A New Species Of The Luvarid Fish Genus Avitoluvarus (Acanthuroidei : Perciformes) From The Eocene Of The Caucasus In Southwest Russia

10 Pages·2001·4.1 MB·

by  A F Bannikov

PROCEEDINGS OF THE BIOLOGICAL SOCIETY OF WASHINGTON
114(3):579-588. 200L
A new species of the luvarid fish genus "fAvitoluvarus (Acanthuroidei:
Perciformes) from the Eocene of the Caucasus in southwest Russia
Alexandre F. Bannikov and James C. Tyler
(AFB) Paleontological Institute, Russian Academy of Sciences, Profsoyuznaya 123,
Moscow 117647, Russia;
(JCT) National Museum of Natural History, Smithsonian Institution,
Washington, D.C. 20560-0106, U.S.A.
—
Abstract. A new species ofthe luvarid fish genus '\Avitoluvarus, A. eocaen-
icus, is described from the middle Eocene of Russia (Kuma Horizon, North
Caucasus) based on a single imprint of its skeleton (incomplete posteriorly).
Avitoluvarus previously has been known only from the terminal (uppermost)
Paleocene of Turkmenistan, where it is represented by two species, A. dianae
and A. mariannae.
The fossil luvarid genus Avitoluvarus signed to the Upper Eocene (Tyler & Ban-
was described recently (Bannikov & Tyler nikov 1992a, Bannikov 1993) but according
1995) from the Danata Formation in Turk- to Cavelier & Pomerol (1986) only the Pria-
menistan, where it is represented by two bonian (but not the Bartonian) should be
&
species, A. dianae Bannikov Tyler, 1995, included in the Upper Eocene, and thus the
and A. mariannae Bannikov & Tyler, 1995. Kuma Horizon fishes are oflate Middle Eo-
The exact age of the fish-bearing layer of cene age.
the Danata Formation in Turkmenistan has A preliminary, and far from complete,
been questioned, with the fishes from this list of Kuma fishes (Bannikov & Parin
layer dated as both Late Paleocene (Danil- 1997) includes at least 27 species repre-
chenko 1968, Bannikov 1985) and Early senting 25 families in 10 orders. Based on
&
Eocene (Tyler Bannikov 1992b, Patter- the recent excavations in 1999, this list can
son 1993, Bannikov & Tyler 1995). We ac- be extended by a number of taxa repre-
cept here the analysis of Muzylev (1994) senting their first discoveries at the generic
indicating that the fish-bearing layer of the and familial levels at the Gorny Luch lo-
Danata Formation is synchronous with the cality; e.g., a champsodontid, anAntigonia-
Upper Thanetian sapropel of more western like caproid, a Seriola-like carangid, a Sar-
regions, corresponding to global Late Pa- da-like scombrid, a percoid fish probably
leocene anoxic events. belonging to a new genus ofuncertain fam-
In 1999, excavations in the North Cau- ily, and the new species of luvarid de-
casus by the first listed author yielded the scribed below. The great majority ofthe tel-
imprint of a skeleton (incomplete posteri- eost species from the Gorny Luch assem-
orly) that represents a new species of Avi- blage are oceanic pelagic (epi- and meso-)
toluvarus. This specimen was found in the and the new species of Avitoluvarus de-
bituminous marls of the Kuma Horizon in scribed below is presumably epipelagic like
the Gorny Luch locality (Pshekha River, the only Recent representative of the fam-
Apsheronsk District, about 0.5 km from the ily, Luvarus imperialis Rafinesque, 1810.
Gorny Luch farmstead). The KumaHorizon The new species represents the first re-
correlates with the Bartonian stage. The cord of the genus Avitoluvarus outside of
Kuma Horizon previously has been as- Turkmenistan and extends the stratigraphic
580 PROCEEDINGS OF THE BIOLOGICAL SOCIETY OFWASHINGTON
distribution of this genus from the Late Pa- two species ofAvitoluvarus, and the single
mm
leocene to the end of the Middle Eocene. specimen of A. eocaenicus (38 SL) is
within the size range of the specimens of
mm
Family Luvaridae Gill, 1885 A. mariannae (34-1m9m0 SL) and A.
Genus Avitoluvarus Bannikov & Tyler, eocaenicus (38—-120 SL).
1995 Description. The holotype of Avitolu-
Avitoluvarus eocaenicus, new species varus eocaenicus is probably a juvenile
Figs. 1-3 based on the presence ofridges on the skull
— bones and serrations on the dorsal- and pel-
Material. Holotype, Paleontological In- vic-fin spines. Such ridges and serrations
stitute (PIN), Moscow, no. 4425/32, esti- are present in other smalljuvenile (less than
mm mm
mated 38 SL, imprint of poorly pre- about 50 SL) luvarids (e.g., as in A.
served skeleton, incomplete posteriorly, mariannae and Luvarus imperialis) but
single plate, and counterpart of the head these are lost at larger sizes. Also, the skel-
and cleithrum. — eton ofA. eocaenicus seems to be about as
Type locality and horizon. Left bank of weakly ossified as in A. mariannae and
Pshekha River, about 0.5 km from the Gor- somewhat less well ossified than in A. di-
ny Luch farmstead, Apsheronsk District, anae.
Krasnodar Region; Kuma Horizon, upper The body is fusiform, and its greatest
part of the Mi—ddle Eocene. depth is about 41% SL. The head is rela-
Etymology. For the Eocene age of the tively long, about 35% SL. The upper head
new species. — profile is gently curved and the mouth is
Diagnosis. Avitoluvarus eocaenicus small (gap equal to about orbit diameter or
has all the available characters (those ofthe less). The round orbit is situated in the low-
caudal peduncle are unknown because of er part ofthe upper halfofthe head and has
the incompleteness posteriorly of the holo- a horizontal diameter of about 23% of the
type) diagnostic of the genus Avitoluvarus head length.
(see diagnosis in Bannikov & Tyler 1995: Skull: The limits of the individual bones
6). Avitoluvarus eocaenicus differs from of the weakly ossified occipital and otic re-
both of the other two species of the genus, gions are not clear, but the frontals seem to
A. dianae and A. mariannae, by the anter- be relatively wide and the supraoccipital
oventral inclination ofthe third to fifth hae- lacks a prominent crest. The bones of the
mal spines, versus these spines being ori- cranial roof (probably of the frontals) have
ented relatively vertically in A. dianae and remnants of larval ridges, and we presume
inclined posteroventrally in A. mariannae that these ridges were serrate as in other
(for comparative data on the two previously acanthuroids even though we cannot be ab-
described species, here and following, see solutely sure of this because of poor pres-
&
Bannikov Tyler 1995). Additionally, A. ervation of the single specimen; one ridge
eocaenicus differs from the other two spe- is along the dorsal midline and another,
cies by a combination character: the haemal shorter one, is anterior to the orbit.
spines of the first two caudal vertebrae are The ethmoid region is exceptionally
slender and make contact with one another weakly ossified. The parasphenoid is slen-
in their middle regions, versus thesehaemal der and slightly convex where it is exposed
spines being slender but not contacting one at the lower edge of the orbit. The shaft of
another in A. mariannae, or relatively stout the hyomandibular is oriented anteroven-
and contacting one another in A. dianae. trally. The pterygoid bones and the palatine
These diagnostic haemal spine characters are unclear. The quadrate is broad, trian-
do not change appreciably ontogenetically gular, and has a small articular facet for the
within the size ranges known for the other lower jaw. The symplectic is an elongate
VOLUME NUMBER
114, 3 581
Fig. 1. Avitoluvarus eocaenicus, new species, photograph ofthe holotype, PIN 4425/32, estimated 38 mm
SL (incomplete posteriorly), wet with alcohol to improve contrast. Middle Eocene of North Caucasus (Kuma
Horizon), Russia. Scale bar is 10 mm.
rod, somewhat stouter posteriorly than an- Several radiating ridges appear to be pre-
teriorly. There is a faint remnant of a struc- sent in the upper part of the preopercle.
ture in the ethmoid region above the ante- The hyoid and branchial arches are not
rior part of the parasphenoid that can be clear, but there are five sabre-like bran-
interpreted as a long slender process or chiostegal rays.
prong of the lachrymal similar to that of Pectoral fin and girdle: The posttemporal
juvenile specimens of Luvarus imperialis is slender and elongate; it extends from the
(see Tyler et al. 1989). posterodorsal aspect of the head ventrally
The lower jaw articulation is situated and slightly posteriorly. A short intercalar
well in front of the level of the anterior process extends anteriorly from the lower
edge of the orbit. The lower jaw is rather part of the posttemporal. The supracleith-
short but deep, with the dentary and artic- rum is elongate and tapered anterodorsally;
ular of about equal size. The dentary bears this bone is disarticulated in the present
a single row of small conical teeth. The al- specimen. The large cleithrum has a gently
veolar process of the premaxilla is slender, curved c-shape, with the upper and lower
elongate, and bears a single row of small ends inclined, respectively, slightly antero-
conical teeth, whereas the ascending pro- dorsally and anteroventrally. The upperlim-
cess is short but prominent. The maxilla its of the postcleithrum are poorly pre-
seems to be relatively wide. served, but the long shaft ofthe bone below
The thin, flat bones of the opercular re- the pectoral-fin base is sturdy and reaches
gion are poorly preserved, with the limits ventrally below the ventral edge of the an-
ofthe individual bones unclear. The opercle terior extension of the first anal-fin ptery-
appears to be subtriangular and has several giophore (this extension is probably dis-
bony ridges radiating from the condylar re- placed postmortem somewhat dorsally) or
gion. The preopercle is curved along its an- almost to the ventral margin of the body.
terior border, with an angle of about 100°. There is no evidence of a division of the
582 PROCEEDINGS OF THE BIOLOGICAL SOCIETY OF WASHINGTON
/ ••• \
Fig. 2. Avitoluvarus eocaenicus, new species, reconstruction of the holotype (for data see legend for Fig.
1). The unstippled parts of the skeleton posterior to the dashed fracture line through the sixth caudal vertebra
are missing and hypothesized here as having the conditions ofall other species ofthe family Luvaridae (e.g., a
total of 13 caudal vertebrae, thepterygial truss extending posteriorly to the ninthcaudalvertebra, fourvertebrae
in the caudal peduncle, and a forked caudal fin deeply overlapping the hypural plate).
VOLUME NUMBER
114, 3 583
long postcleithral shaft into two pieces, and between the rear of the skull and the front
we presume that the postcleithrum is a sin- of the centrum that is the first to be clearly
gle bone (the right and left postcleithra are exposed (presumably the third abdominal
preserved close alongside one another, ex- vertebra). The vertebral column is elevated
cept where they overlap ventrally, and are anteriorly and articulates relatively high on
so shown in the reconstruction). The cora- the rear ofthe cranium. All ofthe preserved
coid is elongate, of decreasing width ante- centra are somewhat elongate anteroposte-
roventrally. riorly. All ofthe preserved neural spines are
The base ofthe pectoral fin is situated in short, slender, and only slightly curved.
the middle of the body, slightly above the Most of the neural spines are inclined pos-
middle of the distance between the verte- teriorly, but the orientation of those of the
bral column and the ventral profile. The last abdominal and first two caudal verte-
pectoral fin is poorly preserved and its brae is close to vertical. The neural spines
length and number of rays cannot be deter- of the first five caudal vertebrae are shorter
mined. than the corresponding haemal spines. The
Pelvic fin and girdle: The pelvis is L- abdominal vertebrae lack parapophyses.
shaped. The long ascending (pubic) process Short and slender ribs (pleurals) are present
is oriented slightly posterodorsally toward on the third to ninth abdominal vertebrae,
the cleithrum. The posterior (ishial) process becoming slightly shorter posteriorly. The
is well developed, reaching almost to the ribs are inclined posteroventrally and reach
level of the anterior end of the anterior ex- to a level less than one-half the distance
tension of the first anal-fin pterygiophore. between the vertebral column and the ven-
The posterior process is tapered and shorter tral profile of the body. Epineurals are not
than the ascending process. There is essen- evident.
tially no anterior (iliac) process. The haemal spine of the first caudal ver-
The pelvic spine is relatively long (lon- tebra is only slightly thicker than that ofthe
ger than the posteriorprocess ofthe pelvis), second caudal vertebra; it is only moderate-
about 8% SL, robust, and bears serrations ly curved anteroventrally in its lower re-
along its anterior edge, being similar to the gion. The first haemal spine is so closely
first dorsal-fin spine except somewhat articulated along its anterioredge to the first
shorter. No soft rays are evident in the pel- anal-fin pterygiophore that its full length
vic fin. The pelvic fin ofluvaroids becomes cannot be determined. The haemal spine of
rudimentary with increasing specimen size, the second caudal vertebra is very slender
&
or it is absent (Bannikov Tyler 1995:33), and almost straight. The second haemal
so it can be anticipated that when larger spine is angled anteriorly from its base at
specimens of A. eocaenicus become avail- the centrum and closely approaches or con-
able the pelvic-fin spines will be much tacts the posterior edge ofthe middle ofthe
shorter (or absent) than in the present spec- first haemal spine, distal to which point of
mm
imen of ca. 38 SL. contact these two haemal spines diverge.
Vertebral column: There are probably 9 The haemal spines ofthe third to sixth cau-
+ 13 = 22 vertebrae, as in all other species dal vertebrae are exceptionally slender,
ofboth fossil and Recent luvarids; ifso, the straight or only slightly curved, and of de-
posterior seven vertebrae are missing in the creasing length posteriorly in the series.
only specimen of A. eocaenicus (the miss- The third to fifth haemal spines are inclined
ing vertebrae are indicated hypothetically anteroventrally, whereas the sixth haemal
by dashed lines in the reconstruction, Fig. spine is oriented slightly posteroventrally,
2). The first two abdominal vertebrae are and close to vertical.
obscured, but can be reasonably estimated The caudal peduncle, caudal skeleton,
on the basis of the space available for them and caudal fin are missing.
584 PROCEEDINGS OF THE BIOLOGICAL SOCIETY OF WASHINGTON
Dorsal and anal fins: The posterior parts has a long, thick anterior extension beneath
ofthese fins are missing and the total num- most of the abdominal cavity, and a long,
ber of dorsal- and anal-fin elements cannot equally stout posterodorsal process whose
be determined exactly. However, based on upper end firmly articulates along the an-
the assumption that the missing parts ofthe terior edge of the lower end of the haemal
dorsal fin are similar to those of other lu- spine of the first caudal vertebra. The sec-
varids, we estimate that there are about 24 ond and more posterior anal-fin pterygio-
dorsal-fin elements. The first dorsal-fin el- phores have the distal ends expanded into
ement is definitely a spine. It is relatively shallow anterior and posterior processes
long, about 11% SL, borne in supernumer- that form a continuous truss that we pre-
ary association on the first dorsal-fin pter- sume must have been extensively interdig-
ygiophore, and bears serrations along its an- itated even though we cannot determine the
terior edge. The second and third elements details ofthis because ofpoor preservation.
are represented by faint remnants; these are As preserved, the ventral edge of the
evidently thicker than the subsequent ele- truss in the abdominal region just behind
ments, and therefore are probably spines. the anterior extension of the first anal-fin
The remaining elements are clearly soft pterygiophore is strongly convex (Figs. 1,
rays that are unsegmented, unbranched, and 2), uniquely so among luvarids. However,
bilaterally paired (as seen in some elements we do not believe that this is an autapo-
that are separated into slightly displacedleft morphy of the new species, but, rather, a
and righthalves). The rays are much shorter postmortem artifact associated with the up-
than the first dorsal-fin spine. The dorsal- ward displacement ofthe anteriorextension
fin pterygiophores are very poorly pre- of the first anal-fin pterygiophore above the
served and the majority of their ventral level ofthe lower end ofthe postcleithrum;
shafts are unclear. The distal expansions of the weak ossification in this juvenile spec-
the dorsal-fin pterygiophores apparently imen would lend itself to this distortion. In
form a continuous truss, but we cannot de- Figure 3 we show this region of the truss
termine its thickness. The upper region of as preserved (A) and as we believe it would
the ventral shaft of the first dorsal-fin pter- appear if undistorted (B).
ygiophore is expanded into lamellar plates With the exception of the first anal-fin
both anteriorly and posteriorly. Based on its pterygiophore, the other pterygiophores are
position, the lower region of the ventral very slender. The shafts of the second to
shaft of the first dorsal-fin pterygiophore is sixth pterygiophores are situated between
situated in what we estimate to be the pre- the haemal spines of the first and second
neural space in front ofthe distal end ofthe caudal vertebrae (first interhaemal space),
neural spine of the first vertebra. and those ofthe seventh to ninth pterygiop-
Only a few of the anal-fin rays are pre- hores are in the second interhaemal space.
served, in about the middle of the fin, but The succeeding three interhaemal spaces
based on the number ofpterygiophores pre- accommodate two or three pterygiophore
served and ofmissing parts being similarto shafts each. The shafts of the anal-fin pter-
those of other luvarids, we estimate that ygiophores are not especially convergent
there are about 22 anal-fin rays. The anal- toward the haemal spines.
fin rays are unsegmented, unbranched, and Scales: Tiny scales cover the entire body,
bilaterally paired, similar to those of the but poorpreservation precludes detailedde-
dorsal fin and ofabout the same length. The scription.
anal-fin pterygiophores are mostlyT-shaped
Discussion
in lateral view, with a dorsally oriented
&
shaft and an anteroposteriorly expanded Bannikov Tyler (1995) proposed nine
distal end. The first anal-fin pterygiophore unequivocal synapomorphies for the super-
VOLUME NUMBER
114, 3 585
Fig. 3. Avitoluvarus eocaenicus, new species: A, structure ofthe anterior pterygiophores ofthe anal fin as
preserved with postulated convex distortion of the lower edge of the first pterygiophore in the region behind
the anteriorprocess, and with upward displacement ofthe anteriorprocess above the level ofthe ventral end of
the postcleithrum; B, our interpretation of how the anterior pterygiophores of the anal-fin would appear if
undistorted.
family Luvaroidea (Luvaridae + fKushlu- imal shafts of a total of seven to 11 (rarely
kiidae) and six unequivocal synapomor- six) anal-fin pterygiophores are situated in
phies for the family Luvaridae (with two the first two interhaemal spaces (here and
species of Paleocene Avitoluvarus, and two following, character numbers are in paren-
species of Luvarus, one Paleocene and one theses and correspond to those in the clad-
Recent; Tyler et al. 1989, previously doc- ogram. Fig. 4).
umented a far larger number of synapo- Within Avitoluvarus, Bannikov & Tyler
morphies for the Luvaridae based only on (1995) identified one autapomorphy of A.
the single Recent species, butmany ofthese dianae as (3) the posterodorsally oriented
features cannot be determined in fossils or proximal shaft of the first anal-fin ptery-
are unique to the Recent species). Within giophore is especially thick and stout, and
the Luvaridae, the preponderance of de- three autapomorphies of a. mariannae as:
rived features were found in Luvarus, (4) the proximal shafts of a total of 10 to
whereasAvitoluvarus was shown to possess 11 anal-fin pterygiophores are situated in
only two unequivocal synapomorphies: (1) the first two interhaemal spaces (the greater
the truss formed by the interdigitation ofthe increase in number of pterygiophores in
distal regions of the dorsal- and anal-fin these two spaces in A. mariannae being a
pterygiophores is relatively shallow and not more derived condition than the lesser in-
extensively interdigitated; and (2) the prox- crease in A. dianae); (5) the ribs are rela-
586 PROCEEDINGS OF THE BIOLOGICAL SOCIETY OFWASHINGTON
CO
LU
O <
<
z
LU
< <UJ < C3O
o
X-J < O D<C D<C
CO Q >
LU
3
<
4-8
1-2
Fig. 4. Cladogram ofthe proposed relationships withinAvitoluvarus. Derived character states, discussedin
the text, are: (1) pterygial truss shallow and not extensively interdigitated; (2) proximal shafts ofa total of7-
11 (rarely 6) anal-fin pterygiophores situated in the first two interhaemal spaces; (3) posterodorsally oriented
proximal shaftofthe first anal-fin pterygiophore especially thick and stout; (4) proximal shafts ofatotal of 10-
11 anal-fin pterygiophores situated in the first two interhaemal spaces (5) ribs relatively short and thin, ending
at the level ofthe upperthird ofthe abdominal cavity; (6) ventral halfofpterygial truss especially shallow; (7)
firsttwohaemal spinesonly slightlythickened; (8) proximal shaftsofthe secondandmanyofthemoreposterior
anal-fin pterygiophores exceptionally slender; (9) third to fifthhaemal spines inclined anteroventrally. Character
states 1-6 and 9 are unequivocal apomorphies, but states 7-8 are equivocal and apomorphic only under the
assumption ofdelayed transformation, favoring independent acquisition over reversal. For data supporting this
and the sistergroup relationships ofKushlukia withAvitoluvarus+Luvarus, see Bannikov & Tyler (1995).
lively short and thin, ending at the level of Moreover, A. mariannae was shown to pos-
about the upper third ofthe abdominal cav- sess two equivocal autapomorphies (which
ity; and (6) the ventral halfof the pterygial assume delayedtransformation, favoringin-
truss formed from the distal ends of the dependent acquisition over reversal, with
anal-fin pterygiophores is very shallow. the ancestor possessing the plesiomorphic
VOLUME NUMBER
114, 3 587
condition ofthe outgroups): (7) the firsttwo one of the three species ofAvitoluvarus (in
haemal spines are only slightly thickened; A. eocaenicus, versus posteroventral in A.
and (8) the proximal shafts of the second mariannae and vertical in A. dianae). It is
and more posterior anal-fin pterygiophores, most parsimonious to propose that the an-
except for the last few, are exceptionally teroventral orientation has arisen indepen-
slender. dently in the single species of Kushlukia,
The new species, A. eocaenicus, shares Luvarus, and Avitoluvarus; therefore, the
four derived character states (5-6 unequiv- anteroventral orientation ofthe third to fifth
ocal, 7-8 equivocal) withA. mariannae and haemal spines is an autapomorphy of A.
lacks the single autapomorphy (3) ofA. di- eocaenicus.
anae, all of which indicates a sister group
relationship between A. eocaenicus and A. Conclusion
mariannae.
The addition of another new species to
For one character (4), A. eocaenicus is the record of fossil luvarids reinforces the
intermediate betweenA. dianae andA. mar- &
notion (Bannikov Tyler 1995:40) that the
iannae. Two to five anal-fin pterygiophore luvaroid fishes were far more diversified in
shafts in the first two interhaemal spa&ces is the early Tertiary (six species in two fami-
ancestral for acanthuroids (Bannikov Ty-
lies, with four species from the uppermost
ler 1995), and the increase in number of Paleocene of the Danata Formation, Turk-
such pterygiophores in Avitoluvarus can be menistan, alone) than they are today, when
considered an ordered transformation series
only a single species survives.
of increasing specialization from the six to
seven inA. dianae to the eight inA. eocaen- Acknowledgments
icus, to the 10—11 inA. mariannae. For this
character within Avitoluvarus, we consider We thank the National Geographic So-
the eight or more pterygiophores in these ciety, Washington DC, for grant no. 6555-
two spaces as a derived condition synapo- 99 that allowed the first listed author to
morphic for A. dianae and A. mariannae. make excavations in the North Caucasus in
We identify an autapomorphy of A. 1999, where the holotype of Avitoluvarus
eocaenicus as (9) the third to fifth haemal eocaenicus was obtained. The manuscript
spines are inclined anteroventrally. We con- was improved by the suggestions of G.D.
sider posteroventral to vertical inclination Johnson, Smithsonian Institution, and an
of the third to fifth haemal spines as ances- anonymous reviewer. The photograph of
tral for acanthuroids because this is the con- the holotype was made by A.V. Mazin
dition found in fossil and Recent scato- (PIN, Moscow).
phagids (the first outgroup for acanthu-
roids), siganids, zanclids, and most acan- Literature Cited
thurids (including those considered to be Bannikov, A. F. 1985. Iskopayemye skumbriyevye
the most basal): see illustrations in Tyler et SSSR [Fossil scombrids ofthe USSR].—Trudy
al. 1989, Blot & Tyler 1991, Tyler & Ban- Paleontologicheskogo Instituta Akademii Nauk
nikov 1997, Tyler & Sorbini 1999). Among SSSR 210:1-111.
the luvarid + tkushlukiid clade (Bannikov . 1993. The succession of the Tethys fish as-
& semblages exemplified by the Eocene lo—calities
Tyler 1995), anteroventral orientation of ofthesouthernpartoftheformerUSSR. Kau-
the third to fifth haemal spines is found pia, Darmstadter Beitrage zur Naturgeschichte
only in one ofthe two species ofKushlukia 2:241-246.
(in K. permira, versus posteroventral ori- , & N. N. Parin. 1997. The fistofmarine fishes
from Cenozoic (Upper Paleocene-Middle Mio-
entation in K. sp.), in one ofthe two species cene) localities in sou—thern European Russia
ofLuvarus (in L. imperialis, versus vertical andadjacentcountries. JournalofIchthyology
to posteroventral in L. necopinatus), and in 37:133-146.
—
588 PROCEEDINGS OF THE BIOLOGICAL SOCIETY OF WASHINGTON
, & J. C. Tyler. 1995. Phylogenetic revision of lution of the biosphere. Issue 1] Nedra, Mos-
the fish families Luvaridae and tKushlukiidae cow, 367 pp.
(Acanthuroidei), with anew—genusandtwonew Patterson, C. 1993. Osteichthyes: Teleostei. Pp. 621-
species ofEocene luvarids. SmithsonianCon- 656 in M. J. Benton, ed.. The fossil record 2.
tributions to Paleobiology 81:1-45. Chapman and Hall, London, 845 pp.
Blot, J., & J. C. Tyler. 1991. New genera and species Tyler, J. C, & A. E Bannikov. 1992a. New genus of
of fossil surgeon fishes and their relatives primitive ocean sunfish with separate premax-
(Acanthuroidei, Teleostei) from the Eocene of illae from the Eocene of so—uthwest Russia
Monte Boica, Italy, with application oft—he Blot (Molidae, Tetraodontiformes). Copeia 1992
formula to both fossil and Recentforms. Mis- (4):1014-1023.
cellanea Paleontologica, Studi e Ricerche sui . 1992b. Aremarkablenew genusoftetraodon-
Giacimenti Tertziari di Bolca, Museo Civico di tiform fish with features of both balistids and
Storia Naturale di Verona 6:13-92. ostraciids from the Eocene ofTurkmenistan.
Cavelier, C, & C.—Pomerol. 1986. Stratigraphy ofthe Smithsonian Contributions to Paleobiology 72:
Paleocene. Bulletin de la Societe de la Geo- 1-14.
logique de France 2(2):255-265. . 1997. Relationships of the fossil and Recent
Danilchenko, P. G. 1968. Ryby verkhnego paleotsena gener—a of rabbitfishes (Acanthuroidei: Sigani-
Turkmenii [Fishes of the Upper Paleocene of dae). Smithsonian Contributions to Paleobi-
Turkmenistan]. Pp. 113-156 in D. V. Obruchev, ology 84:1-35.
ed., Ocherki po filogenii i sistematike iskopay- , & C. Sorbini. 1999. Phylogeny of the fossil
emykh ryb i beschelyustnykh [Outlines on the and Recent genera offishes o—fthe family Sca-
phylogeny and systematics of fossil fishes and tophagidae (Squamipinnes). Bollettino del
agnathans]. Nauka, Moscow, 173 pp. Museo Civico di Storia Naturale di Verona 23:
Muzylev, N. G. 1994. Anoksicheskiye sobytiya pa- 353-393.
leotsena-srednegoeotsena[Anoxiceventsofthe , G. D. Johnson, I. Nakamura, &B.B. Collette.
Paleocene-Middle Eocene]. Pp. 160-166 in 1989. Morphology of Luvarus imperialis (Lu-
A.Y. Rozanov andM.A. Semikhatov, eds.,Eko- varidae), with a phylog—enetic analysis of the
sistemnye perestroiki i evolyutsiya biosfery. Acanthuroidei (Pisces). Smithsonian Contri-
Vypusk 1 [Ecosystemrestructuringandtheevo- butions to Zoology 485:1-78.