aspect Ratio
0.19
Vestibule
Aqueduct Length
0.70
Stapedial Ratio
0.26
Semicircular Canal Orientation
Anterior-Lateral
0.15
Anterior-Posterior
0.07
Lateral-Posterior
0.33
Semicircular Canal Dimensions
Anterior Radius
0.85
Lateral Radius
0.88
Posterior Radius
0.84
Anterior Length
0.79
Lateral Length
0.83
Posterior Length
0.83
Anterior Diameter
0.83
Lateral Diameter
0.81
Posterior Diameter
0.82
Anterior Linear Deviation
0.33
Lateral Linear Deviation
0.64
Posterior Linear Deviation
0.35
Anterior Angular Deviation
0.17
Lateral Angular Deviation
0.19
Posterior Angular Deviation
0.08
Anterior Aspect Ratio
0.19
388
TABLE 5.7. (Continued)
Measurement
Coefficient of Correlation - r
Lateral Aspect Ratio
0.50
Posterior Aspect Ratio
0.61
a
Data logarithmically transformed using the natural logarithm. Values over 0.70 (in bold)
are considered significant.
389
FIGURE 5.70. Bivariate plots of labyrinth dimensions over body mass. All data
logarithmically transformed using the natural logarithm (ln). A, total labyrinth volume
over body mass; B, total length of labyrinth over body mass; C, volume of cochlea over
body mass; D, length of canaliculus cochleae for aqueduct of cochlea over body mass; E,
length of bony channel for aqueduct of vestibule over body mass; F, length of cochlear
canal over body mass; G, length of slender anterior semicircular canal over body mass;
H, length of slender lateral semicircular canal over body mass; I, length of slender
posterior semicircular canal over body mass; J, arc radius of curvature of anterior
semicircular canal over body mass; K, arc radius of curvature of lateral semicircular
canal over body mass; L, arc radius of curvature of posterior semicircular canal over
body mass. The outlier that falls well below regression line in G-L is Tursiops truncatus.
390
ln labyrinth volume
ln body mass
-1.00
1.00
0.00
3.00
4.00
2.00
5.00
6.00
7.00
0.00 10.00 15.005.00
ln labyrinth length
ln body mass
0.00
0.00
1.00
2.00
3.00
2.50
1.50
0.50
3.50
10.00 15.005.00
ln cochlea volume
ln body mass
0.00
-2.00
-1.00
0.00
1.00
3.00
4.00
5.00
6.00
7.00
2.00
10.00 15.005.00
ln coch. aq. length
ln body mass
0.00
-3.00
-2.00
-1.00
0.00
1.00
2.00
3.00
10.00 15.005.00
ln vest. aq. length
ln body mass
0.00
-3.00
-2.00
-1.00
0.00
1.00
2.00
3.00
10.00 15.005.00
ln cochlea length
ln body mass
0.00
0.00
0.50
1.00
1.50
2.00
2.50
3.00
3.50
10.00 15.005.00
ln ant. canal length
ln body mass
0.00 10.00 15.005.00
0.00
0.50
1.00
1.50
2.00
2.50
3.00
3.50
ln lat. canal length
ln body mass
0.00 10.00 15.005.00
0.00
0.50
1.00
1.50
2.00
2.50
3.00
3.50
ln post. canal length
ln body mass
0.00 10.00 15.005.00
0.00
0.50
1.00
1.50
2.00
2.50
3.00
3.50
ln ant. canal radius
ln body mass
0.00
-1.00
-0.50
0.00
0.50
1.00
1.50
2.00
10.00 15.005.00
ln lat. canal radius
ln body mass
0.00
-1.00
-0.50
0.00
0.50
1.00
1.50
2.00
10.00 15.005.00
ln post. canal radius
ln body mass
0.00
-1.00
-0.50
0.00
0.50
1.00
1.50
2.00
10.00 15.005.00
A B C
D E F
G H I
J K L
391
the equation above, the estimated body mass of the specimen is 4.5 kg, which is at the
low range of dog body masses (3.4 to 31.3 kg as reported by Galvão, 1947). In fact, the
breed of dog used in this study is from a Chihuahua, which is among the smallest breeds
of domestic dog. This equation can be used to calculate body masses for extinct taxa. For
example, the estimated body masses of the oreodont Bathygenys reevesi and the fossil
balaenopterid whale are 2.5 kg and 1608.3 kg respectively.
Additional dimensions significantly scale with body mass (see Table 5.7),
including the length and volume of the cochlear canal (r equals 0.84 and 0.93
respectively), lengths of the bony channels for the aqueducts of the cochlea (canaliculus
cochleae; r equals 0.91) and vestibule (r equals 0.70), slender canal lengths of the anterior
(r equals 0.79), lateral (r equals 0.83), and posterior semicircular canals (0.83), as well as
the radii of curvature of the semicircular canal arcs (r equals 0.85 for anterior; 0.88 for
lateral; 0.84 for posterior). In all cases, the mammals with the largest inner ear
dimensions are large-bodied animals (see Figure 5.70). Alternatively, the aspect ratios of
the cochlea (r equals 0.19) and semicircular canal arcs (0.19 for anterior; 0.50 for lateral;
0.61 for posterior) do not correlate with body mass.
The degree of coiling exhibited by the cochlea and body mass do not correlated
with one another (r equals 0.02). That is, large animals, such as Equus caballus, do not
have a significantly more or less coiled cochlea than smaller species. Species with a large
number of cochlear whorls do not have significantly more voluminous cochleae (r equals
0.07), longer cochlear canals (r equals 0.36), or cochlear spirals with higher aspect ratios
(r equals 0.44), as summarized in Table 5.8. Likewise, a long cochlear canal does not
signify a high-spired cochlea (r equals 0.10), as observed in Cavia porcellus. Table 5.9
summarizes correlations among dimensions of the semicircular canals.
392
TABLE 5.8. Coefficients of correlation (r) calculated for dimensions of the cochlea
a
Degree of Coiling
Canal Volume
Canal Length
Aspect Ratio
Degree of Coiling
-
0.07
0.36
0.44
Canal Volume
0.07
-
0.94
0.22
Canal Length
0.36
0.94
-
0.10
Aspect Ratio
0.44
0.22
0.10
-
a
Data logarithmically transformed using the natural logarithm. Values over 0.70 (in bold)
are considered significant.
393
TABLE 5.9. Coefficients of correlation (r) calculated for dimensions of the semicircular
canals
a
Ant
Lat
Post
Radius
Length
Ratio
Radius
Length
Ratio
Radius
Length
Ratio
Radius
-
0.98
0.26
-
0.98
0.26
-
0.99
0.40
Length
0.98
-
0.31
0.98
-
0.29
0.99
-
0.39
Ratio
0.26
0.31
-
0.26
0.29
-
0.40
0.39
-
a
Data logarithmically transformed using the natural logarithm. Values over 0.70 (in bold)
are considered significant.
394
In short, strong correlations were observed between the semicircular canal arc
radii of curvature and slender canal length (r ranging from 0.98 to 0.99). The correlations
between the aspect ratio of a canal arc and respective length or arc radius are not
significant (Table 5.9).
DISCUSSION
General Patterns
That variation was observed across the sample of bony labyrinths of placental
mammals is not surprising, and it has been a long-recognized phenomenon, even before
the seminal works of Gustaf Retzius in the late 19
th
Century (for example, see Retzius,
1884). However, the nature of this variation has received only a superficial treatment in
the scientific literature; exceptions include Caix and Outrequin (1979), Dimopoulos and
Muren (1990), and Chapter 4 of this dissertation. Variation in the degree of coiling in the
cochlea in particular is related to phylogeny (e.g., Meng and Fox, 1995) and function
(West, 1985). The broad range of over 1,400° (nearly 3 turns) within the placental sample
examined here may be a reflection of taxonomic diversity, where at least 5,421 extant
mammal species are recognized (Wilson and Reeder, 2005; Reeder et al., 2007), as well
as physiological diversity, where a range of auditory sensitivities extend from subsonic
(in proboscideans and cetaceans; Payne et al., 1986; Poole et al., 1988; Ketten, 1997) to
ultrasonic frequencies (in some chiropterans, soricid lipotyphlans, and tenrecs; Gould,
1965; Tomasi, 1979; Simmons et al., 1979).
Other general patterns in the bony labyrinth anatomy include the arc radius of
curvature of the anterior semicircular canal being the largest among the three canals in
395
the majority of the mammals examined here (24 out of 32 species). This pattern has been
observed in most mammal species (Curthoys et al., 1977a, b; Blanks et al., 1985; Muren
et al., 1986; Spoor and Zonneveld, 1998; Jeffery and Spoor, 2004; Calabrese and Hullar,
2006; Spoor et al., 2007), and a large anterior semicircular canal arc signifies that the
majority of mammals are most sensitive to rotational head movement in the pitch
(anterior-posterior) direction (Yang and Hullar, 2007). Exceptions include Dasypus
novemcinctus, where the posterior canal is the most sensitive, or Eumetopias jubatus
where the lateral canal would be most sensitive.
The posterior semicircular canal is the least planar of the three canals in most of
the bony labyrinths studied here (15 out of 32 species), and the lateral canal is the most
planar for the majority of taxa (18 out of 32 species). The ratio of the total linear
deviation to the cross-sectional diameter of the semicircular canal is used in the present
study to describe the degree of planar deviation of a semicircular canal, where a ratio
above 1 (linear deviation greater than diameter) is considered substantial. Any
physiological importance of planar deviation has yet to be explored in a rigorous sense,
and such substantial deviation may not have any basis in function. The ratio is used for
descriptive and comparative purposes only. Although the ratio is arbitrary, evidence
suggests that, even in species with broad ranges of planarities (such as in Monodelphis
domestica; Chapter 2), there is not much variation in whether or not the ratio is
substantial (Table 5.10). The degree of deviation exhibited both by the anterior and
posterior semicircular canals is considered substantial in half of the taxa examined here
(16 out of 32), although the deviations of the two canals are not always significant within
the same labyrinth. The deviation of the lateral semicircular canal is significant in only
one quarter of the mammals.
396
TABLE 5.10. Linear deviations of the semicircular canals of Monodelphis domestica
a
Specimens (TMM M)
7595
8261
8265
7536
8266
7539
7542
8267
7545
8268
8273
7599
Linear Deviations
Anterior
0.00
0.08
0.05
0.11
0.11
0.00
0.08
0.07
0.10
0.11
0.11
0.07
Lateral
0.00
0.00
0.00
0.05
0.05
0.00
0.00
0.05
0.05
0.04
0.07
0.06
Posterior
0.00
0.08
0.08
0.07
0.07
0.10
0.06
0.06
0.09
0.09
0.09
0.07
Canal Diameters
Anterior
0.20
0.21
0.29
0.19
0.19
0.21
0.23
0.17
0.17
0.17
0.23
0.18
Lateral
0.27
0.24
0.25
0.19
0.19
0.19
0.22
0.19
0.25
0.20
0.20
0.20
Posterior
0.26
0.18
0.22
0.24
0.19
0.22
0.25
0.22
0.27
0.20
0.24
0.26
Ratios
Anterior
0.00
0.38
0.17
0.58
0.58
0.00
0.35
0.41
0.59
0.65
0.48
0.39
Lateral
0.00
0.00
0.00
0.26
0.26
0.00
0.00
0.26
0.20
0.20
0.35
0.30
Posterior
0.00
0.44
0.36
0.29
0.37
0.45
0.24
0.27
0.33
0.45
0.38
0.27
a
Scanning parameters provided in Chapter 3. All specimens housed at the Texas Natural
Sciences Center, Austin Texas (TMM M). Linear dimensions expressed in millimeters.
397
Functional Considerations
A strong relationship between the size of the bony labyrinth and body mass is to
be expected. This phenomenon causes a tricky situation when morphologies within the
inner ear are used to make functional interpretations. A clear positive correlation between
the arc radius of curvature and sensitivity is evident (Yang and Hullar, 2007), with
absolutely larger canals being more sensitive to rotational movement than smaller canals.
Additionally, the size of the canals has been related to locomotor behavior (e.g., Spoor et
al., 1994, 2002), but this relationship has yet to be tested experimentally.
The size of a semicircular canal arc appears to be correlated with agility (Spoor et
al., 2007). In theory, agile (coded as “fast” by Spoor et al., 2007) mammals will have a
larger semicircular canal arc radius (averaged over the three canals within a labyrinth)
than slower animals of the same body size. The average radius of the anterior, lateral, and
posterior semicircular canals was calculated for each taxon examined in this study, and
the average was divided by body mass in order to normalize the data (Table 5.11). No
correlation is recovered when the ratios of arc radius over body mass are plotted over
agility (based on a six point scoring system developed by Spoor et al., 2007). All data
were logarithmically transformed using the natural log.
Although the radius of curvature does not correlate to the agility scores of Spoor
et al. (2007) when the radius is divided by body mass, the ratios of aquatic taxa are nearly
an order of magnitude smaller than the ratios calculated for terrestrial animals, regardless
of their evolutionary relationships (see Table 5.11). This suggests that bony labyrinth
morphology can be used to identify aquatic tendencies (Spoor et al., 2002; Boyer and
Georgi, 2007; Georgi and Sipla, 2008). For example, the size ratio between the cochlea
and vestibular apparatus of cetaceans is greater than that observed in most mammals, and
398
TABLE 5.11. Ratios of Semicircular Canal Arc Radius of Curvature over Average Body
Mass and Slender Canal Length
a
Taxon
b
Radius / Average Body Mass
Length
c
/ Radius
Ant
Lat
Post
Average
Ant
Lat
Post
Marsupialia
Didelphis
0.0523
0.0331
0.0440
0.0361
5.63
5.47
6.11
Eutheria
Kulbeckia
NA
NA
NA
NA
4.80
4.29
4.75
Ukhaatherium
NA
NA
NA
NA
4.55
4.28
4.88
Zalambdalestes
NA
NA
NA
NA
4.77
4.36
4.53
Zhelestid
NA
NA
NA
NA
4.96
4.40
5.15
Afrotheria
Chrysochloris
2.4675
1.5184
1.5980
1.8613
4.30
3.89
5.07
Hemicentetes
0.9977
0.6182
0.8045
0.8068
4.52
3.59
5.41
Macroscelides
3.4387
2.7400
2.6670
2.9490
4.25
4.00
5.10
Elephantoidea
NA
NA
NA
NA
4.93
4.70
4.41
Orycteropus
0.0052
0.0054
0.0060
0.0055
4.96
5.03
.5.39
Procavia
0.0524
0.0470
0.0570
0.0520
5.14
4.28
4.90
Trichechus
0.0009
0.0009
0.0010
0.0008
4.02
3.18
4.67
Xenarthra
Dasypus
0.0345
0.0336
0.0400
0.0362
5.91
4.63
5.88
Laurasiatheria
Atelerix
0.1431
0.1021
0.1410
0.1288
4.74
4.15
4.74
Balaenopteridae
NA
NA
NA
NA
4.19
4.05
4.94
Bathygenys
NA
NA
NA
NA
5.08
4.68
5.59
Canis
NA
NA
NA
NA
4.97
4.50
5.14
Equus
0.0014
0.0014
0.0010
0.0014
4.79
4.02
5.32
Eumetopias
0.0004
0.0004
0.0004
0.0004
4.33
4.72
4.92
Felis
0.0563
0.0494
0.0559
0.0539
4.57
4.45
4.93
Manis
0.0324
0.0236
0.0369
0.0310
4.52
3.49
4.23
Nycteris
3.3118
2.9696
2.7014
2.9942
4.48
3.91
5.51
Pteropus
0.3604
0.2951
0.3106
0.3221
4.37
4.56
5.20
Rhinolophus
4.8121
4.0187
4.3210
4.3839
4.25
4.64
5.25
Sorex
10.7230
7.9496
10.3611
9.6780
4.91
3.38
5.44
Sus
0.0028
0.0024
0.0025
0.0026
4.86
3.87
4.89
Tadarida
6.9670
6.0395
6.0601
6.3556
4.62
4.45
4.88
Tursiops
0.0007
0.0008
0.0005
0.0006
3.47
3.38
5.17
Euarchontoglires
399
TABLE 5.11. (Continued)
Taxon
b
Radius / Average Body Mass
Length
c
/ Radius
Ant
Lat
Post
Average
Ant
Lat
Post
Cavia
0.2581
0.2156
0.2240
0.2325
4.79
4.13
5.02
Cynocephalus
0.1930
0.1471
0.1695
0.1699
5.15
4.75
4.94
Homo
0.0037
0.0029
0.0039
0.0035
4.61
4.39
4.76
Lepus
0.0996
0.0707
0.0718
0.0807
4.89
4.13
4.80
Macaca
0.0578
0.0528
0.0545
0.0550
4.74
4.29
5.13
Mus
5.0069
3.8899
4.3126
4.4032
4.98
4.12
5.39
Sylvilagus
0.1600
0.1111
0.1241
0.1317
4.84
4.38
5.13
Tupaia
1.3157
1.0953
1.1412
1.1840
5.35
5.46
5.40
a
Taxonomy and systematic arrangement follows Bininda-Emonds et al.
(2007).Institutional abbreviations: AMNH, American Museum of Natural History, New
York, NY; MSW, Mortality South West; SDSNH, San Diego Society of Natural History,
San Diego, CA; TMM, Texas Natural Science Center, Austin, TX.
b
Body mass data from Loughlin et al. (1987) for Eumetopias, Ogden et al (2004) for
Homo, and Silva and Downing (1995) for remaining taxa. Ratio multiplied by 100.
c
Slender canal length defined as length of the semicircular canal minus its ampullated
end.
400
this led to a hypothesis that a reduced vestibular system is an evolutionary response to the
rapid body movements within an aquatic environment exhibited by extant cetaceans
(Spoor et al, 2002). Because the mobility of the head and neck in cetaceans nearly is
eliminated owing to fusion of cervical vertebrae in some taxa, the vestibulo-collic and
vestibulo-ocular reflexes that stabilize the head and eyes during rapid rotations of the
body are no longer effective (see Cox and Jeffery, 2008). Thus, larger, more sensitive
semicircular canals may no longer compensate for agile movements when the head is
unable to move. A reduced vestibular apparatus would reduce sensitivity of the system
(see Yang and Hullar, 2007 for a discussion of canal arc size and sensitivity), and lessen
any ill effects of an agile lifestyle with cervical fusion.
Although a fully aquatic lifestyle, increased agility, and reduced vestibular
systems are observed individually within many mammal taxa, cetaceans are unique in
having the full suite of these characteristics. For example, the vestibule and its associated
semicircular canals contribute a significantly smaller proportion of the entire bony
labyrinth in Rhinolophus ferrumequinum, similar to that observed in cetaceans, but
Rhinolophus does not inhabit an aquatic environment at all. Likewise, sirenians are fully
aquatic, but they are among the least agile mammals (as scored by Spoor et al., 2007) and
their cervical vertebrae are not fused (see Kaiser, 1974).
The low contribution of the vestibule (or inversely, the high contribution of the
cochlea) might be related to an aquatic lifestyle nonetheless. To investigate this
hypothesis, the relative contributions of the cochlea and vestibule are compared between
terrestrial and aquatic taxa. Because the bats are the only true volant mammals and their
ears likely are specialized for aerial locomotion, Chiroptera was not incorporated into this
comparison. The vestibular contribution of Tursiops and the balaenopterid (6.5% and
9.4% respectively) are less than that observed in terrestrial mammals (range of 28.3% for
401
Macroscelides to 69.4% for the elephantoid). The vestibular contribution of Trichechus
(28.9%) is on the low end of the entire mammal range, but it is still greater than the
vestibular apparati of both Macroscelides (28.3%) and Chrysochloris (28.7%), which are
strictly terrestrial. Furthermore, the vestibular apparatus of Eumetopias contributes 46.5%
of total labyrinthine volume, which is only slightly larger than the mean for terrestrial
mammals (43.7%). This suggests that aquatic behavior alone cannot explain a reduced
vestibular system.
Although the ranges of vestibular contribution overlap between the terrestrial and
aquatic samples, the means of each group may differ significantly. The small number of
aquatic species used here limits the effectiveness of statistical analysis. The hypothesis
that the mean contribution of the vestibule differs significantly between terrestrial and
aquatic mammals was tested, with a two-tailed t-test assuming unequal variances
(determined through an F-test). An a priori significance level of 0.05 was selected
(following Hammer and Harper, 2006), and the null hypothesis is that the mean
contributions of the vestibule are equal between the aquatic and terrestrial samples. The
results of the analysis (p=0.007, which is less than the significance level of 0.05) reject
the null hypothesis and suggest that the vestibular contribution of aquatic mammals is
significantly greater than that of terrestrial mammals. However, a more thorough analysis
incorporating larger sample sizes is needed before a formal conclusion can be made. The
aquatic sample is very small, and cetaceans are overrepresented (50.0% of the aquatic
taxa) within the sample, which potentially skews the analysis.
Nonetheless, the ratio of semicircular canal arc over body mass is significantly
reduced compared to terrestrial species. Furthermore, the vestibules of the two aquatic
species Trichechus manatus and Eumetopias jubatus appear as though they have been
402
compressed (see Figures 5.18 and 5.36). The deflation may, in essence, reflect a
reduction of the membranous utricle and saccule within the bony vestibule.
Further aspects of bony labyrinth morphology are thought to be related to aquatic
behavior, namely dimensions of the semicircular canals and their respective arcs. The
ratio between the length of the slender semicircular canal over arc radius reflects the
frequencies of neural impulses transduced from mechanoreceptors within the
membranous labyrinth during rotation of the canal (Boyer and Georgi, 2007). Differences
in the ratio might indicate different locomotor abilities, such as semiaquatic versus fully
terrestrial. The only pattern observed in the data presented here is that the ratios of length
to radius of the anterior semicircular canal of aquatic species tends to be less than that
calculated for their close terrestrial relatives (Table 5.11). Although a pattern is observed,
the sample size and taxonomic resolution of the current study is insufficient to postulate a
formal connection between the ratio of slender duct length to arc radius and locomotor
behavior.
The aspect ratios of the arcs of the anterior and posterior semicircular canals of
aquatic non-mammalian amniotes tend to be significantly lower than their close terrestrial
relatives (Georgi and Sipla, 2008). However, an opposite situation is observed across the
mammalian sample examined here, where the radii of the arcs of the two vertical canals
(anterior and posterior) are greater for every case in which the canals were compared
between aquatic and closely related terrestrial species (see Table 5.4). Methodological
differences in the calculation of aspect ratios between the present investigation and that
of Georgi and Sipla (2008) might account for the discrepancy in the pattern, or else
mammals may in fact exhibit the opposite pattern from other amniotes.
403
Phylogenetic Considerations
There are major structural differences within the membranous labyrinth that likely
contain a phylogenetic signal (Gray 1906, 1907). Two particular features identified by
Gray (1906), pigmentation within the membranes and size of the perilymphatic space
surrounding the membranous semicircular ducts (which are filled with endolymph
themselves). Unfortunately, neither can be assessed from the morphology of the bony
labyrinth alone. Gray (1906) considered the perilymphatic space to be an important
character in regard to mammal phylogeny, and he argued that a large space is ancestral
for mammals (given that he observed a large space within the semicircular canals of
reptiles, including birds). Unfortunately, the caliber of the bony canal approximates the
shape of the membranous duct within, but not necessarily the size (Curthoys et al.,
1977b).
Additional features that may have an importance in determining evolutionary
relationships that can be observed within the bony labyrinth include the size of otoliths
within the vestibular apparatus, coiling of the cochlea, and shape of the cochlear spiral.
The otoliths are tiny in most mammal species, although Gray (1906, 1907, 1908)
observed sizeable otoliths within the labyrinths of the marsupial Petrogale penicillata,
the cetaceans Balaena australis and Phocaena communis, the sirenian Dugong dugon,
and the pinniped carnivorans Phoca vitulina, Halichoerus grypus, and Otaria pulsilla.
However, otoliths were not observed in the CT imagery of any specimen examined in the
present study, including investigated members of Cetacea, Sirenia, and Carnivora.
There are several reasons for the absence of otoliths on the CT scans. The
composition and density of otoliths makes it virtually impossible that they would have
been missed in CT data if present. Indeed, CT scans of many non-mammalian vertebrates
(see scans of fish and squamate reptiles at www.digimorph.org; Maisano and Rieppel,
404
2007) reveal otoliths. Alternatively, otoliths may be lacking within the bony labyrinth at
the time of scanning, either because large otoliths do not occur in life in these species, or
else through loss during skeletal preparation. The latter cannot be ruled out, because
specimens representing the taxa in which Gray observed large otoliths that were used in
this study are dried skulls, and it is conceivable that the otoliths fell out of the ear cavity
once the membranes holding them decayed. Even so, otoliths were not observed in
specimens that were preserved in alcohol, such as Chrysochloris and Atelerix, thereby
preserving the membranous labyrinths with the otoliths supposedly intact.
Two cochlear shapes termed “sharp-pointed” (observed in rodents, lagomorphs,
and non-pinniped carnivorans) and “flattened” (observed in pinnipeds, primates,
cetartiodactyls and perissodactyls) were identified by Gray (1906). The distinction
between the two morphotypes is not clear, although they roughly correspond to the aspect
ratios of the cochlea reported in the present investigation. That is, the taxa with the
“sharp-pointed” condition tend towards high aspect ratios, above 0.55, whereas the aspect
ratios calculated for species with “flattened” cochleae are 0.55 or less. A couple of
exceptions to the generality are Eumetopias, which has a cochlear aspect ratio of 0.68
similar to other carnivorans, and Sus, which has an aspect ratio of 0.71. However, Gray
(1907) described the cochlea of Sus as intermediate between “flattened” and “sharp-
pointed”, but he described the cochleae of pinnipeds as “flattened”.
The cochlea of Cavia porcellus has the highest aspect ratio (1.29), and it is the
only species in this study in which the height of the cochlea is greater than the width.
Similar high-spired cochleae are observed within other caviomorph rodents, including
Hydrochoerus capybara (Gray, 1908), Dolichotis patagonum (http://digimorph.org),
Myocastor coypu (Solntseva, 2007), and Chinchilla laniger (personal observation). A
high-spired cochlea is likely a synapomorphy for caviomorph rodents, although the
405
cochlea of the North American porcupine, Erethizon dorsatum (which is nested well
within Caviomorpha; Huchon and Douzery, 2001; Huchon et al., 2007), is low spired in
Mus musculus and non-caviomorph rodents (Gray, 1907, 1908; personal observation).
Absence of the high-spired cochlea might retention of the ancestral state in Erethizon, but
it more likely is a reversal. A more thorough investigation of the bony labyrinths of
caviomorph and closely related non-caviomorph rodents is needed to fully explore the
issue.
The coiling of the cochlea is phylogenetically informative (suggested by Gray,
1907), and can be used to distinguish therian and non-therian mammals (Rowe, 1988),
where all extant therian cochleae are coiled to at least 360°. One fossil exception
Uchkudukodon nessovi from the Cretaceous of Uzbekistan, which has a cochlea that
completes less than 360° (McKenna et al., 2000). A single turn likely is plesiomorphic
for Eutheria (Meng and Fox, 1995; Chapter 4), and coiling beyond a single turn
developed more than once within placental lineages. The number of cochlear whorls was
not phylogenetically informative in the present study, however, and could not be used to
distinguish major clades within Placentalia. For example, Mus musculus and Pteropus
lyelli both possess a low degree of coiling (628° and 656° respectively), but other
members of their clades possess high degrees of coiling (e.g., 1457° in Cavia porcellus
and 1115° in Rhinolophus ferrumequinum).
The stapedial ratio is an index commonly used in phylogenetic analyses to explore
the relationships between Mesozoic therians (Wible, 1990; Rougier et al., 1998;
Archibald et al., 2001; Ladevèze, 2007). Use of the stapedial ratio in phylogenetic
analyses is perpetuated by the assumption that (with a few exceptions; Segall, 1970;
Ekdale et al., 2004) marsupials tend to have fenestra vestibuli that are more circular (with
a height to width ratio of the stapedial footplate below 1.8; dimensions of the fenestra
406
vestibuli are used as a proxy in the absence of the stapes) than placentals (based on
observations of Segall, 1970). The only marsupial considered here (Didelphis virginiana)
does, in fact, possess a fenestra vestibuli that falls below the 1.8 cut-off (ratio of 1.6).
Among the placentals examined here, however half of the taxa (15; note that no ratio was
calculated for Bathygenys) exhibit the ‘marsupial condition’ (below 1.8) of Segall (1970).
In fact, the ratio for Nycteris is 1.0, which is the observed condition among monotremes
(Segall, 1970). This result indicates that a thorough exploration of the stapedial ratio
across a broad range of marsupial and placental taxa is necessary before using the ratio in
phylogenetic studies.
A further example of the phylogenetic significance of the bony labyrinth of
mammals is the relative contribution of the vestibule to the entire labyrinth. The size of
the vestibular apparatus certainly is correlated to function (as discussed above), but a
reduced vestibule is also a synapomorphy for Cetacea, if only among cetartiodactyls. The
contribution of the vestibule of Rhinolophus is similar to that observed in cetaceans, but
the vestibular (or inversely, cochlear) contribution calculated for other bats fall within the
range of other mammals. Because a large cochlea is phylogenetically informative for
Cetacea, the phenomenon may also be informative within Chiroptera, upon which a
denser sampling of taxa might shed light.
The secondary common crus between the lateral and posterior semicircular canals
is an ancestral feature for Theria. All Cretaceous therians possess the secondary crus
(Meng and Fox, 1995; Chapter 4), although the structure is lost within most extant
placental groups. In fact, the only extant mammals considered in this study having the
secondary crus are the marsupial Didelphis virginiana, as well as the placentals Canis
familiaris and Orycteropus afer. The lateral semicircular canal opens directly into the
407
vestibule at its posterior end in the vast majority of the species examined here (20 out of
32), and absence of a secondary common crus is a synapomorphy for crown Placentalia.
A third state is entry of the lateral semicircular canal into the posterior ampulla
rather than the vestibule, but separate from the posterior canal (a secondary common crus
is not developed). Although the entry of the posterior limb of the lateral semicircular
canal does not express any major pattern with the taxonomic sampling employed by the
current study, potential for informativeness at lower levels is apparent. For example, the
lateral canal opens into the posterior ampulla in the cetaceans, but it opens into the
vestibule in their closest terrestrial relatives. Even so, entry of the lateral canal into the
posterior ampulla is reconstructed as a synapomorphy of Cetacea, as well as Carnivora. A
denser sampling at lower taxonomic levels within these groups, as well as Perissodactyla
and Chiroptera, may reveal phylogenetic patterns that are lost at the coarse resolution of
this study.
CONCLUSIONS
Variation is a naturally occurring phenomenon that is observable at all levels of
morphology, from anatomical variations of DNA molecules to gross variations between
whole organisms. The structure of the otic region is no exception. The present paper
documents the broad morphological diversity exhibited by the inner ear region of
placental mammals. Not surprisingly, many of the individual dimensions of the inner ear
correspond with each other, as well as with overall body size of the individual animal.
Great advancements have been made in our understanding of the physiological
role played by the inner ear labyrinth over the past 50 years, and this information has
been used to make functional and phylogenetic interpretations of fossil vertebrate taxa.
408
Certainly the morphology of the inner ear cavities has physiological correlates, and it
appears that the ratio of the average semicircular canal radius over body mass indeed can
distinguish aquatic from terrestrial mammals. Moreover, a generally reduced vestibular
system may indicate aquatic lifestyles among closely related taxa.
Phylogenetic patterns are preserved within the ear region. The reduced vestibular
apparatus of cetaceans may be rooted in physiology, but the condition, at least when
compared to their terrestrial relatives, is a synapomorphy for Cetacea, and such a
character state can be used to place fossil cetartiodactyls into a phylogenetic framework.
Additionally, the high aspect ratio of the cochlea of caviomorph rodents (excepting
Erethizon dorsatum) appears to be a synapomorphy for Caviomorpha.
The morphology of the bony labyrinth is phylogenetically informative at the
ordinal level, but the morphology of the inner ear does not support any superordinal
relationships. In order to fully understand the functional and evolutionary implications
within the structure of the bony labyrinth, both physiology and phylogeny must be
considered, as these two phenomena are not necessarily mutually exclusive. Future
detailed studies of the inner ear among closely related species will increase our
knowledge of the phylogenetic and functional implications of the inner ear, and foster the
application of bony labyrinth morphology to the biological interpretation of fossil
vertebrates.
409
Appendix 1
Specific CT scanning parameters employed during scanning and post-scanning image
processing. Definitions of parameters are as follows: FR, field of reconstruction is the
dimensions of an individual CT slice, expressed in millimeters; Pixel, interpixel spacing,
or vertical and horizontal dimensions of an individual pixel, expressed in millimeters, and
calculated as FR/Size; Size, number of pixels in a CT slice, either 512X512, or
1024X1024 pixels; Side, perusal from left or right side of skull; Slices, number of CT
slices through inner ear collected in the coronal (original) slice plane; Space, interslice
spacing, or distance between consecutive slices, expressed in millimeters.
Taxon and Specimen No.
Side
Slices
Space
FR
Pixel
Size
Cretaceous Eutheria
Zhelestid
URBAC 99-02
left
379
0.0128
11.5
0.0112
1024
URBAC 99-41
right
336
0.0127
15.0
0.0146
1024
URBAC 99-73
right
530
0.0128
10.5
0.0103
1024
URBAC 03-39
left
536
0.0158
14.9
0.0146
1024
URBAC 04-233
right
536
0.0158
14.9
0.0146
1024
ZIN C. 85511
right
397
0.0127
12.0
0.0117
1024
ZIN C. 85512
left
333
0.0128
11.5
0.0112
1024
Kulbeckia kulbeke
URBAC 98-113
left
126
0.0128
10.0
0.00977
1024
URBAC 00-16
left
446
0.0128
10
0.00977
1024
URBAC 02-56
right
377
0.0128
10.0
0.00977
1024
URBAC 04-36
right
387
0.0157
14.9
0.0146
1024
Ukhaatherium nessovi
PSS-MAE 110
left
59
0.0800
15.0
0.0293
512
Zalambdalestes lechei
PSS-MAE 108
left
66
0.113
24.5
0.0479
512
PSS-MAE 108
right
66
0.113
24.5
0.0479
512
PSS-MAE 130
left
127
0.111
28.0
0.0547
512
PSS-MAE 130
right
127
0.111
28.0
0.0547
512
410
Appendix 1. (Continued).
Taxon and Specimen No.
Side
Slices
Space
FR
Pixel
Size
Extant Placentalia
Dasypus novemcinctus
TMM M-152
left
494
0.0291
25.0
0.0244
1024
TMM M-1065
left
494
0.0291
25.0
0.0244
1024
TMM M-1880
left
494
0.0291
25.0
0.0244
1024
TMM M-1885
left
494
0.0291
25.0
0.0244
1024
TMM M-7417
left
97
0.0371
38.0
0.0371
1024
Tadarida brasiliensis
TMM M-3030
left
380
0.0104
9.9
0.00967
1024
TMM M-3030
right
315
0.0104
9.9
0.00967
1024
TMM M-3110
left
393
0.0104
9.9
0.00967
1024
TMM M-3110
right
388
0.0104
9.9
0.00967
1024
TMM M-6421
left
396
0.0104
9.9
0.00967
1024
TMM M-6653
left
450
0.0104
9.9
0.00967
1024
TMM M-6653
right
351
0.0104
9.9
0.00967
1024
Extant Marsupialia
Monodelphis domestica
TMM M-7536
left
81
0.0625
14.0
0.0273
512
TMM M-7539
left
76
0.0676
14.9
0.0291
512
TMM M-7542
left
96
0.0608
17.9
0.0350
512
TMM M-7545
left
81
0.0682
19.0
0.0371
512
TMM M-7595
left
116
0.0371
10.8
0.0211
512
TMM M-8261
left
69
0.0483
16.5
0.0161
1024
TMM M-8265
left
108
0.0338
13.0
0.0127
1024
TMM M-8266
left
119
0.0355
17.0
0.0166
1024
TMM M-8267
left
91
0.0480
22.0
0.0215
1024
TMM M-8268
left
109
0.0480
22.0
0.0215
1024
TMM M-8273
right
61
0.1170
55.7
0.0543
1024
411
Appendix 2
Extant placentals taken from published material for variation comparisons.
Carnivora
Felis catus (Curthoys et al., 1977a)
Primates
Alouatta seniculus (Spoor and Zonneveld, 1998)
Gorilla gorilla (Spoor and Zonneveld, 1998)
Homo sapiens (Blanks et al., 1975; Curthoys et al., 1977a; Muren et al., 1986;
Curthoys and Oman, 1987; Jeffery and Spoor, 2004)
Hylobates syndactylus (Spoor and Zonneveld, 1998)
Macaca fascicularis (Spoor and Zonneveld, 1998)
Macaca mulatta (Blanks et al., 1985)
Pan paniscus (Spoor and Zonneveld, 1998)
Pan troglodytes (Spoor and Zonneveld, 1998)
Pongo pygmaeus (Spoor and Zonneveld, 1998)
Saimiri sciureus (Blanks et al., 1985)
Rodentia
Cavia porcellus (Curthoys et al., 1975, 1977a; Curthoys and Oman, 1986)
Chinchilla laniger (Hullar and Williams, 2006)
Mus musculus (Calabrese and Hullar, 2006)
Rattus rattus (Curthoys and Oman, 1986).
412
References
Adkins, R. M., and R. L. Honeycutt. 1991. Molecular phylogeny of the superorder
Archonta. Proceedings of the National Academy of Sciences of the United States
of America 88:10317-10321.
Adkins, R. M., E. L. Gelke, D. Rowe, and R. L. Honeycutt. 2001. Molecular phylogeny
and divergence time estimates for major rodent groups: evidence from multiple
genes. Molecular Biology and Evolution 18:777-791.
Aitkin, L., S. Cochran, S. Frost, A. Martsi-McClintock, and B. Masterton. 1997. Features
of the auditory development of the short-tailed Brazilian opossum, Monodelphis
domestica: evoked responses, neonatal vocalizations and synapses in the inferior
colliculus. Hearing Research 113:69-75.
Alexander, R. M. 2002. Stability and maneuverability of terrestrial vertebrates.
Integrative and Comparative Biology 42:158-164.
Alonso, P. D., A. C. Milner, R. A. Ketcham, M. J. Cookson, and T. B. Rowe. 2004. The
avian nature of the brain and inner ear of Archaeopteryx. Nature 430:666-669.
Amrine-Madsen, H., K. –P. Koepfli, R. K. Wayne, and M. S. Springer. 2003. A new
phylogenetic marker, apolipoprotein B, provides compelling evidence for
eutherian relationships. Molecular Phylogenetics and Evolution 28:225-240.
Archibald, J. D. 1979. Oldest known eutherian stapes and a marsupial petrosal bone from
the Late Cretaceous of North America. Nature 281:669-670.
Archibald, J. D. 1996. Fossil evidence for a Late Cretaceous origin of “hoofed”
mammals. Science 272:1150-1153.
Archibald, J. D., and A. O. Averianov. 2003. The Late Cretaceous placental mammal
Kulbeckia. Journal of Vertebrate Paleontology 23:404-419.
Archibald, J. D., and A. O. Averianov. 2005. Mammalian succession in the Cretaceous of
the Kyzylkum Desert. Journal of Mammalian Evolution 12:9-22.
Archibald, J. D., and A. O. Averianov. 2006. Late Cretaceous asioryctitherian eutherian
mammals from Uzbekistan and phylogenetic analysis of Asioryctitheria. Acta
Palaeontologica Polonica 51:351-376.
Archibald, J. D., and A. O. Averianov. 2007. Zhelestids: stem eutherians or basal
laurasiatherians, but no evidence for placental orders in the Cretaceous. Journal of
Vertebrate Paleontology 27(3, Supplement):41A.
Archibald, J. D., A. O. Averianov, and E. G. Ekdale. 2001. Late Cretaceous relatives of
rabbits, rodents, and other extant eutherian mammals. Nature 414:62-65.
Arnason, U., J. A. Adegoke, K. Bodin, E. W. Born, Y. B. Esa, A. Gullberg, M. Nilsson,
R. V. Short, X. Xu, and A. Janke. 2002. Mammalian mitogenomic relationships
413
and the root of the eutherian tree. Proceedings of the National Academy of
Sciences 99:8151-8156.
Asher, R. J. 2001. Cranial anatomy in tenrecid insectivorans: character evolution across
competing phylogenies. American Museum Novitates 3352:1-54.
Asher, R. J. 2005. Insectivoran-grade placentals; pp. 50-70 in K. D. Rose and J. D.
Archibald (eds), The rise of placental mammals: origins and relationships of the
major extant clades. The Johns Hopkins University Press, Baltimore.
Asher, R. J. 2007. A web-database of mammalian morphology and a reanalysis of
placental phylogeny. BMC Evolutionary Biology 7(108):1-10.
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 10:131-194.
Asher, R. J., J. Meng, J. R. Wible, M. C. McKenna, G. W. Rougier, D. Dashzeveg, and
M. J. Novacek. 2005. Stem Lagomorpha and the antiquity of Glires. Science
307:1091-1094.
Averianov, A. O., and J. D. Archibald. 2005. Mammals from the mid-Cretaceous
Khodzhakul Formation, Kyzylkum Desert, Uzbekistan. Cretaceous Research
26:593-608.
Beard, K. C. 1993. Phylogenetic systematics of the Primatomorpha, with special
reference to Dermoptera; pp 129-150 in F. S. Szalay, M. J. Novacek, and M. C.
McKenna (eds.), Mammal Phylogeny: Placentals. Springer-Verlag, New York.
Beck, R. M. D., H. Godthelp, V. Weisbecker, M. Archer, and S. J. Hand. 2008.
Australia’s oldest marsupial fossils and their biogeographical implications. PLoS
One 3(3):e1858.
Bever, G. S. 2006. Studies on post-natal variation and variability in the skeleton and its
paleontological implications. Unpublished PhD Dissertation, The University of
Texas at Austin, 719 p.
Bever, G. S., C. J. Bell, and J. A. Maisano. 2005. The ossified braincase and cephalic
osteoderms of Shinisaurus crocodilurus (Squamata, Shinisauridae).
Palaeontologia Electronica 8(4A):1-36.
Bininda-Emonds, O. R. P., J. L. Gittelman, and A. Purvis. 1999. Building large trees by
combining phylogenetic information: a complete phylogeny of the extant
Carnivora (Mammalia). Biological Reviews 74:143-175.
Bininda-Emonds, O. R. P., M. Cardilla, K. E. Jones, D. E. Ross, R. M. D. Beck, R.
Grenyer, S. A. Price, R. A. Voss, J. L. Gittleman, and A. Purvis. 2007. The
delayed rise of present-day mammals. Nature 446:507-512.
414
Blair, P. 1710-1712a. Osteographia elephantina: or, a full and exact description of all the
bones of an elephant, which died near Dundee, April the 27
th
, 1706. With their
several dimensions. Communicated in a letter to Dr. Hans Sloane, R. S. Secr. By
Mr Patrick Blair, &c. Philosophical Transactions (1683-1775) 27:53-116.
Blair, P. 1710-1712b. A continuation of the osteographia elephantina: or, a description of
the bones of an elephant, which died near Dundee, April the 27
th
, 1706. By Mr.
Patrick Blair. Philosophical Transactions (1683-1775) 27:117-168.
Blair, P. 1717-1719. A description of the organ of hearing in the elephant, with the
figures and situation of the ossicles, labyrinth and cochlea in the ear of that large
animal. Communicated to the Royal Society by Dr. Patrick Blair, R. S. S.
Philosophical Transactions (1683-1775) 30:885-898.
Blanks, R. H. I., I. S. Curthoys, and C. H. Markham. 1972. Planar relationships of
semicircular canals in the cat. American Journal of Physiology 223:55-62.
Blanks, R. H. I., I. S. Curthoys, and C. H. Markham. 1975. Planar relationships of the
semicircular canals in man. Acta Oto-Laryngolica 80:185-196.
Blanks, R. H. I., I. S. Curthoys, M. L. Bennett, and C. H. Markham. 1985. Planar
relationships of the semicircular canals in rhesus and squirrel monkeys. Brain
Research 340:315-324.
Boisvert, C. A. 2005. The pelvic fin and girdle of Panderichthys and the origin of
tetrapod locomotion. Nature 438:1145-1147.
Boulter, M. C., D. Gee, and H. C. Fisher. 1998. Angiosperm radiations at the
Cenomanian/Turonian and Cretaceous/Tertiary boundaries. Cretaceous Research
19:107-112.
Boyden A, Gemeroy D. 1950. The relative position of the Cetacea among the orders of
Mammalia as indicated by precipitin tests. Zoologica 35:145-151.
Boyer, D. M. and J. A. Georgi. 2007. Cranial morphology of a pantolestid eutherian
mammal from the Eocene Bridger Formation, Wyoming, USA: implications for
relationships and habitat. Journal of Mammalian Evolution 14:239-380.
Brinkman, D. 1988. Size-independent criteria for estimating relative age in Ophiacodon
and Dimetrodon (Reptilia, Pelycosauria) from the Admiral and Lower Belle
Plains formations of west-central Texas. Journal of Vertebrate Paleontology
8:172-180.
Brochu, C. A. 1996. Closure of neurocentral sutures during crocodilian ontogeny:
implications for maturity assessment in fossil archosaurs. Journal of Vertebrate
Paleontology 16:49-62.
Bugge, J. 1978. The cephalic arterial system in carnivores, with special reference to the
systematic classification. Acta Anatomica 101:45-61.
415
Butler, P. M. 1972. The problem of insectivore classification; pp. 253-265 in K. A.
Joysey, and T. S. Kemp (eds), Studies in vertebrate evolution. Winchester Press,
New York, New York.
Caix, M., and G. Outrequin. 1979. Variability of the bony semicircular canals. Anatomia
Clinica 1:259-265.
Calabrese, D. R., and T. E. Hullar. 2006. Planar relationships of the semicircular canals in
two strains of mice. Journal for the Association for Research in Otolaryngology
7:151-159.
Campbell CBG. 1966. The relationships of the tree shrews: the evidence of the nervous
system. Evolution 20:276-281.
Cao, Y., N. Okada, and M. Hasegawa. 1997. Phylogenetic position of guinea pigs
revisited. Molecular Biology and Evolution 14:461-464.
Cao Y., J. Adachi , T. Yano, and M. Hasegawa. 1994. Phylogenetic place of guinea pigs:
no support of the rodent-polyphyly hypothesis from maximum-likelihood
analyses of multiple protein sequences. Molecular Biology and Evolution 11:593-
604.
Cao Y., K. S. Kim, J. H. Ha, and M. Hasegawa. 1999. Model dependence of the
phylogenetic inference: relationship among carnivores, perissodactyls and
cetartiodactyls as inferred from mitochondrial genome sequences. Genes and
Genetic Systems 74:211-217.
Carlsson, A. 1922. Über die Tupaiidae und ihre beziehungen zu den Insectivora und den
Prosimiae. Acta Zoologica 3:227-270.
Carrasco, M. A. 1998. Variation and its implication in a population of Cupidinimus
(Heteromyidae) from Hepburn’s Mesa, Montana. Journal of Vertebrate
Paleontology 18:391-402.
Chester, S., E. Sargis, E. Szalay, J. D. Archibald, and A. Averianov. 2007. Functional
analysis of mammalian humeri from the Late Cretaceous of Uzbekistan. Journal
of Vertebrate Paleontology 27(3, Supplement):58A.
Chester, S., E. Sargis, F. Szalay, J. D. Archibald, and A. O. Averianov. 2008. Therian
femora from the Late Cretaceous of Uzbekistan. Journal of Vertebrate
Paleontology 28(3, Supplement):63A.
Christiansen, P. 2004. Body size in proboscideans, with notes on elephant metabolism.
Zoological Journal of the Linnean Society 140:523-549.
Cifelii, R. L. 1982. The petrosal structure of Hyopsodus with respect to that of some other
ungulates, and its phylogenetic implications. Journal of Paleontology 56:795-805.
416
Cifelli, R. L. 1990. Cretaceous mammals of southern Utah. IV. Eutherian mammals from
the Wahweap (Aquilan) and Kaiparowits (Judithian) formations. Journal of
Vertebrate Paleontology 10:346-360.
Clack, J. A. 2002. Patterns and processes in the early evolution of the tetrapod ear.
Journal of Neurobiology 53:251-264.
Clark, C. T. and K. K. Smith. 1993. Cranial osteogenesis in Monodelphis domestica
(Didelphidae) and Macropus eugenii (Macropodidae). Journal of Morphology
215:119-149.
Clarke, A. H. 2005. On the vestibular labyrinth of Brachiosaurus brancai. Journal of
Vestibular Research 15:65-71.
Clarke, J. A., C. P. Tambussi, J. I. Noriega, G. M. Erickson, and R. A. Ketcham. 2005.
Definitive fossil evidence for the extant avian radiation in the Cretaceous. Nature
433:305-308.
Clemens, W. A. 1968. Origin and early evolution of marsupials. Evolution 22:1-18.
Colbert, M. W., and T. Rowe. 2008. Ontogenetic sequence analysis: using parsimony to
characterize developmental sequences and sequence polymorphism. Journal of
Experimental Zoology (Molecular and Developmental Evolution) 310B:398-416.
Coleman, M. N., and M. W. Colbert. 2007. Technical note: CT thresholding protocols for
taking measurements on three-dimensional models. American Journal of Physical
Anthropology 133:723-725.
Cooper, A., and D. Penny. 1997. Mass survival of birds across the Cretaceous-Tertiary
boundary: molecular evidence. Science 275:1109-1113.
Cope, D. A., and M. G. Lacy. 1995. Comparative application of the coefficient of
variation and range-based statistics for assessing the taxonomic composition of
fossil samples. Journal of Human Evolution 29:549-576.
Court N. 1990. Periotic anatomy of Arsinotherium (Mammalia, Embrithopoda) and its
phylogenetic implications. Journal of Vertebrate Paleontology 10:170-182.
Court N. 1992a. The skull of Arsinotherium (Mammalia, Embrithopoda) and the higher
order interrelationships of ungulates. Palaeovertebrata 22:1-43.
Court, N. 1992b. Cochlea anatomy of Numidotherium koholense: auditory acuity in the
oldest known proboscidean. Lethaia 25:211-215.
Court, N. 1994. The periotic of Moeritherium (Mammalia, Proboscidea): homology or
homoplasy in the ear region of Tethytheria McKenna, 1975?. Zoological Journal
of the Linnean Society 112:13-28.
Court, N. and J. –J. Jaeger. 1991. Anatomy of the periotic bone in the Eocene
probsoidean Numidotherium koholense: an example of parallel evolution in the
inner ear of tethytheres. Comptes Rendus de l’Académie des Sciences. Série II,
417
Mécanique, Physique, Chimie, Sciences de l’Univers, Sciences de la Terre,
312:559-565.
Cox, P. G., and N. Jeffery. 2008. Geometry of the semicircular canals and extraocular
muscles in rodents, lagomorphs, felids and modern humans. Journal of Anatomy
213:583-596.
Cracraft, J. 1986. The origin and early diversification of birds. Paleobiology 12:383-399.
Curthoys, I. A., and C. M. Oman. 1986. Dimensions of the horizontal semicircular duct,
ampulla and utricle in rat and guinea pig. Acta Oto-laryngolica 101:1-10.
Curthoys, I. A., and Oman. 1987. Dimensions of the horizontal semicircular duct,
ampulla and utricle in the human. Acta Oto-laryngolica 103:254-261.
Curthoys, I. S., R. H. I. Blanks, and C. H. Markham. 1977a. Semicircular canal radii of
curvature (R) in cat, guinea pig and man. Journal of Morphology 151:1-16.
Curthoys, I. S., C. H. Markham, and E. J. Curthoys. 1977b. Semicircular duct and
ampulla dimensions in cat, guinea pig and man. Journal of Morphology 151:17-
34.
Curthoys, I. S., E. J. Curthoys, R. H. I. Blanks, and C. H. Markham. 1975. The
orientation of the semicircular canals in the guinea pig. Acta Otolaryngolica
80:197-205.
de Beer, G. R. 1947. How animals hold their heads. Proceedings of the Linnean Society
of London 159:125-139.
de Jong, W. W., A. Zweers, and M. Goodman. 1981. Relationship of aardvark to
elephants, hyraxes and sea cows from #-crystallin sequences. Nature 292:538-
540.
de Jong, W. W., M. A. M. van Dijk, C. Poux, G. Kappé, T. van Rheede, and O. Madsen.
2003. Indels in protein-coding sequences of Euarchontoglires constrain the
rooting of the eutherian tree. Molecular Phylogenetics and Evolution 28:328-340.
D’Erchia, A. M., C. Gisso, G. Pesole, C. Saccone, and U. Arnason. 1996. The guinea-pig
is not a rodent. Nature 381:597-600.
Dercum, F. 1879. On the morphology of the semicircular canals. American Naturalist
13:366-374.
Dimopoulos, P., and C. Muren. 1990. Anatomic variations of the cochlea and relations to
other temporal bone structures. Acta Radiologica 31:439-444.
Douzery, E. J. P., and D. Huchon. 2004. Rabbits, if anything, are likely Glires. Molecular
Phylogenetics and Evolution 33:922-935.
Eales, N. B. 1926. The anatomy of the head of a foetal African elephant, Elephas
africanus (Loxodonta africana). Transactions of the Royal Society of Edinburgh
54:491-551.
418
Easteal, S. 1990. The pattern of mammalian evolution and the relative rate of molecular
evolution. Genetics 124:165-173.
Eizirik, E., W. J. Murphy, and S. J. O’Brien. 2001. Molecular dating and biogeography of
the early placental mammal radiation. Journal of Heredity 92:212-219.
Ekdale, E. G. 2005. Ontogeny of the inner ear of mammals: implications for the
phylogenetic assessment of fossils. Journal of Vertebrate Paleontology (3,
Supplement) 25:53A.
Ekdale, E. G., J. D. Archibald, and A. O. Averianov. 2004. Petrosal bones of placental
mammals from the Late Cretaceous of Uzbekistan. Acta Palaeontological
Polonica 49:161-176.
Emerson, G. L., C. W. Kilpatrick, B. E. McNiff, J. Ottenwalder, M. W. Allard. 1999.
Phylogenetic relationships of the order Insectivora based on complete 12S rRNA
sequences from mitochondria. Cladistics 15:221-230.
Eugenín, J. and J. G. Nicholls. 1997. Chemosensory and cholinergic stimulation of fictive
respiration in isolated CNS of neonatal opossum. Journal of Physiology 501:425-
437.
Evans, G. L. 1961. The Friesenhahn Cave. Bulletin of the Texas Memorial Museum 2:7-
22.
Evans, H. E. 1993. Miller’s Anatomy of the Dog (third edition). Saunders, Philadelphia,
1113 p.
Fadem, B. H., G. L. Trupin, E. Maliniak, J. L. VandeBerg, and V. Hayssen. 1982. Care
and breeding of the gray, short-tailed opossum (Monodelphis domestica).
Laboratory Animal Science 32:405-409.
Fischer, M. S. 1990. Un trait unique de l’orielle des elephants et des siréniens
(Mammalia): un paradoxe phylogénétigue. Comptes Rendus de l’Académie des
Sciences. Série II, Sciences de la Vie 311:157-162.
Fischer, M. S. and P. Tassy. 1993. The interrelation between Proboscidea, Sirenia,
Hyracoidea, and Mesaxonia: the morphological evidence; pp. 217-234 in F. S.
Szalay, M. J. Novacek, and M. C. McKenna (eds.), Mammal Phylogeny:
Placentals. Springer-Verlag, New York.
Fleischer, G. 1976. Hearing in extinct cetaceans as determined by cochlear structure.
Journal of Paleontology 50:133-152.
Flower, W. H. 1883. On the arrangement of the orders and families of existing
Mammalia. Proceedings of the Zoological Society of London 1883:178-186.
Flynn, J. J., and G. D. Wesley-Hunt. 2005. Carnivora; pp 175-198 in K. D. Rose, and J.
D. Archibald (eds.), The rise of placental mammals: origins and relationships of
the major extant clades. The Johns Hopkins University Press, Baltimore.
419
Fox, R. C. 1989. The Wounded Knee local fauna and mammalian evolution near the
Cretaceous-Tertiary boundary, Saskatchewan, Canada. Palaeontographica
Abteilung A 208:11-59.
Frye, M. S., and S. B. Hedges. 1995. Monophyly of the order Rodentia inferred from
mitochondrial DNA sequences of the genes for 12S rRNA, 16S rRNA, and tRNA-
valine. Molecular Biology and Evolution 12:168-176.
Galvão, P. E. 1947. Heat production in relation to body weight and body surface.
Inapplicability of the surface law on dogs of the tropical zone. American Journal
of Physiology 148:478-489.
Gatesy, J., C. Hayashi, M. A. Cronin, and P. Arctander. 1996. Evidence from milk casein
genes that cetaceans are close relatives of hippopotamid artiodactyls. Molecular
Biology Evolution 13:954-963.
Gatesy, J., M. Milinkovitch, V. Waddell, and M. Stanhope. 1999. Stability of cladistic
relationships between Cetacea and higher-level artiodactyl taxa. Systematic
Biology 48:6-20.
Gaudin, T. J., and A. A. Biewener. 1992. The functional morphology of xenarthrous
vertebrae in the armadillo Dasypus novemcinctus (Mammalia, Xenarthra). Journal
of Morphology 214:63-81.
Geisler, J.H. 2001. New morphological evidence for the phylogeny of Artiodactyla,
Cetacea, and Mesonychidae. American Museum Novitates 3344:1-53.
Geisler, J. H. and Z. Luo. 1996. The petrosal and inner ear of Herpetocetus sp.
(Mammalia: Cetacea) and their implications for the phylogeny and hearing of
archaic mysticetes. Journal of Paleontology 70:1045-1066.
Geisler, J. H., and Z. Luo. 1998. Relationships of Cetacea to terrestrial ungulates and the
evolution of cranial vasculature in Cete; pp. 163-212 in J. G. M Thewissen (ed.),
The Emergence of Whales. Plenum Press, New York.
Geisler, J. H., and J. M. Theodor. 2009. Hippopotamus and whale phylogeny. Nature
458:E1-E4.
Geisler, J. H., M. D. Uhen. 2003. Morphological support for a close relationship between
hippos and whales. Journal of Vertebrate Paleontology 23:991-996.
George, C. O., E. L. Lundelius, Jr., L. Meissner, R. S. Toomey, III, and R. W. Graham.
2007. Late Quaternary cave sites of central Texas: field trip guidebook. Prepared
for the Society of Vertebrate Paleontology, Austin,TX, Oct. 15-16, 67
th
Annual
Meeting, 48 p.
Georgi, J. A., and J. S. Sipla. 2008. Comparative and functional anatomy of balance in
aquatic reptiles and birds; pp. 233-256 in J. G. M. Thewissen and S. Numella
(eds.), Sensory Evolution on the Threshold: Adaptations in Secondarily Aquatic
Vertebrates. University of California Press, Berkeley, California.
420
Gheerbrant, E., and H. Astibia. 1994. Un nouveau mammifère du Maastrichtien de Laño
(Pays Basque espagnol). Comptes Rendus de l’Académie des Sciences, Série II ,
Sciences de la Terre et des Planètes 318:1125-1131.
Gheerbrant, E., D. P. Domning, and P. Tassy. 2005. Paenungulata (Sirenia, Proboscidea,
Hyracoidea, and Relatives), pp. 84-105 in K. D. Rose, and J. D. Archibald (eds.),
The Rise of Placental Mammals: Origins and Relationships of the Major Extant
Clades. The Johns Hopkins University Press, Baltimore.
Gidley, J. W. 1912. The lagomorphs an independent order. Science 36:285-286.
Gingerich, P. D. 2005. Cetacea; pp. 234-252 in K. D. Rose, and J. D. Archibald JD (eds.),
The Rise of Placental Mammals: Origins and Relationships of the Major Extant
Clades. The Johns Hopkins University Press, Baltimore.
Gingerich, P. D., M. ul Haq, I. S. Zalmout, I. H. Khan, and M. S. Malkani. 2001. Origin
of whales from early artiodactyls: hands and feet of Eocene Protocetidae from
Pakistan. Science 293:2239-2242.
Goodman, M., C. A. Porter, J. Czelusniak, S. L. Page, H. Schneider, J. Shoshani, G.
Gunnell, and C. P. Groves. 1998. Toward a phylogenetic classification of
Primates based on DNA evidence complemented by fossil evidence. Molecular
Phylogenetics and Evolution 9:585-598.
Gosman, J. H., and R. A. Ketcham. 2009. Patterns in ontogeny of human trabecular bone
from SunWatch Village in the prehistoric Ohio Valley: general features of
microarchitectural change. American Journal of Physical Anthropology 138:318-
332.
Gosselin-Ildari, A. D. 2006. Functional morphology of the bony labyrinth in Primates.
Unpublished Undergraduate Honors Thesis, The University of Texas at Austin, 52
p.
Gosselin-Ildari, A. D., and E. C. Kirk. 2007. Functional morphology of the primate
cochlea. American Journal of Physical Anthropology 132(S44):118.
Gould, E. 1965. Evidence for echolocation in the Tenrecidae of Madagascar. Proceedings
of the American Philosophical Society 109:352-360.
Graham, R. W. 1976. Pleistocene and Holocene mammals, taphonomy, and paleoecology
of the Friesenhahn Cave local fauna, Bexar County, Texas. Unpublished Ph.D.
Dissertation, The University of Texas at Austin, 233 p.
Graur, D., and D. G. Higgins. 1994. Molecular evidence for the inclusion of cetaceans
within the order Artiodactyla. Molecular Biology and Evolution 11:357-364.
Graur, D., W. A. Hide, and W. –H. Li. 1991. Is the guinea-pig a rodent? Nature 351:649-
652.
421
Graur, D., W. A. Hide, A. Zharkikh, and W. –H. Li. 1992. The biochemical phylogeny of
guinea-pigs and gundis, and the paraphyly of the order Rodentia. Comparative
Biochemistry and Physiology 101B:495-498.
Gray, A. A. 1903. On a method of preparing the membranous labyrinth. Journal of
Anatomy and Physiology 37:379-381.
Gray, A. A. 1905. A demonstration of specimens prepared by the author’s method to
show the membranous labyrinth in man and lower animals. Transactions of the
Otological Society of the United Kingdom 6:9-13.
Gray, A. A. 1906. Observations on the labyrinth of certain animals. Proceedings of the
Royal Society of London. Series B, Containing Papers of a Biological Character
78:284-296.
Gray, A. A. 1907. The labyrinth of animals: including mammals, birds, reptiles and
amphibians. Volume 1. J. and A. Churchill, London, 98 pp.
Gray, A. A. 1908. The labyrinth of animals: including mammals, birds, reptiles and
amphibians. Volume 2. J. and A. Churchill, London, 252 pp.
Graybeal, A., J. J. Rosowski, D. R. Ketten, and A. W. Crompton. 1989. Inner-ear
structure in Morganucodon, an early Jurassic mammal. Zoological Journal of the
Linnean Society 96:107-117.
Gregory, W. K. 1910. The orders of mammals. Bulletin of the American Museum of
Natural History 27:1-524.
Grenyer, R., and A. Purvis. 2003. A composite species-level phylogeny of the
‘Insectivora’ (Mammalia: Order Lipotyphla Haeckel, 1866). Journal of Zoology,
London 260:245-257.
Guild, S. R. 1921. A graphic reconstruction method for the study of the organ of Corti.
Anatomical Record 22:141-157.
Halpern, M., Y. Daniels, and I. Zuri. 2005. The role of the vomeronasal systemin food
preferences of the gray short-tailed opossum, Monodelphis domestica. Nutrition
and Metabolism 2(6):1-6.
Hammer, Ø., and D. Harper. 2006. Paleontological Data Analysis. Blackwell Publishing,
Malden, Massachussets, 351 pp.
Harvati, K. A. and T. D. Weaver. 2006. Human cranial anatomy and the differential
preservation of population history and climate signatures. Anatomical Record
288A:1225-1233.
Hawkins, J. E. 2004. Sketches of otohistory. Part 1: otoprehistory: how it all began.
Audiology and Neuro-otology 9:66-71.
Hedges, S. B., P. H. Parker, C. G. Sibley, and S. Kumar. 1996. Continental breakup and
the ordinal diversification of birds and mammals. Nature 381:226-229.
422
Heusner, A. A. 1991. Body mass, maintenance and basal metabolism in dogs. Journal of
Nutrition 121:S8-S17.
Holroyd, P. A., J. C. Mussell. 2005. Macroscelidea and Tubulidentata; pp. 71-83 in K. D.
Rose, J. D. Archibald (eds.), The Rise of Placental Mammals: Origins and
Relationships of the Major Extant Clades. The Johns Hopkins University Press,
Baltimore.
Honeycutt, R. L., and R. M. Adkins. 1993. Higher level systematics of eutherian
mammals: an assessment of molecular characters and phylogenetic hypotheses.
Annual Reviews in Ecology and Systematics 24:279-305.
Hooker, J. J. 2005. Perissodactyla; pp. 199-214 in K. D. Rose, and J. D. Archibald (eds.),
The Rise of Placental Mammals: Origins and Relationships of the Major Extant
Clades. The Johns Hopkins University Press, Baltimore.
Hoyte, D. A. N. 1961. The postnatal growth of the ear capsule in the rabbit. American
Journal of Anatomy 108:1-16.
Hublin, J. –J, F. Spoor, M. Braun, F. Zonneveld, and S. Condemi. 1996. A late
Neanderthal associated with Upper Paleaeolithic artifacts. Nature 381:224-226.
Huchon, D., and J. P. Douzery. 2001. From the Old World to the New World: a
molecular chronicle of the phylogeny and biogeography of hystricognath rodents.
Molecular Phylogenetics and Evolution 20:238-251.
Huchon, D., F. M. Catzeflis, and E. J. P. Douzery. 1999. Molecular evolution of the
nuclear von Wollebrand factor gene in mammals and the phylogeny of rodents.
Molecular Biology and Evolution 16:577-589.
Huchon, D., P. Chevret, U. Jordan, C. W. Kilpatrick, V. Ranwez, P. D. Jenkins, J.
Brosius, and J. Schmitz. 2007. Multiple molecular evidences for a living
mammalian fossil. Proceedings of the National Academy of Sciences of the
United States of America 104:7495-7499.
Huchon, D., O. Madsen, M. J. J. B. Sibbald, K. Ament, M. J. Stanhope, F. Catzeflis, W.
W. de Jong, and E. J. P. Douzery. 2002. Rodent phylogeny and a timescale for the
evolution of Glires: evidence from an extensive taxon sampling using three
nuclear genes. Molecular Biology and Evolution 19:1053-1065.
Hullar, T. E. 2006. Semicircular canal geometry, afferent sensitivity, and animal
behavior. Anatomical Record 288A:466-472.
Hullar, T. E. and C. D. Williams. 2006. Geometry of the semicircular canals of the
chinchilla (Chinchilla laniger). Hearing Research 213:17-24.
Hunt, R. M., Jr. 1987. Evolution of the aeluroid Carnivora: significance of auditory
structure in the nimravid cat Dinictis. American Museum Novitates 2886:1-74.
423
Hunt, R. M., Jr. 1989. Evolution of the aeluroid Carnivora: significance of the ventral
promontorial process of the petrosal, and the origin of basicranial patterns in
living families. American Museum Novitates 2930: 1-32.
Hyrtl, J. 1845. Verleichende-anatomische Untersuchungen über das innere Gehörorgan
des menschen und der Säugethiere. Prague, 139 pp.
Igarashi, M. 1967. Dimensional study of the vestibular apparatus. Laryngoscope 77:1806-
1817.
Jane"ka, J. E., W. Miller, T. H. Pringle, F. Wiens, A. Zitzmann, K. M. Helgen, M. S.
Springer, W. J. Murphy. 2007. Molecular and genomic data identify the closest
living relative of Primates. Science 318:792-794.
Jeffery, N., and F. Spoor. 2004. Prenatal growth and development of the modern human
labyrinth. Journal of Anatomy 204:71-92.
Jeffery, N. and F. Spoor. 2006. The primate subarcuate fossa and its relationship to the
semicircular canals part I: prenatal growth. Journal of Human Evolution, 51:537-
549.
Jones, G. M. and K. E. Spells. 1963. A theoretical and comparative study of the
functional dependence of the semicircular canal upon its physical dimensions.
Proceedings of the Royal Society of London B 157:403-419.
Kaiser, H. E. 1974. Morphology of the Sirenia: a macroscopic and X-ray atlas of the
osteology of recent species. S. Karger, Basel, Switzerland, 75 pp.
Kay, R. F., C. Ross, and B. A. Williams. 1997. Anthropoid origins. Science 275:797-804.
Ketten, D. R. 1997. Structure and function in whale ears. Bioacoustics 8:103-135.
Kielan-Jaworowska, Z. 1984. Evolution of the therian mammals in the Late Cretaceous
of Asia. Part VI. Endocranial casts of eutherian mammals. Palaeontologica
Polonica 46:157-171.
Kielan-Jaworowska, Z., R. Cifelli, and Z. –X. Luo. 2004. Mammals from the Age of
Dinosaurs. Columbia University Press, New York, 630 pp.
Kirk, E. C., and A. D. Gosselin-Ildari. 2009. Cochlear labyrinth volume and hearing
abilities in Primates. Anatomical Record 292:765-776.
Kleineidam, R. G., G. Pesole, H. L. Breukelman, J. J. Beintema, and R. A. Kastelein.
1999. Inclusion of cetaceans within the order Artiodactyla based on phylogenetic
analysis of pancreatic ribonuclease genes. Journal of Molecular Evolution 48:360-
368.
Kumar, S., and S. B. Hedges. 1998. A molecular timescale for vertebrate evolution.
Nature 392:917-920.
Kusewitt, D. F., T. A. Sherburn, K. B. Miska, G. B. Tafoya, J. M. Gale, and R. D. Miller.
1999. The p53 tumor suppressor gene of the marsupial Monodelphis domestica:
424
cloning of exons 4-11 and mutations in exons 5-8 in ultraviolet radiation-induced
corneal sarcomas. Carcinogenesis 20:963-968.
Ladevèze, S. 2004. Metatherian petrosals from the late Paleocene of Itaboraí, Brazil, and
their phylogenetic implications. Journal of Vertebrate Paleontology 24:202-213.
Ladevèze, S. 2007. Petrosal bones of metatherian mammals form the Late Palaeocene of
Itaboraí (Brazil), and a cladistic analysis of petrosal features in metatherians.
Zoological Journal of the Linnean Society, 150:85-115.
Larsell, O., E. McCrady, Jr, and A. A. Zimmermann. 1935. Morphological and functional
development of the membranous labyrinth in the opossum. Journal of
Comparative Neurology 63:95-118.
Lambert, W. D. and J. Shoshani. 1998. Proboscidea, pp. 606-621 in C. M. Janis, K. M.
Scott, and L. L. Jacobs (eds.), Evolution of Tertiary Mammals of North America,
Volume 1: Terrestrial Carnivores, Ungulates, and Ungulatelike Mammals.
Cambridge University Press, Cambridge.
Lavergne, A., E. Douzery, T. Stichler, F. M. Catzeflis, and M. S. Springer. 1996.
Interordinal mammalian relationships: evidence for paenungulate monophyly is
provided by complete mitochondrial 12S rRNA sequences. Molecular
Phylogenetics and Evolution 6:245-258.
Lawrence, D. R. 1968. Taphonomy and information losses in fossil communities.
Geological Society of America Bulletin 79:1315-1330.
Laws, R. M. 1966. Age criteria for the African elephant, Loxodonta a. africana. East
African Wildlife Journal 4:1-37.
Lee, F. S. 1898. Functions of the ear and the lateral line in fishes. American Journal of
Physiology 1:128-144.
Li W-H, Hide WA, Graur D. 1992. Origin of rodents and guinea-pigs. Nature 359:277-
278.
Lillegraven, J. A. 1976. A new genus of therian mammal from the Late Cretaceous “El
Gallo Formation,” Baja California, Mexico. Journal of Paleontology 50:437-443.
Lin, Y. –H., P. J. Waddell, and D. Penny. 2002. Pika and vole mitochondrial genomes
increase support for both rodent monophyly and glires. Gene 294:119-129.
Linnaeus, C. 1748. Systema Naturae sistens Regna tria Naturae, in Classes et Ordines,
Genera et Species. Sixth Edition. Kiesewetteri, Stockholm, 224 pp.
Linnaeus C. 1758. Systema Naturae per Regna tria Naturae, secundum Classes, Ordines,
Genera, Species, cum Characteribus, Differentiis, Synonymis, Locis. Volume 1.
Tenth Edition. Laurentius Salvius, Stockholm, 376 pp.
425
Liu, F. –G., M. M. Miyamoto, N. P. Freire, P. Q. Ong, M. R. Tennant, T. S. Young, and
K. F. Gugel. 2001. Molecular and morphological supertrees for eutherian
(placental) mammals. Science 291:1786-1789.
Loughlin, T. R., M. A. Perez, R. L. Merrick. 1987. Eumetopias jubatus. Mammalian
Species 283:1-7.
Luckett, W. P., and J. –L. Hartenberger. 1993. Monophyly or polyphyly of the order
Rodentia: possible conflict between morphological and molecular interpretations.
Journal of Mammalian Evolution 1:127-147.
Lundelius, E. L., Jr. and M. B. Collins. 1999. Speleological contexts of micromammal
assemblages from caves and rockshelters in the Texas Hill Country: retrospects
and prospects. International Union of Quaternary Research, XV International
Congress Book of Abstracts :112.
Luo, Z., and E. R. Eastman. 1995. Petrosal and inner ear of a squalodontid whale:
implications for the evolution of hearing in odontocetes. Journal of Vertebrate
Paleontology 15:431-442.
Luo, Z., and P. D. Gingerich. 1999. Terrestrial Mesonychia to aquatic Cetacea:
transformation of the basicranium and evolution of hearing in whales. University
of Michigan Papers in Paleontology 31:1-98.
Luo, Z., and D. R. Ketten. 1991. CT scanning and computerized reconstructions of the
inner ear of multituberculate mammals. Journal of Vertebrate Paleontology
11:220-228.
Luo, Z., and K. Marsh. 1996. Petrosal (periotic) and inner ear of a Pliocene kogiine whale
(Kogiinae, Odontoceti): implications on relationships and hearing evolution of
toothed whales. Journal of Vertebrate Paleontology 16:328-348.
Ma, D. –P., A. Zharkikh, D. Graur, J. L. Vandeberg, and W. –H. Li. 1993. Structure and
evolution of opossum, guinea pig, and porcupine cytochrome b genes. Journal of
Molecular Evolution 36:327-334.
MacFadden, B. J., and G. S. Morgan. 2003. New oreodont (Mammalia, Artiodactyla)
from the Late Oligocene (Early Arikareean) of Florida. Bulletin of the American
Museum of Natural History 279:368-396.
MacIntyre, G. T. 1972. The trisulcate petrosal pattern of mammals, pp. 275-303 in T.
Dobzhansky, M. K. Hecht, and W. C. Steere (eds.), Evolutionary Biology, Vol. 6.
Appleton-Century-Crofts, New York.
MacPhee, R. D. E. 1981. Auditory regions of primates and eutherian insectivores:
morphology, ontogeny, and character analysis; in F. S. Szalay (ed.), Contributions
to Primatology, Volume 18. S. Krager, Basel, 282 pp.
Macrini, T.E. 2004. Monodelphis domestica. Mamm Species 760:1-8.
426
Maddison, W. P., and D. R. Maddison. 2008. Mesquite: a modular system for
evolutionary analysis. Version 2.5 http://mesquiteproject.org.
Madsen, O., M. Scally, C. J. Douady, D. J. Kao, R. W. DeBry, R. Adkins, H. M. Amrine,
M. J. Stanhope, W. W. de Jong, M. S. Springer. 2001. Parallel adaptive radiations
in two major clades of placental mammals. Nature 409:610-614.
Maisano, J. A., and O. Rieppel. 2007. The skull of the Round Island boa, Casarea
dussumieri Schlegel, based on high-resolution X-ray computed tomography.
Journal of Morphology 268:371-384.
Maisano, J. A., C. J. Bell, J. A. Gauthier, and T. Rowe. 2002. The osteoderms and
palpebral in Lanthanotus borneensis (Squamata: Anguimorpha). Journal of
Herpetology 36:678-682.
Malia, M. J., R. M. Adkins, and M. W. Allard. 2002. Molecular support for Afrotheria
and the polyphyly of Lipotyphla based on analyses of the growth hormone
receptor gene. Molecular Phylogenetics and Evolution 24:91-101.
Manley, G. A. 1972. A review of some current concepts of the functional evolution of the
ear in terrestrial vertebrates. Evolution 26:608-621.
Manley, G. A. 2000. Cochlear mechanisms from a phylogenetic viewpoint. Proceedings
of the National Academy of Sciences of the United States of America 97:11736-
11743.
Manoussaki, D., E. K. Dimiriadis, and R. S. Chadwick. 2006. Cochlea’s graded curvature
effect on low frequency waves. Physical Review Letters 96(088701):1-4.
Manoussaki, D., R. S. Chadwick, D. R. Ketten, J. Arruda, E. K. Dimitriadis, and J. T.
O’Malley. 2008. The influence of cochlear shape on low-frequency hearing.
Proceedings of the National Academy of Sciences of the United States of America
105:6162-6166.
Martins, E. P. 1999. Estimation of ancestral states of continuous characters: a computer
simulation study. Systematic Biology 48:642-650.
McBee, K., and R. J. Baker. 1982. Dasypus novemcinctus. Mamm Spec 162:1-9.
McCrady, E., Jr. 1938. The embryology of the opossum. American Anatomical Memoirs
Number 16, 228 p.
McKenna, M. C. 1975. Toward a phylogenetic classification of the Mammalia; pp. 21-46
in W. P. Luckett, and F. S. Szalay (eds.), Phylogeny of the Primates, Plenum
Publishing Company, New York.
McKenna, M. C. 1992. The alpha crystalline A chain of the eye lens and mammalian
phylogeny. Annales Zoologici Fennici 28:349-360.
McKenna, M. C., and S. K. Bell. 1997. Classification of Mammals. Columbia University
Press, New York, 631 pp.
427
McKenna, M. C., Z. Kielan-Jaworowska, and J. Meng. 2000. Earliest eutherian mammal
skull, from the Late Cretaceous (Coniacian) of Uzbekistan. Acta Palaeontologica
Polonica 41:1-54.
McLaren, I. A. 1960. Are the Pinnipedia biphyletic? Systematic Zoology 9:18-28.
Meade, G. E. 1961. The saber-toothed cat, Dinobastis serus. Bulletin of the Texas
Memorial Museum 2:27-60.
Meng, J. 2004. Phylogeny and divergence of basal Glires. Bulletin of the American
Museum of Natural History 285:93-109.
Meng, J., and R. C. Fox. 1995. Osseous inner ear structures and hearing in early
marsupials and placentals. Zoological Journal of the Linnean Society 115:47-71.
Meng, J. and A. R. Wyss. 1995. Monotreme affinities and low-frequency hearing
suggested by multituberculate ear. Nature 377:141-144.
Meng, J, and A. R. Wyss. 2001. The morphology of Tribosphenomys (Rodentiaformes,
Mammalia): phylogenetic implications for basal Glires. Journal of Mammalian
Evolution 8:1-71.
Meng, J., and A. R. Wyss. 2005. Glires (Lagomorpha, Rodentia); pp. 145-158 in K. D.
Rose, and J. D. Archibald (eds.), The Rise of Placental Mammals: Origins and
Relationships of the Major Extant Clades. The Johns Hopkins University Press,
Baltimore.
Meng, J., Y. Hu, and C. Li. 2003. The osteology of Rhombomylus (Mammalia, Glires):
implications for phylogeny and evolution of Glires. Bulletin of the American
Museum of Natural History 275:1-247.
Mercury Computer Systems. 2003. Amira 3.1. Konrad-Zuse-Zentrum für
Infromationstechnik, Berlin.
Mess, A., and A. M. Carter. 2006. Evolutionary transformations of fetal membrane
characters in Eutheria with special reference to Afrotheria. Journal of
Experimental Zoology (Moleular Development and Evolution) 306B:140-163.
Miao, D. 1988. Skull morphology of Lambdopsalis bulla (Mammalia, Multituberculata)
and its implications to mammalian evolution. Contributions to Geology,
University of Wyoming Special Paper 4:1-104.
Miller, M. R. 1966a. The cochlear duct of lizards. Proceedings of the California Academy
of Sciences 33:255-359.
Miller, M. R. 1966b. The cochlear ducts of Lanthanotus and Anelytropsis with remarks
on the familial relationship between Anelytropsis and Dibamus. Occasional
Papers of the California Academy of Sciences 60:1-15.
Miller, M. R. 1968. The cochlear duct of snakes. Proceedings of the California Academy
of Sciences 35:425-475.
428
Misawa, K., and A. Janke. 2003. Revisiting the Glires concept – phylogenetic analysis of
nuclear sequences. Molecular Phylogenetics and Evolution 28:320-327.
Miyamoto, M. M., and M. Goodman. 1986. Biomolecular systematics of eutherian
mammals: phylogenetic patterns and classification. Systematic Zoology 35:230-
240.
Moody, P. A., V. A. Cochran, H. Drugg H. 1949. Serological evidence on lagomorph
relationships. Evolution 3:25-33.
Morsli, H., D. Choo, A. Ryan, R. Johnson, and D. K. Wu. 1998. Development of the
mouse inner ear and origin of its sensory organs. Journal of Neuroscience
18:3327-3335.
Mouchaty, S. K., A. Gullber, A. Janke, and U. Arnason. 2000. Phylogenetic position of
the tenrecs (Mammalia: Tenrecidae) of Madagascar based on analysis of the
complete mitochondrial genome sequence of Echinops telfairi. Zoologica Scripta
29:307-317.
Murata, Y., M. Nikaido, T. Sasaki, Y. Cao, Y. Fukumoto, M. Hasegawa, N. Okada. 2003.
Afrotherian phylogeny as inferred from complete mitochondrial genomes.
Molecular Phylogenetics and Evolution 28:253-260.
Muren, C., G. Ruhn, and H. Wilbrand. 1986. Anatomic variations of the human
semicircular canals: a radioanatomic investigation. Acta Radiologica Diagnosis
27:157-163.
Murphy, W. J., E. Eizirik, W. E. Johnson, Y. P. Zhang, O. A. Ryder, S. J. O’Brien.
2001a. Molecular phylogenetics and the origins of placental mammals. Nature
409:614-617.
Murphy, W. J., E. Eizirik, S. J. O’Brien, O. Madsen, M. Scally, C. J. Douady, E. Teeling,
O. A. Ryder, M. J. Stanhope, W. W. de Jong, M. S. Springer. 2001b. Resolution
of the early placental mammal radiation using Bayesian phylogenetics. Science
294:2348-2351.
Naylor, G. J. P., and D. C. Adams. 2001. Are the fossil data really at odds with the
molecular data? Morphological evidence for cetartiodactyla phylogeny
reexamined. Systematic Biology 50:444-453.
Nessov, L. A. 1985. [New mammals from the Cretaceous of Kyzylkum]. Vestnik
Leningradskogo Universiteta 17:8-18. [Russian]
Nessov, L. A. 1993. [New Mesozoic mammals of middle Asia and Kazakhstan, and
comments about evolution of theriofaunas of Cretaceous coastal plains of ancient
Asia]. Trudy Zoologicheskogo Instituta 249:105-133. [Russian 105-132; English
132-133]
429
Nessov, L. A., J. D. Archibald, and Z. Kielan-Jaworowska. 1998. Ungulate-like
mammals from the Late Cretaceous of Uzbekistan and a phylogenetic analysis of
Ungulatomorpha. Bulletin of Carnegie Museum of Natural History 34:40-88.
Nikaido, M., Y. Cao, M. Harada, N. Okada, and M. Hasegawa. 2003. Mitochondrial
phylogeny of hedgehogs and monophyly of Eulipotyphla. Molecular
Phylogenetics and Evolution 28:276-284.
Nilsson, M. A., A. Gullberg, A. E. Spotorno, U. Arnason, and A. Janke. 2003. Radiation
of extant marsupials after the K/T boundary: evidence from complete
mitochondrial genomes. Journal of Molecular Evolution 57:3-12.
Norman, J. E., and M. V. Ashley. 2000. Phylogenetics of Perissodactyla and tests of the
molecular clock. Journal of Molecular Evolution 50:11-21.
Norris, C. A. 1994. The periotic bones of possums and cuscuses: cuscus polyphyly and
the division of the marsupial family Phalangeridae. Zoological Journal of the
Linnean Society 111:73-98.
Novacek, M. J. 1977. Aspects of the problem of variation, origin and evolution of the
eutherian auditory bulla. Mammal Review 7:131-150.
Novacek, M. J. 1986. The skull of leptictid insectivorans and the higher-level
classification of eutherian mammals. Bulletin of the American Museum of
Natural History 183:1—111.
Novacek, M. J. 1992a. Fossils, topologies, missing data, and the higher level phylogeny
of eutherian mammals. Systematic Biology 41:58-73.
Novacek, M. J. 1992b. Mammalian phylogeny: shaking the tree. Nature 356:121-125.
Novacek, M. J. 1993. Reflections on higher mammalian phylogenetics. Journal of
Mammalian Evolution 1:3-30.
Novacek, M. J. and A. Wyss. 1986a. Origin and transformation of the mammalian stapes.
Contributions to Geology, University of Wyoming, Special Paper 3:35-53.
Novacek, M. J. and A. Wyss. 1986b. Higher-level relationships of the Recent eutherian
orders: morphological evidence. Cladistics 2:257-287.
Novacek, M. J., A. R. Wyss AR, M. C. McKenna. 1988. The major groups of eutherian
mammals; pp. 31-71 in M. J. Benton MJ (ed.), The Phylogeny and Classification
of the Tetrapods, Volume 2: Mammals. Clarendon Press, Oxford.
Ogden, C. L., C. D. Fryar, M. D. Carroll, and K. M. Flegal. 2004. Mean body weight,
height, and body mass index, United States 1960-2002. Advance Data from Vital
and Health Statistics; no 347. Hyattsville, Maryland. 20 p.
O’Leary, M. A. 1999. Parsimony analysis of total evidence from extinct and extant taxa
and the cetacean-artiodactyl question (Mammalia, Ungulata). Cladistics 15:315-
330.
430
O’Leary, M. A, and J. H. Geisler. 1999. The position of Cetacea within Mammalia:
phylogenetic analysis of morphological data from extinct and extant taxa.
Systematic Biology 48:455-490.
O’Leary, M. A, and M. D. Uhen. 1999. The time of origin of whales and the role of
behavioral changes in the terrestrial-aquatic transition. Paleobiology 25:534-556.
Owen, R. 1848. Description of teeth and portions of jaws of two extinct anthracotheroid
quadrupeds (Hyopotamus vectianu and Hyop. bovinus) discovered by the
Marchioness of Hastings in the Eocene deposits on the N.W. coast of the Isle of
Wight: with an attempt to develop Cuvier’s idea of the classification of
pachyderms by the number of their toes. Quarterly Journal of the Geolological
Society 4:103-141.
Olson, E. C. 1944. Origin of mammals based upon cranial morphology of the therapsid
suborders. Geological Society of America Special Papers 55: 1-136.
Payne, K. B., W. R. Langbauer, and E. M. Thomas. 1986. Infrasonic calls of the Asian
elephant (Elephas maximus). Behavioral Ecology and Sociology 18:297-301.
Pol, C., A. D. Buscalioni, J. Carballeira, V. Francés, N. L. Martinez, B. Marandat, J. J.
Moratalla, J. L. Sanz, B. Sigé, and J. Villatte. 1992. Reptiles and mammals from
the Late Cretaceous new locality Quintanilla del Coco (Burgos Province, Spain).
Neues Jahrbuch für Geologie und Paläontologie, Abhandlungen 184:279-314.
Poole, J. H., K. B. Payne, W. R. Langbauer, and C. J. Mos. 1988. Social contexts of some
very low frequency calls of African elephants. Behavioral Ecology and
Sociobiology 22:385-392.
Porter, C. A., M. Goodman, and M. J. Stanhope. 1996. Evidence on mammalian
phylogeny from sequences of Exon 28 of the von Willebrand Factor gene.
Molecular Phylogenetics and Evolution 5:89-101.
Poux, C., and E. J. P. Douzery. 2004. Primate phylogeny, evolutionary rate variations,
and divergence times: a contribution from the nuclear gene IRBP. American
Journal of Physical Anthropology 124:1-16.
Prothero, D. R., and R. J. Emry. 2004. The Chadronian, Orellan, and Whitneyan North
American land mammal ages; pp. 156-168 in M. O. Woodburne (ed.), Late
Cretaceous and Cenozoic Mammals of North America: Biostratigraphy and
Geochronology. Columbia University Press, New York.
Quiroga, J. C. 1979. The inner ear of two cynodonts (Reptilia – Therapsida) and some
comments on the evolution of the inner ear from pelycosaurs to mammals.
Gegenbauers Morphologisches Jahrbuch 2:178-190.
Radinsky, L. B. 1964. Paleomoropus, a new Early Eocene chalicothere (Mammalia,
Perissodactyla), and a revision of Eocene chalicotheres. American Museum
Novitates 2179:1-28.
431
Ramprashad, F., J. P. Landolt, K. E. Money, D. Clark, AND J. Laufer. 1979. A
morphometric study of the cochlea of the little brown bat (Myotis lucifugus).
Journal of Morphology 160:345-358.
Rasmussen, D. T., M. Gagnon M, and E. L. Simons EL. 1990. Taxepody in the carpus
and tarsus of Oligocene Pliohyracidae (Mammalia: Hyracoidea) and the phyletic
position of hyraxes. Proceedings of the Nationall Academy of Sciences of the
United States of America 87:4688-4691.
Raup, D. M., and J. J. Sepkoski, Jr. 1982. Mass extinctions in the marine fossil record.
Science 215:1501-1503.
Reeder, D. M., K. M. Helen KM, and D. E. Wilson. 2007. Global trends and biases in
new mammal species discoveries. Occasional Papers, Museum of Texas Tech
University 269:1-35.
Reimer, K. 1996. Ontogeny of hearing in the marsupial, Monodelphis domestica, as
revealed by brainstem auditory evoked potentials. Hearing Research 92:143-150.
Retzius, G. 1884. Das Gehörorgan der Wirbelthiere: morphologisch-histologische
studien. II. Das Gehörorgan der Reptilien, der Vögel und der Säugethiere. Samson
and Wallin, Stockholm, 368 pp.
Rogers, S. W. 1998. Exploring dinosaur neuropaleobiology: computed tomography
scanning and analysis of an Allosaurus fragilis endocast. Neuron 21:673-679.
Rogers, S. W. 1999. Allosaurus, crocodiles, and birds: evolutionary clues from spiral
computed tomography of an endocast. Anatomical Record Part B – New
Anatomist 257:162-173.
Rose, K. D., R. J. Emry, T. J. Gaudin, and G. Storch. 2005. Xenarthra and Pholidota; pp.
106-126 in K. D. Rose, and J. D. Archibald (eds.), The rise of placental mammals:
origins and relationships of the major extant clades. The Johns Hopkins
University Press, Baltimore.
Rougier, G. W., J. R. Wible, and M. J. Novacel. 1998. Implications of Deltatheridium
specimens for early marsupial history. Nature 396:459-463.
Rowe, T. 1986. Osteological diagnosis of Mammalia, L. 1758, and its relationship to
extinct Synapsida. Ph.D. dissertation, University of California, Berkeley, CA, 446
pp.
Rowe, T. 1988. Definition, diagnosis, and origin of Mammalia. Journal of Vertebrate
Paleontology 8:241-264.
Rowe, T. 1996. Coevolution of the mammalian middle ear and neocortex. Science
273:651-654.
Rowe, T. 1997. Response to “Comparative rates of development in Monodelphis and
Didelphis. Science 275:684.
432
Rowe, T., W. Carlson, and W. Bottorf. 1995. Thrinaxodon: Digital atlas of the skull. CD-
ROM (Second Edition for Windows and Macintosh platforms). University of
Texas Press, Austin.
Rowe, T., J. Kappelman, W. D. Carlson, R. A. Ketcham, and C. Denison. 1997. High-
resolution computed tomography: a breakthrough technology for earth scientists.
Geotimes 42:23-27.
Ruf, I., Z. –X. Luo, J. R. Wible, and T. Martin. 2009. Petrosal anatomy and inner ear
structures of the Late Jurassic Henkelotherium (Mammalia, Cladotheria,
Dryolestoidea): insight into the early evolution of the ear region in cladotherian
mammals. Journal of Anatomy 214:679-693.
Ryals, B. M., H. B. Creech, and E. W. Rubel. 1984. Postnatal changes in the size of the
avian cochlear duct. Acta Otolaryngolica 98:93-97.
Sánchez-Villagra, M. R., and T. Schmelzle. 2007. Anatomy and development of the bony
inner ear in the woolly opossum, Caluromys philander (Didelphimorphia,
Marsupialia). Mastozoología Neotropical 14:53-60.
Sánchez-Villagra, M. R., Y. Narita, and S. Kuratani. 2007. Thoracolumbar vertebral
number: the first skeletal synapomorphy for afrotherian mammals. Systematics
Biodiversity 5:1-7.
Sanderson, M. J., A. Purvis, and C. Henze. 1998. Phylogenetic supertrees: assembling the
trees of life. Trends in Ecology and Evolution 13:105-109.
Saunders, J. J. 1977. Late Pleistocene vertebrates of the Western Ozark Highland,
Missouri. Illinois State Museum Reports of Investigations 33:1-118.
Schmidt, R. S. 1964. Phylogenetic significance of lizard cochlea. Copeia 1964:542-549.
Schmitz, J., and H. Zischler. 2003. A novel family of tRNA-derived SINEs in the colugo
and two new retrotransposable markers separating dermopteras from primates.
Molecular Phylogenetics and Evolution 28:341-349.
Schmitz, J., M. Ohme, B. Suryobroto, and H. Zischler. 2002. The colugo (Cynocephalus
variegates, Dermoptera): the primates’ gliding sister? Molecular Biology and
Evolution 19:2308-2312.
Schuknecht, H. F. 1953. Techniques for study of cochlear function and pathology in
experimental animals. Archives of Otolaryngology 58:377-397.
Segall, W. 1970. Morphological parallelisms of the bulla and auditory ossicles in some
insectivores and marsupials. Fieldiana Zoology 51:169-205.
Seiffert, E. R. 2007. A new estimate of afrotherian phylogeny based on simultaneous
analysis of genomic, morphological, and fossil evidence. BMC Evolutionary
Biology 7(224):1-13.
433
Sheehan, P. M. and D. E. Fastovsky. 1992. Major extinctions of land-dwelling
vertebrates at the Cretaceous-Tertiary boundary, eastern Montana. Geology
20:556-560.
Shoshani, J. 1986. Mammalian phylogeny: comparison of morphological and molecular
results. Molecular Biology and Evolution 3:222-242.
Shoshani, J., M. C. McKenna M. 1998. Higher taxonomic relationships among extant
mammals based on morphology, with selected comparisons of results from
molecular data. Molecular Phylogenetics and Evolution 9:572-584.
Shoshani, J., C. P. Groves, E. L. Simons, and G. F. Gunnell. 1996. Primate phylogeny:
morphological vs molecular results. Molecular Phylogenetics and Evolution
5:102-154.
Shubin, N. H., E. B. Daeschler, and M. I. Coates. 2004. The early evolution of the
tetrapod humerus. Science 304:90-93.
Shute, C. C. D., and A. D’A. Bellairs. 1953. The cochlear apparatus of Geckonidae and
Pygopodidae and its bearing on the affinities of these groups of lizards.
Proceedings of the Zoological Society of Lond 123:695-709.
Silcox, M. T., J. I. Bloch, E. J. Sargis, D. M. Boyer. 2005. Euarchonta (Dermoptera,
Scandentia, Primates); pp. 127-144 in K. D. Rose, and J. D. Archibald (eds.), The
Rise of Placental Mammals: Origins and Relationships of the Major Extant
Clades. The Johns Hopkins University Press, Baltimore.
Silva, M., and J. A. Downing. 1995. CRC Handbook of Mammalian Body Masses. CRC
Press, Boca Raton, Florida, 359 pp.
Simmons, J. A., M. B. Fenton, and M. J. O’Farrell. 1979. Echolocation and pursuit of
prey by bats. Science 203:16-21.
Simmons, N. B., K. L. Seymour, J. Habersetzer, and G. F. Gunnell. 2008. Primitive Early
Eocene bat from Wyoming and the evolution of flight and echolocation. Nature
451:818-821.
Simpson, G. G. 1945. The principles of classification and a classification of mammals.
Bulletin of the American Museum of Natural History 84:1-350.
Sisson, S. and J. D. Grossman. 1938. The Anatomy of the Domestic Animals (third
edition, revised). W. B. Saunders Company, Philadelphia, 972 pp.
Solntseva, G. N. 2007. Morphology of the Auditory and Vestibular Organs in Mammals,
with Emphasis on Marine Species. Pensoft and Brill, Sofia, 244 pp.
Spoor, F. and F. Zonneveld. 1995. Morphometry of the primate bony labyrinth: a new
method based on high-resolution computed tomography. Journal of Anatomy
186:271-286.
434
Spoor, F., and F. Zonneveld. 1998. Comparative review of the human bony labyrinth.
Yearbook of Physical Anthropology 41:211-251.
Spoor, F., B. Wood, and F. Zonneveld. 1994. Implications of early hominid labyrinthine
morphology for evolution of human bipedal locomotion. Nature 369:645-648.
Spoor, F., B. Wood, and F. Zonneveld. 1996. Evidence for a link between human
semicircular canal size and bipedal behaviour. Journal of Human Evolution
30:183-187.
Spoor, F., S. Bajpai, S. T. Hussain, K. Kumar, and J. G. M. Thewissen. 2002. Vestibular
evidence for the evolution of aquatic behavior in early cetaceans. Nature 417:163-
166.
Spoor, F., T. Garland, Jr., G. Krovitz, T. M. Ryan, M. T. Silcox, and A. Walker. 2007.
The primate semicircular canal system and locomotion. Proceedings of the
National Academy of Sciences of the United States of America 104:10808-10812.
Springer, M. S., H. M. Amrine, A. Burk, M. J. Stanhope. 1999. Additional support for
Afrotheria and Paenungulata, the performance of mitochondrial versus nuclear
genes, and the impact of data partitions with heterogeneous base composition.
Systematic Biology 48:65-75.
Springer, M. S., W. J. Murphy, E. Eizirik, S. J. O’Brien. 2003. Placental mammal
diversification and the Cretaceous-Tertiary boundary. Proceedings of the National
Academy Sciences 100:1056-1061.
Springer, M. S., G. C. Cleven, O. Madsen, W. W. de Jong, V. G. Waddell, H. M. Amrine,
and M. J. Stanhope. 1997. Endemic African mammals shake the phylogenetic
tree. Nature 388:61-64.
Springer, M. S., M. Westerman, J. R. Kavanagh, A. Burk, M. O. Woodburne, D. J. Kao,
and C. Krajewski. 1998. The origin of the Australasian marsupial fauna and the
phylogenetic affinities of the enigmatic monito del monte and marsupial mole.
Proceedings of the Royal Society of London B 265:2381-2386.
Stan!k, V. J. 1933. K topografické a srovnávací anatomii sluchového orgánu na$ich
chiropter. Nákladem %eské Akademie v!d a Um!ní, Praze (Prague), 67 pp.
Stanhope, M. J., J. Czelusniak, J. –S. Si, J. Nickerson, and M. Goodman M. 1992. A
molecular perspective on mammalian evolution from the gene encoding
interphotoreceptor retinoid binding protein, with convincing evidence for bat
monophyly. Molecular Phylogenetics and Evolution 1:148-160.
Stanhope, M. J., V. G. Waddell, O. Madsen, W. de Jong, S. B. Hedges, G. C. Cleven, D.
Kao, and M. S. Springer. 1998. Molecular evidence for multiple origins of
Insectivora and for a new order of endemic African insectivore mammals.
Proceedings of the National Academy of Sciences of the United States of America
95:9967-9972.
435
Tedford, R. H. 1976. Relationship of pinnipeds to other carnivores (Mammalia).
Systematic Zoology 25:363-374.
Teeling, E. C., M. Scally, D. J. Kao, M. L. Romagnoll, M. S. Springer, and M. J.
Stanhope. 2000. Molecular evidence regarding the origin of echolocation and
flight in bats. Nature 403:188-192.
Teeling, E. C., M. S. Springer, O. Madsen, P. Bates, S. J. O’Brien, and W. J. Murphy.
2005. A molecular phylogeny for bats illuminates biogeography and the fossil
record. Science 307:580-584.
Theodor, J. M, and S. E. Foss SE. 2005. Deciduous dentitions of Eocene cebochoerid
artiodactyls and cetartiodactyl relationships. Journal of Mammalian Evolution
12:161-181.
Theodor, J. M., K. D. Rose, and J. Erfurt. 2005 Artiodactyla; pp. 215-233 in K. D. Rose,
and J. D. Archibald (eds.), The Rise of Placental Mammals: Origins and
Relationships of the Major Extant Clades. The Johns Hopkins University Press,
Baltimore.
Thewissen, J. G. M., and S. I. Madar. 1999. Ankle morphology of the earliest cetaceans
and its implications for the phylogenetic relations among ungulates. Systematic
Biology 48:21-30.
Thewissen, J. G. M., E. M. Williams, L. J. Roe, and S. T. Hussain ST. 2001. Skeletons of
terrestrial cetaceans and the relationship of whales to artiodactyls. Nature
413:277-281.
Tomasi, T. E. 1979. Echolocation by the short-tailed shrew Blarina brevicauda. Journal
Mammal 60:751-759.
Tykoski, R. S. 2005. Anatomy, ontogeny, and phylogeny of coelophysoid theropods.
Unpublished Ph.D. Dissertation, The University of Texas at Austin, 572 pp.
Tykoski, R. S., T. B. Rowe, R. A. Ketcham, and M. W. Colbert. 2002. Calsoyasuchus
valliceps, a new crocodyliform from the Early Jurassic Kayenta Formation of
Arizona. Journal of Vertebrate Paleontology 22:593-611.
Tomasi, T. E. 1979. Echolocation by the short-tailed shrew Blarina brevicauda. Journal
of Mammalogy 60:751-759.
Ursing, B. M., and U. Arnason. 1998. Analyses of mitochondrial genomes strongly
support a hippopotamus-whale clade. Proceedings of the Royal Society of London
B 265:2251-2255.
Vaughn, T. A., J. M. Ryan, and N. J. Czaplewski. 2000. Mammalogy, Fourth Edition.
Saunders College, Philadelphia, 565 pp.
436
van der Klaauw, C. J. 1931. The auditory bulla in some fossil mammals: with a general
introduction to this region of the skull. Bulletin of the American Museum of
Natural History 67:1-352.
van Dijk, M. A. M., O. Madsen, F. Catzeflis, M. J. Stanhope, W. W. de Jong, and M.
Pagel. 2001. Protein sequence signatures support the African clade of mammals.
Proceedings of the National Academy of Sciences 98:188-193.
Van Kampen, P. N. 1904. De Tympanaalstreek van den Zoogdierschedel. P. N. Van
Kampen and Zoon, Amsterdam, 378 pp.
Van Valen, L. 1965. Treeshrews, primates, and fossils. Evolution 19:137-151.
Van Valen, L. 1971. Adaptive zones and the orders of mammals. Evolution 25:420-428.
Volume Graphics. 2004. VGStudio Max 1.2. Heidelberg, Germany.
Waddell, P. J., N. Okada, M. Hasegawa. 1999. Towards resolving the interordinal
relationships of placental mammals. Systematic Biology 48:1-5.
Wang, Z., G. B. Hubbard, S. Pathak, and J. L. VandeBerg. 2003. In vivo opossum
xenograft maodel for cancer research. Cancer Research 63:6121-6124.
Watson, M. 1874. Contributions to the anatomy of the Indian elephant, Part IV. Muscles
and blood-vessels of the face and head. Journal of Anatomy and Physiology
9:118-133.
West, C. D. 1985. The relationship of the spiral turns of the cochlea and the length of the
basilar membrane to the range of audible frequencies in ground dwelling
mammals. Journal of the Acoustical Society of America 77:1091-1101.
Wever, E. G., J. G. McCormick, J. Palin, and S. H. Ridgway. 1971a. The cochlea of the
dolphin, Tursiops truncatus: general morphology. Proceedings of the National
Academy of Sciences, USA 68:2381-2385.
Wever, E. G., J. G. McCormick, J. Palin, and S. H. Ridgway. 1971b. Cochlea of the
dolphin, Tursiops truncatus: the basilar membrane. Proceedings of the National
Academy of Sciences, USA 68:2708-2711.
Wible, J. R. 1987. The eutherian stapedial artery: character analysis and implications for
superordinal relationships. Zoological Journal of the Linnean Society 91:107-135.
Wible, J. R. 1990. Petrosals of Late Cretaceous marsupials from North America, and a
cladistic analysis of the petrosal in therian mammals. Journal of Vertebrate
Paleontology 10:183-205.
Wible, J. R. 2003. On the cranial osteology of the short-tailed opossum Monodelphis
brevicaudata (Didelphidae, Marsupialia). Annals of the Carnegie Museum
72:137-202.
Wible, J. R. and D. L. Davis. 2000. Ontogeny of the chiropteran basicranium, with
reference to the Indian false vampire bat, Megaderma lyra, pp. 214-246 in R. A.
437
Adams, S. C. Pedersen (eds.), Ontogeny, Functional Ecology, and Evolution of
Bats. Cambridge University Press, Cambridge.
Wible, J. R. and J. R. Martin. 1993. Ontogeny of the tympanic floor and roof in
archontans, pp. 111-148 in R. D. E. MacPhee (ed.), Primates and Their Relatives
in Phylogenetic Perspective. Plenum Press, New York.
Wible, J. R. and M. J. Novacek. 1988. Cranial evidence for the monophyletic origin of
bats. American Museum Novitates, Number 2911:1-19.
Wible, J. R., M. J. Novacek, and G. W. Rougier. 2004. New data on the skull and
dentition in the Mongolian Late Cretaceous eutherian mammal Zalambdalestes.
Bulletin of the American Museum of Natural History 281:1-144.
Wible, J. R., G. W. Rougier, and M. J. Novacek. 2005. Anatomical evidence for
superordinal eutherian taxa in the Cretaceous, pp. 15-36 in K. D. Rose and J. D.
Archibald (eds.), The Rise of Placental Mammals: Origins and Relationships of
the Major Extant Clades. Johns Hopkins University Press, Baltimore.
Wible, J. R., G. W. Rougier, M. J. Novacek, and R. J. Asher. 2007. Cretaceous eutherians
and Laurasian origin for placental mammals near the K/T boundary. Nature
447:1003-1006.
Wible, J. R., G. W. Rougier, M. J. Novacek, and M. C. McKenna. 2001. Earliest
eutherian ear region: a petrosal referred to Prokennalestes from the Early
Cretaceous of Mongolia. American Museum Novitates 3322:1-44.
Wiens, J. J., R. M. Bonett, and P. T. Chippindale. 2005. Ontogeny discombobulates
phylogeny: paedomorphosis and higher-level salamander relationships.
Systematic Biology 54:91-110.
Wilson, D. E., D. M. Reeder. 2005. Mammal Species of the World: A Taxonomic and
Geographic Reference (Third Edition). The Johns Hopkins University Press,
Baltimore, 2142 pp.
Wilson, J. A. 1971. Early Tertiary vertebrate faunas, Vieja Group, Trans-Pecos Texas:
Agrichoeridae and Merycoidontidae. Bulletin of the Texas Memorial Museum
18:1-83.
Witmer, L. M., S. Chatterjee, J. Franzosa, and T. Rowe. 2003. Neuroanatomy of flying
reptiles and implications for flight, posture and behaviour. Nature 425:950-953.
Wood, A. E. 1957. What, if anything, is a rabbit? Evolution 11:417-425.
Wood, A. E. 1962. The Early Tertiary rodents of the family Paramyidae. Transactions of
the American Philosophical Society 52:3-261.
Wyss, A. R. 1987. The walrus auditory region and the monophyly of pinnipeds.
American Museum Novitates 2871:1-7.
438
Yang, A. and T. E. Hullar. 2007. Relationship of semicircular canal size to vestibular-
nerve afferent sensitivity in mammals. Journal of Neurophysiology 98:3197-3205.
439
Vita
Eric Gregory Ekdale was born in Salt Lake City, UT. He attended Judge Memorial
Catholic High School, and he graduated in 1995. He matriculated at Augustana College
in Rock Island, IL, and he received a Bachelor of Arts degree in Biology and
Scandinavian Studies (separate majors) in the spring of 1999. During his education at
Augustana College, he also enrolled in summer courses at the University of Utah in Salt
Lake City, UT, but no degree was conferred from that institution. In the fall of 1999, he
attended classes at San Diego State University in San Diego, CA, where he earned a
Masters of Science degree in Biology (with an emphasis in Systematics, Evolutionary
and Organismal Biology) in May 2002. He entered the graduate school at The University
of Texas at Austin in 2002, where he immediately began work towards a Doctor of
Philosophy degree in Geological Sciences. During his time at The University of Texas,
Eric presented at the annual meetings of the Society of Vertebrate Paleontology and
Texas Academy of Sciences, almost on a yearly basis. During the 2008-2009 academic
year, he was awarded the Hogg Fellowship from The University of Texas at Austin. The
Hogg Fellowship is one of the highest honors awarded by the university to graduates
students.
Permanent address: 3816 South Lamar Blvd, Apt. #1804, Austin, TX 78704
This dissertation was typed by the author.
