BANNIKOV

CONTRIBUTIONS JO PALEOBIOLOGY

NUMBER 72

A Remarkable New Genus of
Tetraodontiform Fish with
Features of Both Balistids
and Ostraciids from the
Eocene of Turkmenistan

, rjw

m Wi Mr

ISH

mM :

y 'A

SERIES PUBLICATIONS OF THE SMITHSONIAN INSTITUTION

Emphasis upon publication as a means of “diffusing knowledge” was expressed by the first
Secretary of the Smithsonian. In his formal plan for the Institution, Joseph Henry outlined a
program that included the following statement: “It is proposed to publish a series of reports,
giving an account of the new discoveries in science, and of the changes made from year to year
in all branches of knowledge.” This theme of basic research has been adhered to through the
years by thousands of titles issued in series publications under the Smithsonian imprint,
commencing with Smithsonian Contributions to Knowledge in 1848 and continuing with the
following active series:

Smithsonian Contributions to Anthropology
Smithsonian Contributions to Astrophysics
Smithsonian Contributions to Botany
Smithsonian Contributions to the Earth Sciences
Smithsonian Contributions to the Marine Sciences
Smithsonian Contributions to Paleobiology
Smithsonian Contributions to Zoology
Smithsonian Folklife Studies
Smithsonian Studies in Air and Space
Smithsonian Studies in History and Technology

In these series, the Institution publishes small papers and full-scale monographs that report
the research and collections of its various museums and bureaux or of professional colleagues
in the world of science and scholarship. The publications are distributed by mailing lists to
libraries, universities, and similar institutions throughout the world.

Papers or monographs submitted for series publication are received by the Smithsonian
Institution Press, subject to its own review for format and style, only through departments of the
various Smithsonian museums or bureaux, where the manuscripts are given substantive review.
Press requirements for manuscript and art preparation are outlined on the inside back cover.

Robert McC. Adams

Secretary

Smithsonian Institution

SMITHSONIAN CONTRIBUTIONS TO PALEOBIOLOGY • NUMBER 72

A Remarkable New Genus of
Tetraodontiform Fish with
Features of Both Balistids
and Ostraciids from the
Eocene of Turkmenistan

James C. Tyler
and Alexandre F. Bannikov

SMITHSONIAN INSTITUTION PRESS
Washington, D.C.

1992

ABSTRACT

Tyler, James C., and Alexandre F. Bannikov. A Remarkable New Genus of Tetraodontiform
Fish with Features of Both Balistids and Ostraciids from the Eocene of Turkmenistan.
Smithsonian Contributions to Paleobiology, number 72, 14 pages, 4 figures, 1 table, 1992.—A
new genus and species of tetraodontiform fish, Eospinus daniltshenkoi, is described from the
Lower Eocene of Turkmenistan (Danatinsk Formation). It is referred to the Balistoidea because
it has three large dorsal-fin spines; the pelvic fin reduced to a rudimentary but prominent
structure apparently composed of two partially fused spines at the posterior end of the pelvis; and
enlarged scale plates that form an incomplete carapace or loosely articulated armature around
much of the body. It differs from all other balistoids in having a long median spine projecting
forward from the snout and another spine projecting posteriorly from the middle of each side of
the body; lower jaw teeth twice as long as the upper jaw teeth; and in lacking encasing scales
around the rudimentary pelvic spine. Eospinus is the first record of a balistoid fish from the
Eocene with three dorsal-fin spines and the pelvic spines fused together at the end of the pelvis,
as otherwise only occurs in balistids, which are first recorded from the Oligocene.

Official publication date is handstamped in a limited number of initial copies and is
recorded in the Institution’s annual report, Smithsonian Year. Series cover design; The
trilobite Phacops rana Green.

Library of Congress Cataloging-in-Publication Data
Tyler, James C., 1935—

A remarkable new genus of Tfetraodonuform fish with features of both Balistids and Ostraciids from the Eocene of
"Rirkmenistan / James C. Tyler and Alexandre F. Bannikov,
p. cm. - (Smithsonian contributions to paleobiology ; no. 72.)

Includes bibliographical references.

1. Eospinus daniltshenkoi—Turkmenistan. 2. Paleontology—Eocene. 3. Paleontology—Turkmenis¬
tan. I. Bannikov, A. F. (Aleksandr Fedorovich) II. Series.

QE852.T48T95 1992 567'.5-dc20 92-23372

© The paper used in this publication meets the minimum requirements of the American
National Standard for Permanence of Paper for Printed Library Materials Z39.48—1984.

Contents

Page

Introduction. 1

Acknowledgments. 1

Eospinus, new genus. 2

Type Species. 2

Etymology. 2

Diagnosis. 2

Eospinus daniltshenkoi, new species. 2

Material . 2

Etymology. 2

Diagnosis . 2

Type Locality and Horizon. 2

Ichthyofaunal Associations. 2

Description. 2

Analysis of Characters of Eospinus . 5

Number and Placement of Dorsal-fin Spines. 8

Length of Dorsal-fin Spines. 8

Pelvic Fin. 9

Carapace Scale Plates.10

Carapace Spines.11

Teeth .12

Plesiomorphic Features.12

Phylogenetic Conclusion.12

Literature Cited.14

iii

A Remarkable New Genus of
Tetraodontiform Fish with
Features of Both Balistids
and Ostraciids from the
Eocene of Turkmenistan

James C. Tyler
and Alexandre F. Bannikov

Introduction

Examination of the specimens of tetraodontiform and
possibly related fishes in the collections of the Paleontological
Institute in Moscow of the Russian Academy of Sciences has
revealed new and unusual members of several families. Among
them is a single specimen of a balistoid with specializations
that include a long, anteriorly directed median spine on the
snout, another spine directed posteriorly on each side of the
body, and much larger teeth in the lower jaw than in the upper
jaw. It has a covering of enlarged ostraciid-like scale plates
forming a partial carapace around the body, three dorsal spines
placed over the rear of the skull like balistids but without a
locking mechanism, and a rudimentary pelvic fin at the end of
the pelvis like balistids but without encasing scales. The
uniqueness of this heavily armored species was recognized by
P.G. Daniltshenko, the pioneering chronicler of the fossil fish
fauna of Russia and adjacent regions, and a sketch he made of
it accompanies the specimen in the collection. The specimen
has remained undescribed, apparently because of uncertainty
about its familial or even ordinal placement.

We place this new form in the Tetraodontiformes because it
has a reduced number of caudal-fin rays (12), a reduced number
of vertebrae (probably 18-20), and because it lacks anal-fin

James C. Tyler, National Museum of Natural History, Smithsonian
Institution, Washington, D C. 20560. Alexandre F. Bannikov, Paleon¬
tological Institute, Russian Academy of Sciences, Profsoyuznaya 123,
117647 Moscow, Russia.

spines. We place it in the Balistoidea (sensu Winterbottom,
1974: the balistids, including monacanthids; the ostraciids,
including aracanids; and the Eocene spinacanthids) because it
has a reduced number (3) of dorsal-fin spines, a reduced pelvic
fin (a rudimentary spine at the end of the pelvis), a reduced
number of teeth, which are of heavy conical shape, and
enlarged scale plates forming a partial carapace with spiny
processes. In several ways the new form is anatomically
intermediate between the ostraciid trunkfishes and the clade
composed of the balistid triggerfishes and their Eocene
spinacanthid sister group. We place the new form incertae sedis
in an unresolved trichotomy with the balistids and spinacan¬
thids. This is reflected in the following hierarchical synopsis.

Order TETRAODONTIFORMES Berg (1940), Winterbottom (1974), Tyler
(1980)

Suborder TETRAODONTOIDEI (this and other intraordinal categories
sensu Winterbottom, 1974)

Superfamily BALISTOIDEA
Family Ostraciidae
Incertae sedis: Eospinus
Family SP1 NACANTHIDAE
Family BALISTIDAE

Acknowledgments.— We appreciate the support of the
Smithsonian Institution’s Office of Fellowships and Grants and
the Russian Academy of Sciences for travel funds that
facilitated the completion of this and other projects involving
fossil tetraodontiforms. We thank R. Winterbottom, Royal
Ontario Museum, Toronto, and K. Matsuura, National Science

1

2

SMITHSONIAN CONTRIBUTIONS TO PALEOBIOLOGY

Museum, Tokyo, for their helpful reviews of the initial
manuscript, which was further improved by the preacceptance
reviews of C. Patterson, the Natural History Museum (Lon¬
don), J.D. Stewart, Los Angeles County Museum of Natural
History, and V.G. Springer, National Museum of Natural
History (Smithsonian Institution), and comments by G.D.
Johnson and J.T. Williams, National Museum of Natural
History. For much help with the cladistic analysis, we are
greatly indebted to R. Winterbottom. At the National Museum
of Natural History invaluable help was received from R. Clark
for photographs, J. Lombardo for typing, and F. Grady for
specimen conservation. At the Smithsonian Institution Press
we thank J.C. Warren for his careful copy editing and
typesetting and D.M. Tyler for preparing the table of
characteristics of the taxa.

Eospinus, new genus

Type Species. —Eospinus daniltshenkoi n. sp., by mono-
typy.

ETYMOLOGY. —Eo for the Eocene age, and spinus for the
three large dorsal-fin spines, the carapace spines, and the fused
pelvic-fin spine; masculine.

DIAGNOSIS. —Eospinus differs from all other Balistoidea by
the presence of a large anteriorly directed median spine on the
snout and a posteriorly directed spine on each side of the body,
heterodont dentition, absence of a locking mechanism of the
three dorsal-fin spines, and absence of encasing scales around
the rudimentary pelvic fin at the end of the pelvis.

Eospinus daniltshenkoi, new species

Figures 1-3

Material. —Holotype and only known specimen. Paleon¬
tological Institute (PIN), Moscow, no. 2179-101, 31.5 mm SL,
in part and counterpart, the more complete of the two halves
being that with the head to the left, collected 16 October 1959
(Figure 1).

ETYMOLOGY.—The patronym is in honor of PG. Danil-
tshenko, now retired from his grand work at the Paleontological
Institute in Moscow of describing the fossil fishes of Russia
and adjacent regions, who first recognized the uniqueness of
this species. We spell the name in the manner he prefers for its
Latinization, with a ts rather than a c, although in literature
references below we use the more commonly cited western
spelling Danilchenko.

Diagnosis. —As for the monotypic genus.

Type Locality and Horizon. —Middle part of Danatinsk
Formation, Lower Eocene of southwestern Turkmenistan, close
to the village of Uylya-Kushlyuk (38°38'N, 55°48'E). The
Danatinsk Formation is composed of deposits from the Upper
Paleocene to the Middle Eocene (Grossheim and Korobkov,
1975) and has a thickness of 200 m of mottled clays and

argillaceous marls. The lower boundary of the Eocene in these
deposits is marked by the replacement of foraminifera of the
Acarinina subsphaerica Zone complex by those of the
Globorotalia subbotinae Zone complex. The fish-bearing layer
of brown clays (whose total thickness is about 9 m) lies at the
base of the lower Eocene portion of the Danatinsk Formation
(Solun, 1975). Danilchenko (1968) was in error in supposing
that this layer belonged to the lower part of the Danatinsk
Formation and therefore was Upper Paleocene in age.

Ichthyofaunal Associations. —Up to now 23 species of
teleost fishes have been described from the Lower Eocene of
the middle part of the Danatinsk Formation (Danilchenko,
1968; Bannikov, 1985). These are mostly species of such
pelagic families as the Scombridae, Carangidae, Clupeidae,
Menidae, Chanidae, Turkmenidae, Kushlukiidae, and Luvar-
idae. However, at least one species, the acanthuroid Sigano-
pygaeus rarus Danilchenko, is a member of the inshore,
benthic family Siganidae, often associated with coral reefs and
adjacent grass flats. We suppose that Eospinus, being heavily
armored and probably slow moving, was benthic in habitat and
perhaps associated with reefs, as would be expected of a
balistoid fish.

Description. —The specimen is only moderately well
preserved. The tear-drop shaped body tapers from a deep head
to a short, deep, caudal peduncle; the eye is of moderate size in
the upper part of the head; the mouth is small and slightly
underslung, with distinct teeth of moderate (upper jaw) to large
(lower jaw) size; especially prominent features are the long
dorsal-fin spines and the median carapace spine on the snout.
Very little of the internal structure is exposed because of the
covering of enlarged but often fragmented or incomplete scale
plates. The part of the parasphenoid under the orbit, the
opercular bones, and the posterior edge of the pectoral girdle
and postcleithrum are recognizable. There is no indication of a
nasal organ. Nothing of diagnostic interest internally is clear,
except in a portion of the vertebral column that curves upward
over the abdominal region from the level of about the base of
the posteriorly directed carapace spine. Somewhat vague
contours and vertical grooves in the impression of the column
offer faint evidence for the size of the individual centra. Our
subjective interpretation is that the vertebrae were relatively
large and few in number, consistent with the low number of
about 18-20 as found in triacanthoid and most balistoid
tetraodontiforms.

Measurements are standard length (SL) 31.5 mm, greatest
depth of body 14.3 mm, approximate diameter of eye 4.0 mm,
distance between tip of snout and base of spiny dorsal fin 13.4
mm, length of base of spiny dorsal fin 7.0 mm (with position of
rear edge of recumbent third dorsal spine estimated as though
it were erected), distance between posterior end of base of spiny
dorsal fin and upper base of caudal fin 16.9 mm, caudal
peduncle depth 4.6 mm, approximate length of caudal fin 7.0
mm (position of rear edge estimated), length of incomplete first
dorsal spine 13.0 mm (based on the thickness and tapering of

FIGURE 1.—Photograph of the holotype of Eospinus daniltshenkoi, new species, 31.5 mm SL, PIN 2179-101,
Lower Eocene of Turkmenistan.

the preserved portion, the intact length is estimated to be about
23 mm), length of second dorsal spine 19.7 mm, length of third
dorsal spine 13.1 mm, length of pelvic-fin apparatus protruding
from ventral profile of body 4.3 mm, length of incomplete
spine projecting anteriorly from snout 8.8 mm (intact length
perhaps about 10 to 12 mm), length of longest tooth in lower
jaw 1.4 mm, length of longest tooth in upper jaw approximately
0.7 mm.

The body is covered with enlarged scale plates bearing
numerous low tubercles or granulations on their surface, with
the most central tubercle often larger than the others. Although
the peripheral edges of the plates are not well preserved, the
plates appear to have varied from irregularly rounded to almost
hexagonal. The plates are largest and in closest contact with
one another in about the middle of the body, apparently
forming an almost continuous covering of armor in the region
that supports the posteriorly directed carapace spine and the
rudimentary pelvic fin. It is possible that some of the plates
were interdigitated, although there is no evidence of the type of
more extensive suturing that characterizes the carapace of
ostraciids. The scale plates of the head and caudal peduncle are
somewhat smaller than those of the middle of the body and are
more distantly spaced from one another.

Both the median snout spine and the posterior body spine on
each side of the body appear to be processes of the enlarged
dermal scale plates with which they are continuous basally;
thus, both spines were probably fixed and inflexible. Continu¬
ing anteriorly from the base of the posterior body spine is a
ridge along the enlarged scale plates, probably forming a crest
along the body to below the pectoral-fin base. An even more
prominent crest, along what are probably relatively consoli¬
dated scale plates, is present ventrolaterally from the region of
the base of the rudimentary pelvic fin to the lower region of the
cheek. Because this crest terminates at the base of the pelvic fin,
we suppose that the pelvic girdle, which is not exposed, is
positioned internal to it. The centers of the scale plates along
this crest apparently were especially upraised, forming a
spinous ventrolateral ridge. The snout spine bears shallow
longitudinal grooves and low serrations, while the posterior
body spine appears to be similarly but slightly less ornamented.

All three dorsal-fin spines are long and robust and at least the
first two bear low serrations along the anterior and, especially
on the second spine, lateral edges. The distal end of the first
dorsal spine is missing, but based on the thickness and degree
of tapering of the basal region we estimate that the first spine
was only slightly longer (73% SL) than the second (63% SL),

4

SMITHSONIAN CONTRIBUTIONS TO PALEOBIOLOGY

while by measurement the length of the third (42% SL) is
slightly more than half that of the second. The large rounded
base of the second dorsal spine does not appear to make close
contact with the posterior edge of the base of the first spine.
Therefore, we think it unlikely that the complex locking
mechanism of the first spine by the second spine as found in
Recent and fossil (as early as the Oligocene) balistids is
present. The origin of the spiny dorsal fin is at a level about half
way between the eye and the region of the gill slit and
pectoral-fin base, and the anterior end of its supporting basal
pterygiophores (which cannot be seen because of overlying
large scale plates) probably articulated with the posterodorsal
surface of the skull.

The base of the caudal fin is preserved and there appear to
have been 12 rays, but we cannot determine how many were
branched or unbranched because the distal part of the fin is only
a faint impression. There are no remains of the pectoral, soft
dorsal, or anal fins, and these are shown hypothetically by
dashed lines in the reconstruction (Figure 2). We presume that
there were no anal-fin spines because if such were present they
could be expected to be as easily visible as the dorsal-fin
spines, the rudimentary pelvic-fin spine and the carapace
spines.

The pelvic fin, although rudimentary, protrudes prominently

from the ventral contour of the body. The apparatus as exposed
is seen in dorsoventral view (Figure 3) and, therefore, its
bilateral symmetry is clear, especially with the clay-like matrix
having been carefully removed from along its length. The bony
element appears to represent the consolidation of the right and
left pelvic spines, as it is a continuous, solid, uninterrupted
piece without cross-striations (segmentations) or branching as
would be found in a soft ray. The two halves of the spine are
fully fused to one another medially except at the distal end.
Indentations and irregularities medially along the length of the
spine also are evidence of the region of fusion. Strong spiny
processes are present along its lateral edges, the more distal of
which tend to be retrorse. These spiny processes appear to be
fully continuous outgrowths of the spine itself. The base of the
pelvic spine abuts against the enlarged scale plate at the
posterior end of the ventrolateral ridge along the body armor.
At the end of the ventrolateral ridge this scale plate curves
medially to meet the base of the spine and presumably overlies
the posterior end of the pelvic girdle where the spine would
probably be attached. Just above the base of the pelvic spine the
body is indented, undoubtedly for the anus. There is no
evidence of elongate scales between the anus and the base of
the spine, such as those in Recent and fossil balistids that
strengthen the somewhat expansible dewlap or fan of skin in

FIGURE 2.—Reconstruction of the holotype of Eospinus daniltshenkoi , new species, 31.5 mm SL, PIN 2179-101,
Lower Eocene of Turkmenistan.

NUMBER 72

5

FIGURE 3.—Rudimentary pelvic-fin element of the holotype of Eospinus daniltshenkoi, new species, as seen in
dorsoventral view, base of element to left, the pelvic spines from the right and left sides fused together except
distally.

this region. Neither is there evidence of a distinct series of
scales forming a sheath around the pelvic spine.

The lower jaw teeth are bluntly conical, slightly curved
posteriorly toward the distal end, and twice as long as those of
the upper jaw. The first four teeth on the side of the jaw are
clearly indicated (the second and third by impressions only)
and are of about similar length and width, while the few more
posterior teeth (probably two) apparently decrease in size
(these are not so evident as the first four). The upper jaw teeth
are more numerous, with about eight teeth clearly evident in a
single row. At least some of the upper jaw teeth, especially
those toward the middle of the series, have more constricted
distal regions than in the lower jaw, forming blunt nipples. In
neither jaw is there evidence of inner series teeth, although such
teeth, especially if small, could be hidden behind the large outer
teeth.

Analysis of Characters of Eospinus

For purposes of establishing the systematic position of
Eospinus we accept Winterbottom’s (1974) higher classifica¬
tion and phylogenetic analysis of the relationships of the
Tetraodontifonnes, i.e., the Suborder Triacanthoidei (Triacan-
thodidae and Triacanthidae) is the sister group of the Suborder
Tetraodontoidei (all of the other families; see Table 1 for a
synopsis of the classification of tetraodontiforms and Figure 4
for their phylogenetic relationships). Within the Tetraodonto¬
idei, the Superfamily Balistoidea (with, among other charac¬
ters, individual teeth protruding from the jaws; plesiomorphic)
is the sister group of the Superfamily Tetraodontoidea (teeth
incorporated with jaw bones into a beak; apomorphic). Within
the Balistoidea, the Spinacanthidae (two Eocene species) is the
sister group of the Balistidae (including monacanthids), while
the sister group of the balistid-spinacanthid clade is the
Ostraciidae (including aracanids). We follow these categories
in the text discussion and in Table 1 summarizing the
characters of Eospinus in comparison to all of the other major

groups of tetraodontiforms, but in the cladogram (Figure 4) we
follow the scheme of Tyler (1980) and recognize the
monacanthids and aracanids at the familial level rather than the
subfamilial as accorded them in Winterbottom (1974, on which
the cladogram is based).

On the basis of 96 osteological and myological autapomor-
phies, Winterbottom and Tyler (1983) defined the Recent
Balistoidea (20 autapomorphies), Balistidae (20), and Ostra¬
ciidae (56). Only a few of these are external features that can be
seen in both fossil and Recent species. The Spinacanthidae
were not treated by Winterbottom and Tyler (1983) and are
analyzed here on the basis of the descriptions in Tyler (1980).
While the phylogeny of the Recent balistoids as proposed by
Winterbottom (1974) and Winterbottom and Tyler (1983) is
robust in osteological and myological characters, the placement
of the poorly known Eocene spinacanthids (single specimens of
each of the two species, with only external features evident)
with the balistoids and its sister group relationship with the
balistids is based on far fewer characters (elongate ethmoid
region, eye high in head, spiny dorsal fin forward on head over
eye; Winterbottom, 1974). Even with the limitations from the
lack of knowledge of the osteology of spinacanthids we accept
these relationships of the balistoids and believe that the few
additional features analyzed here for Eospinus will not
substantially change them. Obviously, however, as more and
more data become available in new taxa like Eospinus with
unique combinations of balistoid characters, or from additional
specimens of spinacanthids, the previous phylogenetic hypoth¬
esis for balistoids should be reanalyzed rather than having the
new data appended to it.

On the basis of a phylogenetic analysis of data from early life
history stages, Leis (1984) tentatively placed the ostraciids in
an unresolved trichotomy with the diodontids and molids as the
sister group of the tetraodontids, which in turn were the sister
group of the balistids (including monacanthids). The placement
of the ostraciids with the tetraodontoids was based on nine
characters, but was proposed only with the caution that no data
were available for two of the included families, the Triodont-

TABLE 1.—Synopsis of the phylogenetically informative features of Eospinus daniltshenkoi , new species, in
comparison to the comparable conditions in all of the other groups of Tfetraodontiformes. (- indicates feature is
not applicable.)

* Two derived genera have long tubular snouts in which the number of teeth may be reduced, and absent entirely in the upper jaw.
X Several derived genera with teeth either compressed and rounded to truncate distally, or with reduced conical teeth.

§ A few genera with secondarily derived scales of moderate size.

00 One species with enlarged midbody scales forming a partial carapace without spines in large specimens.

NUMBER 72

7

Suborder Triacanthoidei

ORDER TETRAODONTIFORMES

Suborder Tetraodontoidei

Superfamily Triacanthoidea (triacanthoids)

Superfamily Balistoidea (balistoids)

Superfamily Tetraodontoidea (tetraodontoids)

FIGURE 4.—Cladogram of relationships within the Tetraodontiformes, based on Winterbottom (1974), except for
changes in the placement of a few fossil groups (Cryptobalistes as incertae sedis rather than a subfamily of
Triacanthodidae; Zignoichthyidae as more closely related to Tetraodontidae than to Eoplectidae) and the
recognition of families within the balistoids (aracanids, ostraciids, balistids, and monacanthids as families rather
than subfamilies).

idae and Aracanidae, the two most moiphologically primitive
of their respective lines. Although we highly value the evidence
offered by early life history stages, in this case we believe that
the 20 synapomorphies from the osteology and myology of
adults by which Winterbottom and Tyler (1983) defined the
balistid-ostraciid clade offer far stronger evidence for the
monophyly of that group than do the nine synapomorphies of
early life history stages by which Leis (1984) related ostraciids
to tetraodontoids, especially as such data were absent for
critically important groups of both ostraciids and tetraodon¬
toids.

In Rosen’s (1984) proposal of the zeiforms as the primitive
sister group of the tetraodontiforms, a rigorous analysis of the

inter-relationships of the families of the latter was not intended,
and Rosen’s placement of the ostraciids in an unresolved
trichotomy with the triodontids and other tetraodontoids based
on a single synapomorphy (dorsal fins and radials remote from
occipital region of skull) and four others, all reductive, linking
ostraciids and tetraodontoids (absence of pelvis, ribs, and
uncinate process on first epibranchial, and consolidation of the
caudal skeleton) obviously is weak in comparison to the
evidence presented by Winterbottom and Tyler (1983).

The few known and almost exclusively external characters of
Eospinus are compared below with those of spinacanthids,
balistids, and ostraciids, and with those of their triacanthoid and
tetraodontoid outgroups. The triacanthoids are, in general, the

8

SMITHSONIAN CONTRIBUTIONS TO PALEOBIOLOGY

most plesiomorphic tetraodontiforms. Among the tetraodon-
toids we compare Eospinus especially to the Eocene Eoplectus ,
the most morphologically primitive member of the Superfamily
Tetraodontoidea, and to Triodon, the most generalized extant
member of the superfamily and the sister group of all of the
other extant families of that group.

Number and Placement of Dorsal-fin Spines.—
Eospinus has three dorsal spines. The plesiomorphic condition
for tetraodontiforms is six spines, as found in all Triacanthoidei
and in Eoplectus, the most generalized representative of the
Tetraodontoidea. Within the Tetraodontoidei, reduction in the
number of dorsal spines has taken place independently in the
Balistoidea (5-6 spines in Spinacanthidae, 2-3 or rarely 1 in
Balistidae, and none in Ostraciidae) and Tetraodontoidea (6
spines in the Eocene Eoplectidae, 1-3 rudimentary spines or
none in Triodontidae, and none in the Eocene Zignoichthyidae,
Tetraodontidae, Diodontidae, and Molidae).

Winterbottom’s (1974) phylogeny of the balistoids indicates
that an a posteriori synapomorphy of the ostraciids is the
posterior migration of the basal pterygiophore of the spiny
dorsal fin toward the origin of the soft dorsal fin and the
complete loss of the spiny dorsal fin, with the pterygiophores
becoming the prominent supraneural element found just in
front of the first basal pterygiophore of the soft dorsal fin in all
ostraciids. Conversely, the sister group relationship of balistids
and spinacanthids is supported by the anterior migration of the
dorsal spines and their pterygiophores to at least the rear of the
skull, which condition is apomorphic relative to the position
just behind the rear of the skull in the triacanthoid and
eoplectid-based outgroups. Spinacanthids retained five or six
dorsal spines from the ancestral triacanthoid-eoplectid condi¬
tion and these spines migrated even further forward onto the
top of the skull over the eye and greatly increased in length
relative to the moderate length in triacanthoids and eoplectids.
The reduction of the number of dorsal spines in balistids to
three in balistins and to two (rarely one) in monacanthins is
independent of the loss of spines in ostraciids. Balistins
retained the position of the spiny dorsal fin and their supports
at the rear of the skull, plesiomorphic for the balistid-
spinacanthid clade, while monacanthins further specialized by
the migration of the spines further forward onto the top of the
skull, usually to the region of the eye or even in front of the eye.
The forward migration of the spiny dorsal fin in monacanthins
is here interpreted as independent of that in spinacanthids.

The presence of three dorsal spines in Eospinus relates it to
balistids within the spinacanthid-balistid clade, while the origin
of the spines at the rear of the skull is the ancestral condition for
that clade.

On the basis only of the number of dorsal spines and not of
their position (nor of the supraneural in their absence),
spinacanthids with five or six spines would be the sister group
of both balistids and ostraciids, with three or fewer spines being
a synapomorphy for a balistid-ostraciid clade and the complete
loss of the dorsal spines a synapomorphy of ostraciids.

It is clear that the position of the spiny dorsal fin was
anterior, toward the rear of the skull, in the ancestral balistoid.
One could hypothesize that the spines were lost in ostraciids
while still in this position and that only then did the basal
pterygiophores, having lost their original function of spine
support, migrate posteriorly to the origin of the soft dorsal fin
to become the supraneural. However, in triodontids, which are
the only tetraodontiforms with a rudimentary spiny dorsal fin,
the spiny dorsal fin and its pterygiophores have migrated
posteriorly to just in front of the soft dorsal fin. Moreover,
when the rudimentary spiny dorsal fin is absent (Indian Ocean
populations) in the only Recent species, Triodon macropterus
Lesson, the basal pterygiophores are entirely comparable to the
posteriorly placed supraneural of ostraciids. This analogous
positioning of the pterygial supports of the absent spiny dorsal
fin in triodontids and ostraciids is in conformity with
Winterbottom’s (1974) contention of the sister group relation¬
ship of spinacanthids and balistids, both of which have anterior
migration of the spiny dorsal fin, and the sister group
relationship of that clade with the ostraciids, which have
posterior migration of the pterygiophores and loss of the spines.
This scenario also is consistent with Rosen’s (1984) hypothesis
of ostraciid relationships.

Eospinus lacks a mechanism by which the second dorsal
spine locks the first spine in an erect position. Such a complex
specialized locking mechanism is present among tetraodonti¬
forms (and all other fishes) only in the Balistidae (present in all
species except for a few highly specialized monacanthins that
have lost the second spine). If Eospinus is more closely related
to balistids than to other balistoids this would indicate that the
reduction in the number of dorsal spines to three in Eospinus
and balistids took place in that lineage prior to the development
of the locking mechanism in balistids alone, in which case
Eospinus would be the sister group of balistids.

Length of Dorsal-fin Spines. — Eospinus has exception¬
ally long and robust dorsal spines, the first being about 73% SL
and the second 63% SL. The plesiomorphic condition for
tetraodontiforms is a moderate dorsal-spine length of 17%-
43% SL as found in the Triacanthoidei (for specimens over 60
mm SL; smaller specimens with somewhat longer spines,
Tyler, 1968). In the only group of Tetraodontoidea with a
relatively well-developed spiny dorsal fin, the primitive Eocene
Eoplectus (Tyler, 1973b), the spines are incompletely pre¬
served, but from the shape and size of the parts that remain the
spines were obviously of only moderate length and probably
somewhat shorter than in triacanthoids. In the Balistidae the
longest (first) dorsal spine is plesiomorphically moderate in
length, usually between 20% and 30% SL, and always less than
40% SL. In the Spinacanthidae the dorsal spines are exception¬
ally long and strong in both of the monotypic genera, 54% and
92% SL in Protobalistum imperiale (Massalongo) and
Spinacanthus cuneiformis (de Blainville), respectively. The
greatly increased length of the dorsal spines is a synapomorphy

NUMBER 72

9

of the two genera of Spinacanthidae, as is their origin far
forward on the head over the small and highly placed eye.

A hypothesis of moderate dorsal-spine length as found in
triacanthoids and eoplectids being ancestral for balistoids and
increased length acquired by spinacanthids and Eospinus (one
step if spinacanthids and Eospinus are sister groups or two
steps if balistids and Eospinus are sister groups) is more
parsimonious than increased length being ancestral for balis¬
toids and there having been reduction in the length of the spine
in balistoids and ostraciids (three steps). Within the balistoids,
however, it is equally parsimonious to assume that the
spinacanthid-balistid ancestor had increased spine length that
was reversed in balistids or that this ancestor had moderate
length and that spinacanthids and Eospinus independently
acquired increased length if, as evidenced by the number of
dorsal spines, Eospinus and balistids are sister groups (two
steps in both scenarios).

The evidence from the 73% SL dorsal-spine length in
Eospinus indicates relationship with spinacanthids, while the
evidence from the number of spines indicates relationship with
balistids. If Eospinus is proposed as the sister group of
spinacanthids on the basis of dorsal-spine length, then the
number of spines had to be reduced from the ancestral six to
three or fewer independently in Eospinus and balistids (two
steps). If Eospinus is proposed as the sister group of balistids on
the basis of the number of spines, then the size of the spines had
to be increased from the ancestral moderate size to large
independently in Eospinus and spinacanthids (two steps). None
of the other characteristics of Eospinus supports one versus the
other of these equally parsimonious alternatives.

Pelvic Fin. — Eospinus has the pelvic fin reduced to a pair of
partially fused and relatively large rudimentary spines placed
posteriorly at the end of the pelvis. The plesiomorphic
condition for tetraodontiforms is a pelvic fin of one spine and
two (triacanthoids) to four (eoplectids) rays placed thoracically
along the middle of the length of the pelvis below the pectoral
fin (condition of pelvis unknown in eoplectids but pelvic fin
thoracic). Among the alternate outgroups for tetraodontiforms,
most acanthuroids and some zeiforms have a pelvic fin with a
spine and five well-developed rays (in some zeiforms up to
seven rays if a spine is present or up to 10 rays if the spine is
absent; rays reduced to three or four in some specialized
acanthuroids). The pelvic fin in zeiforms and acanthuroids is
like that of morphologically primitive tetraodontiforms in
being placed thoracically along the side of the pelvis, which
usually has a prominent posterior process behind the fin; i.e.,
the fin is never at the extreme posterior end of the pelvis as in
balistids.

The pelvic fin of tetraodontiforms has become much
reduced, apparently independently, in both lineages of the
Tetraodontoidei. In the Tetraodontoidea the pelvic fin is
completely lost in all groups except the Eoplectidae, although
the Triodontidae retain the pelvis. In the Balistoidea the pelvic
fin is either reduced to a complex rudimentary structure at the

end of the pelvis in front of the anus or absent altogether. In the
Balistidae the pelvic fin rudiment is thought to be the partial
fusion product of the pelvic spines from both sides of the body.
Except at its distal end, this rudimentary fused spine is
surrounded by a series of enlarged encasing scales (in balistins
and most monacanthins, while the rudiment and its encasing
scales are secondarily further reduced or completely lost in
some of the more specialized monacanthins, even though the
pelvis is always present: Tyler, 1962; Matsuura, 1979). In the
Ostraciidae both the pelvis and pelvic fin are completely lost. In
the Eocene Spinacanthidae there is no clear evidence in either
of the single specimens of the two species of a pelvic fin or
pelvis, and the pelvic fin must be considered to have been either
extremely rudimentary or entirely lost.

The spiny structure at the end of the pelvis in Eospinus is less
specialized than that of balistids because of its larger or less
apomorphically reduced size and its lack of development of the
specialized series of encasing scales that surround the rudiment
in the more primitive balistids (subsequently lost by some
specialized monacanthins). The spiny processes along the
lateral edges of the reduced and partially fused pelvic spines in
Eospinus also can be interpreted as plesiomorphic relative to
the smooth edges of the smaller rudiment in balistids because
such spiny processes are present on the pelvic spines of both the
more generalized triacanthoid (the triacanthodids) and eoplec-
tid-based outgroups.

In assessing the condition of the pelvic fin in balistoids we
presume that there is an ordered, polarized, transformation
series from the ancestral condition of a relatively well-
developed pelvic fin with a spine and at least a few rays
(triacanthoids, eoplectids, extraordinal outgroups) to a progres¬
sively smaller rudimentary fin before it is entirely lost. In this
scenario the pelvic fin became reduced to a large rudiment in
the ancestral balistoid, a condition retained by Eospinus, while
balistids further reduced the size of the rudiment and ostraciids
lost it. Given the frequency of reductive tendencies among
many lineages of tetraodontiforms in both pelvic- and
dorsal-fin spines, with the numbers and size progressively
reduced and then eventually lost, this scenario seems far more
reasonable to us than to assume that the ancestral balistoid
entirely lost the pelvic fin and it was regained to much different
degrees by balistids and Eospinus.

Any further analysis of the pelvic fin in balistoids is severely
limited by the unknown condition in spinacanthids. Neverthe¬
less, on the basis only of its relatively large and ornamented
rudimentary pelvic fin Eospinus is the sister group of a clade
including at least balistids and ostraciids (and spinacanthids if
their pelvic fin is no more than a small rudiment, if not absent
entirely). The ancestor of such a clade would have had the
pelvic fin further reduced to a small rudiment, which was
retained by balistids and lost by ostraciids, while the balistids
developed a complex specialized series of encasing scales
around the rudiment.

Without knowing the condition of the pelvic fin in

10

SMITHSONIAN CONTRIBUTIONS TO PALEOBIOLOGY

spinacanthids we cannot assess the implications of the
placement of the rudimentary pelvic fin at the posterior end of
the pelvis just in front of the anus in Eospinus and balistids,
versus thoracically placed in all other tetraodontiforms with a
pelvic fin (triacanthoids and eoplectids) and in the extraordinal
outgroups. If one presumes that the loss of the pelvic fin in
ostraciids took place after its migration posteriorly to the end of
the pelvis, as was probably the comparable situation with the
posterior migration of the spiny dorsal fin associated with its
loss, then the position of the pelvic-fin rudiment in Eospinus
and balistids is plesiomorphic for balistoids.

We note that if Eospinus had encasing scales around the
rudimentary pelvic spine we would accept this as important
evidence linking it with the balistids (just as we would if
Eospinus had a locking mechanism between the first and
second dorsal spines). We believe it unlikely that the loss of the
pelvic fin in ostraciids, even if it involved migration to the
posterior end of the pelvis, proceeded through a stage in which
the rudimentary spine became encased in the specialized
series of flexible scale segments found in balistids. This
complex apparatus in balistids serves variously to wedge the
species into crevices and to flare a dewlap of skin to increase
apparent body size in a presumably protective-defensive
behavior, and the development of this apparatus has apparently
been a key factor in the diversification of the group. We doubt
that ostraciids also developed this complex pelvic apparatus
only to subsequently lose it before their probable pre-Eocene
origins (the ostraciin Eolactoria and the aracaniin Proaracana,
both without a pelvis or pelvic fin of any kind, being known
from the early Eocene). The balistid pelvic apparatus (and the
dorsal-spine locking mechanism) is fully developed in all of the
fossil balistids. Balistins are first recorded from the Oligocene
by the three species of Balistomorphus from Switzerland and
the monotypic Oligobalistes from Russia (Tyler, 1980), while
monacanthins are first known from the Pliocene by several
species of Aluterus and Cantherhines reported by Sorbini
(1987, 1991) from Italy.

Carapace Scale Plates. — Eospinus has enlarged tubercu-
late scale plates over the entire body, the largest of which are
between 12% and 13% SL and tend to form a continuous
carapace around the middle of the body, although the plates are
not as regularly hexagonal or extensively sutured to one
another as in ostraciids. In ostraciids the mostly hexagonal
tuberculate scale plates are firmly sutured to one another to
form a rigid carapace over most of the body, with scale plates
of up to 13% SL (larger in the Eocene Proaracana, up to 24%
SL).

In spinacanthids one genus, Spinacanthus, has moderately
enlarged scales of 4.8% SL, while the other, Protobalistum, has
greatly enlarged scales of 12.3% SL, and in both cases the
scales bear numerous tubercles on the surface, like those of
Eospinus and the ostraciids. In Spinacanthus the plates are
more or less isolated and only cover a total of about one-fourth
of the surface of the body, without much contact between the

plates except ventrally. In Protobalistum, however, the scale
covering is much more complete, like that of Eospinus. In the
anterior half of the body (exclusive of the head) of Protobalis¬
tum many of the plates are distinctly hexagonal and articulate
with one another by interdigitations, just as in ostraciids. This
is similar to the situation in Eospinus, in which the greatest
consolidation of the scale plates and their closest articulation to
one another is in the anterior half of the body below the spiny
dorsal fin and in front of the pelvic spine and the posteriorly
oriented carapace spine.

Scale size in balistids is not well quantified in the literature.
Therefore, we made a survey of the size of the scales from the
side of the body (i.e., excluding the sometimes larger and more
elongate scales at the rear of the abdomen associated with the
expansible dewlap) from specimens in the collections of the
National Museum of Natural History. Examination of 21
species of balistin triggerfishes (the more generalized of the
two groups of balistids), representing nearly all of the genera,
shows that the greatest dimension of the scales is usually
between 2%-3% (range l%-4%) SL, but as much as 5% SL in
Pseudobalistes flavomarginatus (Riippell) and 7% SL in
Odonus niger (Riippell). The thick scales in the aberrant
Xenobalistes (enlarged coracoid, postcleithra, and frontal),
which has some superficial ostraciid characteristics, are about
3.3% SL (Matsuura, 1981:194). The scales in the Oligocene
genera Balistomorphus and Oligobalistes are only impressions
that are difficult to measure, but from their pattern they
obviously were small and in the size range of 2%-3% SL like
those of most Recent balistins. The scales of monacanthin
filefishes tend to be slightly smaller and thinner than those of
balistins, and to bear upright spinules, while in some
specialized species the scales are almost entirely lost. Thus, the
scales in balistids are small, except in a few derived genera of
balistins with scales of moderate size, even though the basal
scale plate may be thicker in balistins than in triacanthoids.

Among the Tetraodontoidea, the monotypic Eocene Zigno-
ichthys (Zignoichthyidae) has small scales (0.9% SL) with
upright spinules very similar to those of the Triacanthodidae,
the most generalized family of the order, while the monotypic
Eocene Eoplectus (Eoplectidae) has moderately enlarged scales
(5.8% SL) with stellate radiations from the base of a central
upright spinule. In the Triodontidae, another relatively general¬
ized family of tetraodontoids, the scales are relatively small
(3% SL) except where more elongate in the expansible dewlap
of the abdomen. Elsewhere in the Tetraodontoidea the scales
often are enlarged as either fixed or erectable spines in most
Tetraodontidae and Diodontidae, except where the spines are
secondarily reduced or lost in some tetraodontids. However,
enlarged plate-like scales similar to those of ostraciids are
found in the tetraodontid Ephippion and in the molid Ranzania.
The scale plates of Ephippion are up to 9.2% SL in large adults
and form a firmly sutured girdle around the middle of the body
(more extensively so than in Eospinus, although the latter has
a girdle at a much smaller body size than does Ephippion). The

NUMBER 72

11

scale plates of Ranzania are relatively thin and small, up to
1.2% SL, but form a complete, if somewhat flexibly sutured,
carapace around the entire body. These two genera of different
families of the Tetraodontoidea apparently have developed a
carapace-like structure independently of one another and of
Eospinus, Protobalistum , and the ostraciids among the Balis-
toidea.

In the triacanthoids, which overall are the most morphologi¬
cally plesiomorphic group of tetraodontiforms, the scales are
relatively small (l%-3% SL), with either upright spinules
(triacanthodids) or emarginate ridges (triacanthids), including
the only Eocene representative, Protacanthodes. Because
triacanthoids have small scales, the moderately enlarged
plate-like scales of some spinacanthids ( Spinacanthus ), balis-
tids ( Pseudobalistes and Odonus), and eoplectids ( Eoplectus)
and the greatly enlarged plates of Eospinus, ostraciids, and
some spinacanthids ( Protobalistum ) and tetraodontids ( Ephip -
pion) are derived conditions. It seems clear that many of these
plate-like enlargements are independent acquisitions (e.g., in
tetraodontids and balistids; as are the smaller plate-like scales
forming the carapace in the molid Ranzania). Enlarged scales,
as either plates or spines, are of such frequent if phylogeneti-
cally sporadic occurrence among spinacanthids, eoplectids,
ostraciids, tetraodontids, and diodontids, that it also seems clear
that scale enlargement is an especially homoplastic feature in
tetraodontiforms (as is extreme reduction and loss of scales).
Nevertheless, the small scales of triacanthoids must be
considered the primitive condition, and enlargement apomor-
phic.

Given that small scales are primitive for the Triacanthoidei
(all species with small scales) and Tetraodontoidea (small
scales in morphologically primitive forms like the Eocene
Zignoichthys and the Eocene to Recent Triodon, although
moderate size scales in the Eocene Eoplectus ), it is equally
parsimonious (3 steps) to assume that the scales in the ancestral
balistoids were of either large or small size, but not of moderate
size (4 steps). If the balistoid ancestral line had small scales like
triacanthoids, then this requires changes for the moderate scale
size in Spinacanthus, the large scale size in Protobalistum , and
the large scale size of ostraciids. If the balistoid ancestral line
had large scales, then this requires changes from the small scale
size of triacanthoids, the moderate scale size of Spinacanthus,
and the small scale size of nearly all balistids.

Our preferred hypothesis is that the ancestral balistoids had
small scales like the triacanthoids and some of the more
primitive tetraodontoids, simply because we have no particular
reason to believe that a large-scaled balistoid ancestral line is
more probable than the proposed small-scaled ancestry. If one
accepts small scales as ancestral to balistoids, then the
enlargement of scales in Eospinus relates it to either ostraciids
or spinacanthids ( Protobalistum ) in the accepted phylogeny or
indicates, contrary to other evidence, that spinacanthids and
ostraciids are sister groups to which Eospinus has an
unresolved relationship. None of these alternatives are refuted

by the possibility that the ancestral balistoids had large scales,
in which case the large scales of Eospinus are phyletically
uninformative within the balistoids.

In Protobalistum the partial carapace has a ventrolateral
ridge, similar to the ridges in this position in most ostraciids
and in Eospinus. Just as with scale enlargement, we postulate
that the carapace ridge in Protobalistum is an independent
acquisition from that in Eospinus and ostraciids. Thus, ridges
are interpreted as a common functional aspect of carapace
construction irrespective of the phylogenetic origin of the
carapace.

Carapace Spines. — Eospinus has a large median carapace
spine directed forward from the snout and another fixed spine
directed posteriorly from the middle of each side of the body.
No such spines are present in spinacanthids and balistids
(among the latter neither the flexible quill-like spines found on
the side of the body of males of the monacanthin Amanses
scopas (Cuvier) nor the enlarged tubercles and spiny processes
borne on the scales of the caudal peduncle in many balistins and
monacanthins seem homologous to the fixed carapace spines of
Eospinus). Among ostraciids only the Eocene Eolactoria has a
median, but much shorter, carapace spine on the snout, in the
same position as that of Eospinus (the Recent Ostracion
rhinorhynchus Bleeker has a blunt, rounded enlargement on the
snout, but not a spine). Bilateral carapace spines are present in
many ostraciids, either on the supraorbital ridge or along the
dorsolateral and ventrolateral ridges of the carapace, but none
has a large spine projecting posteriorly from the middle of each
side of the body.

While the median snout spines of Eospinus and Eolactoria
could be considered a synapomorphy, we discount the
significance of this evidence because of the great degree of
homoplasy in carapace spines in ostraciids. For example,
seemingly identical large carapace spines bilaterally in front of
the eyes have been acquired independently by genera in two
different phyletic lines of Recent ostraciins, and this is also
homoplastic to the occurrence of such preorbital spines in the
Oligocene Oligolactoria (Tyler and Gregorova, 1991). More¬
over, the numerous spiny scale processes of pelagic larval
molids, which apparently act as flotation and defensive devices,
include a long anteriorly projecting median snout spine and one
projecting posteriorly from each side of the middle of the body
(especially in Masturus). These are similar in size and
placement to those of Eospinus but they are not associated with
a carapace. As molids are one of the most derived groups of
tetraodontoids and more closely related to tetraodontids and
diodontids than to ostraciids, it is reasonable to assume that the
molid snout and body spines have been acquired independently
from those of ostraciids. Because there is no other evidence of
a close relationship between Eospinus and Eolactoria, we
postulate that the snout spines in these two genera (and in larval
molids) are homoplastic, and that the presence of carapace
spines of any kind in Eospinus is not evidence of relationship
with ostraciids.

12

SMITHSONIAN CONTRIBUTIONS TO PALEOBIOLOGY

Teeth. —Eospinus has a moderate number (6 to 8 on each
side) of basically conical teeth; those of the lower jaw are
stoutly conical and larger than those above, and some of the
latter are constricted distally. The plesiomorphic condition for
tetraodontiforms is relatively numerous (modally 12 or more
on each side), simple, conical teeth without distal constrictions
and of equal (small to moderate) size in both jaws. This is the
condition found in all species of the Triacanthodidae except for
several derived genera with thin, wide, and distally rounded or
truncate scraping teeth (sometimes reduced in number or absent
in the upper jaw) and/or long tubular snouts. In the Triacanth-
idae, the more derived of the two families of Triacanthoidei, the
teeth are larger, more incisor-like, and reduced in number to
four or five on each side in an outer series and one or two in an
inner series on each side.

Within the Tetraodontoidei, all Tetraodontoidea have the
derived condition of the teeth incorporated with the jaw bones
into a specialized beak, while the Balistoidea have teeth that
vary from conical to heavy incisors, as follows. Of the two
species of Spinacanthidae, Spinacanthus cuneiformis has about
seven stoutly conical teeth on each side of the lower jaw (the
upper jaw teeth are less visible in the single specimen), some of
which taper distally into a strongly constricted blunt nipple. In
the other species, Protobalistum imperiale , there are four or
five teeth on each side of both jaws, some with straight or
rounded distal edges as in incisors or molars, but others with a
distinct blunt nipple-like cusp. In the Balistidae the teeth are
relatively few, large, compressed, and more-or-less notched
and incisor-like, three to four in an outer series on each side,
and two or three in an inner series in the upper jaw and two to
four in a single series in the lower jaw. In the Ostraciidae the
teeth are of moderate size and more or less conical but usually
constricted distally, often into a blunt nipple, with between
three and nine (usually four to six) in a single series on each
side of both jaws. In the two Eocene genera of ostraciids,
Eolactoria and Proaracana , the teeth are somewhat larger,
heavier, and more notched distally than in the Recent species
and bear at least some similarity to the outer teeth of balistids
(Tyler, 1973a).

Relative to the simple conical teeth of triacanthodids, the
teeth of all three families of balistoids have a derived condition
of being enlarged to some degree, less numerous, and in
various ways notched, distally constricted, or incisor-like.
Within the balistoids, the basically conical, distally constricted
teeth of ostraciids and spinacanthids are less derived than are
the larger, less numerous, incisor-like teeth in the outer series of
balistids. The teeth of Eospinus are more similar to those of
spinacanthids and ostraciids than to those of balistids in that
they are basically conical and some of them are distally
constricted. However, the lower jaw teeth of Eospinus are
enlarged, being at least as large as those in the two Eocene
genera of ostraciids and not much smaller than those of
balistids; and some of the teeth in the spinacanthid Protobalis¬
tum are incisor-like.

Thus, there are significant differences in the size and shape
of the teeth within each of the families of balistoids, and even
within each of the species of the two monotypic genera of
spinacanthids, Protobalistum and Spinacanthus. Given this
degree of ambiguity, we interpret the teeth of Eospinus as
simply relating it to the balistoids but not to a particular family
within it. The dentition of Eospinus is best considered as
modified for its function in a unique mouth in which the upper
jaw teeth are only half as long as and more distally constricted
than those in the lower jaw.

Plesiomorphic Features. — Eospinus has 12 principal
caudal-fin rays as also found in triacanthoids, spinacanthids,
balistids, eoplectids, triodontids, and zignoichthyids rather than
the derived condition of 11 or less found in ostraciids,
tetraodontids, diodontids, and molids. In Eospinus the eye is of
moderate size and placed in about the middle of the upper half
of the head, as it is also in triacanthoids and most other
tetraodontiforms. The eye is somewhat more elevated in some
balistids and in most ostraciids, but these never have the eye so
small and highly placed just under the spiny dorsal-fin origin as
in spinacanthids, the latter size and placement being a
synapomorphy of Spinacanthus and Protobalistum. In Eo¬
spinus the soft dorsal and anal fins are not preserved, but based
on the space available for them, they obviously were
short-based, as in triacanthodids (the most morphologically
generalized tetraodontiforms), ostraciids, and most Tetraodont¬
oidea (except moderate in the Eocene Eoplectus and secondar¬
ily elongate in Chonerhinos and Xenopterus). In triacanthids
(including the Eocene protacanthodins) the soft dorsal fin is
secondarily elongate, while in balistids both the soft dorsal and
anal fin are apomorphically elongate.

Phylogenetic Conclusion

The structure of the spiny dorsal fin, pelvic fin, scale plates,
and teeth in Eospinus are all phylogenetically informative but
some of the evidence conflicts with the accepted phytogeny of
the balistoids and leaves the exact relationships of Eospinus
uncertain.

The presence of three dorsal spines relates Eospinus to
balistids, while the lack of a locking mechanism excludes
Eospinus from the balistids as presently defined and places it as
the sister group of the balistids. The position of the spiny dorsal
fin over the rear of the skull relates Eospinus to spinacanthids
and balistids. The increased length of the dorsal spines relates
Eospinus to spinacanthids. The presence of a relatively large
rudimentary pelvic spine at the end of the pelvis without
surrounding encasing scales relates Eospinus to balistoids as
this is the presumed ancestral condition of reduction in that
clade but, without the condition of the pelvic fin being known
in spinacanthids, it is not possible to further use this character
to establish relationships within the clade. The shape of the
teeth (conical with distal constrictions) relates Eospinus to

NUMBER 72

13

balistoids but not to any particular family because such teeth
are plesiomorphic for balistoids. The enlargement of the scale
plates into a partial carapace either relates Eospinus to both
spinacanthids and ostraciids or is primitive for balistoids.

Although we believe that the evidence presented here
abundantly supports our contention that Eospinus is a balistoid,
we cannot determine on the basis of its few external characters
whether Eospinus is more closely related to balistids or to
spinacanthids. Given the tenuousness of the evidence from
Eospinus and the weakness of the cladogram at the spinacan-

thid-balistid node, we prefer not to make a formal rearrange¬
ment of these taxonomic groupings in order to accommodate
Eospinus either in an expanded concept of the balistids or as a
categorically equivalent sister group of them or of spinacan¬
thids or of both. Rather, we adopt a conservative course and
simply place Eospinus incertae sedis with the spinacanthid-
balistid clade of balistoids.

We are confident that the exposure of even a few internal
osteological features in future specimens of Eospinus will
allow for its precise placement among the balistoids.

Literature Cited

Bannikov, A.F.

1985. Iskopayemye skumbriyevye SSSR [Fossil Scombrids of the USSR].
Trudy Paleontologicheskogo lnstituta AN SSSR, 210:1-111. [In
Russian.]

Berg, L.

1940. Classification of Fishes, Both Recent and Fossil. Travaux de
I'Institut Zoologique de I'Academie des Sciences de I’URSS,
5:87-517.

Danilchenko, P.G.

1968. Ryby verkhnego paleotsena Tlirkmenii [Fishes of the Upper
Paleocene of Dirkmenia], In D.V. Obruchev, editor, Ocherki po
filogenii i sistematike iskopayemykh ryb i beschelyustnykh, pages
113-156. Moscow: Nauka. [In Russian.]

Grossheim, V.A., and I.A. Korobkov, editors

1975. Stratigrafia SSSR: Paleogenovaya sistema [Stratigraphy of the
USSR: Paleogene System], 524 pages. Moscow: Nedra. [In
Russian.]

Leis, J.M.

1984. Tetraodontiformes: Relationships. In H.G. Moser, W.J. Richards,
D.M. Cohen, M.P. Fahay, A.W. Kendall, Jr., and S.L. Richardson,
editors. Ontogeny and Systematics of Fishes. American Society of
Ichthyologists and Herpetologists, Special Publication, 1:459-463.

Matsuura, K.

1979. Phylogeny of the Superfamily Balistoidea (Pisces: Tfetraodonti-
formes). Memoirs of the Faculty of Fisheries, Hokkaido University,
26(1—2):49—169.

1981. Xenobalistes tumidipectoris, a New Genus and Species of Trigger-
fish (Tetraodontiformes, Balistidae) from the Marianas Islands.
Bulletin of the National Science Museum, Tokyo, series A,
7(4): 191-200.

Rosen, D.E.

1984. Zeiforms As Primitive Plectognath Fishes. American Museum
Novitates, 2782:1-45.

Solun, V.I., editor

1975. Paleogen Turkmenii: Biostratigraficheskoye raschleneniye i korre-
latsia [The Paleogene of Tlirkmenia: Biostratigraphical Dismember¬
ment and Correlation]. 195 pages. Ashkhabad: Ylym. [In Russian.]

Sorbini, L.

1987. Biogeography and Climatology of Pliocene and Messinian Fossil
Fish of Eastern-central Italy. Bollettino del Museo Civico di Storia
Naturale di Verona, 14:1-85.

1991. The Survival of Pliocene Ichthyofauna of Tfethyan Origin in Central
Italy. Saito Ho-on Kai Special Publication, 3:459-465.

Tyler, J.C.

1962. The Pelvis and Pelvic Fin of Plectognath Fishes; a Study in
Reduction. Proceedings of the Academy of Natural Sciences of
Philadelphia, 114(7):207-250.

1968. A Monograph on Plectognath Fishes of the Superfamily Triacantho-
idea Academy of Natural Sciences of Philadelphia, Monograph,
16:1-364.

1973a. A New Species of Boxfish from the Eocene of Monte Bolca Italy,
the First Unquestionable Fossil Record of the Ostraciontidae. Museo
Civico di Storia Naturale di Verona, Studi e Ricerche sui Giacimenti
Terziari di Bolca, 2:103-127.

1973b. A New Species of Triacanthodid Fish (Plectognathi) from the
Eocene of Monte Bolca, Italy, Representing a New Subfamily
Ancestral to the Triodontidae and the Other Gymnodonts. Museo
Civico di Storia Naturale di Verona, Studi e Ricerche sui Giacimenti
Terziari di Bolca, 2:128-156.

1980. Osteology, Phylogeny, and Higher Classification of the Fishes of the
Order Plectognathi (Tfetraodontiformes). National Oceanic and
Atmospheric Administration Technical Report, National Marine
Fisheries Service Circular, 434:1-422.

Tyler, J.C., and R. Gregorova

1991. A New Genus and Species of Boxfish (Ostraciidae, Tfetraodonti¬
formes) from the Oligocene of Moravia, the Second Fossil
Representative of the Family. Smithsonian Contributions to Paleo¬
biology, 71:1-20.

Winterbottom, R.

1974. The Familial Phylogeny of the Tfetraodontiformes (Acanthopterygii:
Pisces) as Evidenced by Their Comparative Myology. Smithsonian
Contributions to Zoology, 155:1-201.

Winterbottom, R., and J.C. "fyler

1983. Phylogenetic Relationships of Aracanin Genera of Boxfishes
(Ostraciidae: Tfetraodontiformes). Copeia, 1983(4):902-915.

14

REQUIREMENTS FOR SMITHSONIAN SERIES PUBLICATION

Manuscripts intended for series publication receive substantive
review (conducted by their originating Smithsonian museums or
offices) and are submitted to the Smithsonian Institution Press
with Form SI-36, which must show the approval of the appropriate
authority designated by the sponsoring organizational unit. Re¬
quests for special treatment—use of color, foldouts, case-bound
covers, etc.—require, on the same form, the added approval of
the sponsoring authority.

Review of manuscripts and art by the Press for requirements
of series format and style, completeness and clarity of copy, and
arrangement of all material, as outlined below, will govern, within
the judgment of the Press, acceptance or rejection of manuscripts
and art.

Copy must be prepared on typewriter or word processor,
double-spaced, on one side of standard white bond paper (not
erasable), with IV 4 " margins, submitted as ribbon copy (not
carbon or xerox), in loose sheets (not stapled or bound), and
accompanied by original art. Minimum acceptable length is 30
pages.

Front matter (preceding the text) should include: title page
with only title and author and no other information, abstract page
with author, title, series, etc., following the established format;
table of contents with indents reflecting the hierarchy of heads in
the paper; also, foreword and/or preface, if appropriate.

First page of text should carry the title and author at the top
of the page; second page should have only the author’s name
and professional mailing address, to be used as an unnumbered
footnote on the first page of printed text.

Center heads of whatever level should be typed with initial
caps of major words, with extra space above and below the head,
but no other preparation (such as all caps or underline, except for
the underline necessary for generic and specific epithets). Run-in
paragraph heads should use period/dashes or colons as neces¬
sary.

Tabulations within text (lists of data, often in parallel columns)
can be typed on the text page where they occur, but they should
not contain rules or numbered table captions.

Formal tables (numbered, with captions, boxheads, stubs,
rules) should be submitted as carefully typed, double-spaced copy
separate from the text; they will be typeset unless otherwise
requested. If camera-copy use is anticipated, do not draw rules
on manuscript copy.

Taxonomic keys in natural history papers should use the
aligned-couplet form for zoology and may use the multi-level
indent form for botany. If cross referencing is required between
key and text, do not include page references within the key, but
number the keyed-out taxa, using the same numbers with their
corresponding heads in the text.

Synonymy in zoology must use the short form (taxon, author,
yearpage), with full reference at the end of the paper under
“Literature Cited." For botany, the long form (taxon, author,
abbreviated journal or book title, volume, page, year, with no
reference in “Literature Cited”) is optional.

Text-reference system (author, year:page used within the text,
with full citation in “Literature Cited” at the end of the text) must
be used in place of bibliographic footnotes in all Contributions
Series and is strongly recommended in the Studies Series:
“(Jones, 1910:122)" or “...Jones (1910:122)." If bibliographic

footnotes are required, use the short form (author, brief title, page)
with the full citation in the bibliography.

Footnotes, when few in number, whether annotative or biblio¬
graphic, should be typed on separate sheets and inserted imme¬
diately after the text pages on which the references occur. Exten¬
sive notes must be gathered together and placed at the end of
the text in a notes section.

Bibliography, depending upon use, is termed “Literature Cited,"
“References," or “Bibliography." Spell out titles of books, articles,
journals, and monographic series. For book and article titles use
sentence-style capitalization according to the rules of the lan¬
guage employed (exception: capitalize all major words in English).
For journal and series titles, capitalize the initial word and all
subsequent words except articles, conjunctions, and prepositions.
Transliterate languages that use a non-Roman alphabet according
to the Library of Congress system. Underline (for italics) titles of
journals and series and titles of books that are not part of a series.
Use the parentheses/colon system for volume (number):
pagination: “10(2):5-9." For alignment and arrangement of ele¬
ments, follow the format of recent publications in the series for
which the manuscript is intended. Guidelines for preparing bibli¬
ography may be secured from Series Section, SI Press.

Legends for illustrations must be submitted at the end of the
manuscript, with as many legends typed, double-spaced, to a
page as convenient.

Illustrations must be submitted as original art (not copies)
accompanying, but separate from, the manuscript. Guidelines for
preparing art may be secured from Series Section, SI Press. All
types of illustrations (photographs, line drawings, maps, etc.) may
be intermixed throughout the printed text. They should be termed
Figures and should be numbered consecutively as they will
appear in the monograph. If several illustrations are treated as
components of a single composite figure, they should be desig¬
nated by lowercase italic letters on the illustration; also, in the
legend and in text references the italic letters (underlined in copy)
should be used: “Figure 9b.” Illustrations that are intended to
follow the printed text may be termed Plates, and any components
should be similarly lettered and referenced: “Plate 9p." Keys to
any symbols within an illustration should appear on the art rather
than in the legend.

Some points of style: Do not use periods after such abbrevi¬
ations as “mm, ft, USNM, NNE.” Spell out numbers "one" through
“nine" in expository text, but use digits in all other cases if possible.
Use of the metric system of measurement is preferable; where
use of the English system is unavoidable, supply metric equiva¬
lents in parentheses. Use the decimal system for precise meas¬
urements and relationships, common fractions for approximations.
Use day/month/year sequence for dates: “9 April 1976.” For
months in tabular listings or data sections, use three-letter abbre¬
viations with no periods: “Jan, Mar, Jun,” etc. Omit space between
initials of a personal name: “J.B. Jones."

Arrange and paginate sequentially every sheet of manu¬
script in the following order: (1) title page, (2) abstract, (3) con¬
tents, (4) foreword and/or preface, (5) text, (6) appendixes, (7)
notes section, (8) glossary, (9) bibliography, (10) legends, (11)
tables. Index copy may be submitted at page proof stage, but
plans for an index should be indicated when manuscript is sub¬
mitted.

mm

tof V73Py

^.. ■y>,jS *X

r®Svl

^-4

&Bfi , fjf . y Jtr r ^i^rL

SSPk'* 1

> •>

- •*.•' " •-- '3wj?;

s ^,^ iL v . , AJ®>'

\ TV

i 4jK-

*■'■ Ir'*

’-_ f * Va^

' S* - I ^75r.':» •* - i/V

l\

'' Vr V*~ ' ,y $R

ItV -\v;*5> - \ i