co — CO £ o

INSTITUTION NOIlfUllSNI NVIN0SH1IINS S3iavaail LIBRARIES SMITHSON

CO 21 CO Z v.^. CO

s ^ws?T

5 ■^wr a (|r *31 * a 5

m 9 M £ 9 %M> £ Wv I

> > *<Wr ^\ > x^Ua^J

■V Z CO 2 00 * Z

S3iavaan libraries Smithsonian institution NouniixsNi nvinoshj

'?&***&/ m v/j'm&y — x^r*®r^y

O \^V. PC^X V^M-tlXy^ Q

INSTITUTION2 N0lirU!lSNl“JNVIN0SHimSZS3 lavaail^LIBRAR! ES^SMITHSOP

CD
2J

*J >

o/

—

m \w m

u> " ± co !=

SBIBVaail LIBRARIES SMITHSONIAN INSTITUTION NOIlfUllSNI NVINOSHJ

■ z \ ^ 2: co 2:

< ^ S X^STTT^X < S <

voxw- - % !■ " ', i '^5;

CO ^ CO **" Z CO

INSTITUTION NOIlfUllSNI NVINOSHIIINS SliaVHan LIBRARIES SMITHSOP,

CO z ^ ^ 5 CO

o

_J z -J z _

S3 iava an libraries Smithsonian institution NoiinuiSNi nvinoshj

2: r~ a z r-

m r; m

co — co _; co

INSTITUTION NOIlfUllSNI NVINOSHlItNS S3 I HVB8 II LIBRARI ES^SMITHSOI

I i I i I

V £ ■ OT Z w * • Z

saiavaan libraries Smithsonian institution noixoiiisni nvinosh.

q x^/y . xcumv>^ q

__ _ Z _J Z

INSTITUTION NOIlfUllSNI NVINOSHIIIMS SBIdVdaiT LIBRARIES SMITHSO;
z r- » z r~ z

m

ED

n m rn | '*<nmgy m

^INSTITUTION ^ NOlIfUIXSNl NVINOSHIIINS^SS I ava a ll^LI B RAR I ES^SMITHSOr

o w -a»x M 3 /0'

| (II il) g £ TO, • 8

> W I

WS3 I Hva a II2 Ll B RAR 1 ES^SMITHSONIAN _ INSTITUTION ^NOIinilXSNI_NVINOSH.

CO

""INSTITUTION NOliniliSNI NVINOSHIIINS S3iyVd9ll L I B R A R I ES SMITHSO

T! X/* osvvvx " VVAbiy m 2 m __

A- (S) — CO —

S3IHVB9I1 LIBRARIES SMITHSONIAN INSTITUTION NOliniliSNI NVINOSH

i>3iavaa 2 •. CO Z r W Z ^ ^

Wfa. £ mA. 5 o X Jggk; 1 |

I ^2^ 1 , t % _

>- ^ >' 2 X^os^J

INSTITUTION U>N0llfUUSNl_NVIN0SHllWs"s3 I UVa 8 IT_LI B RAR I ES^SMITHSO
SSB/fc, w \k 111 th 114 .o^ll

S3iavaan libraries Smithsonian iNSTiTUTi0N_N0iiniiiSNi_NViN0SH
i— 3?*’ r~

XQ^5cjx ^ n ^ Xoit^x m

CO ± CO ± CO

INSTITUTION NOliniliSNI . NVIN0SH11WS S3 I 9V9 9 ,1.1 L I B RAR I ES^SMITHSC

Z ,\v CO -r (n Z

S3 [HVyailZuBRAR1 E S^SMITHSON I AN __ INSTITUTION NOI mills Nl__NVINOSI
to co

CO

o Xfr-P-SX ” "<^rr O o

INSTITUTION^ NOliniliSNI^NVINOSHill/MS _^S3 I UVd a IT^LI B RAR I ES^SMITHS

O Q Nq

ifSl

Sf-7'73

C-z-l-Z£f

NUMBER 219
JANUARY 26, 1972

NEW SPECIES OF SALAMANDERS
(GENUS BOLITOGLOSSA ) FROM
COLOMBIA, ECUADOR and PANAMA

By Arden H. Brame, Jr. and David B. Wake

/

CONTRIBUTIONS IN SCIGNCC

NATURAL HISTORY MUSEUM

LOS ANGELES COUNTY

CONTRIBUTIONS IN SCIENCE is a series of miscellaneous technical papers
in the fields of Biology, Geology and Anthropology, published at irregular intervals
by the Natural History Museum of Los Angeles County. Issues are numbered sep-
arately, and numbers run consecutively regardless of subject matter. Number 1 was
issued January 23, 1957. The series is available to scientific institutions and scien-
tists on an exchange basis. Copies may also be purchased at a nominal price. Inquiries
should be directed to Virginia D. Miller, Natural History Museum of Los Angeles
County, 900 Exposition Boulevard, Los Angeles, California 90007.

INSTRUCTIONS FOR AUTHORS

Manuscripts for CONTRIBUTIONS IN SCIENCE may be in any field of
Life or Earth Sciences. Acceptance of papers will be determined by the amount
and character of new information. Although priority will be given to manuscripts
by staff members, or to papers dealing largely with specimens in the collections of
the Museum, other technical papers will be considered. All manuscripts must be
recommended for consideration by the curator in charge of the proper section or
by the editorial board. Manuscripts must conform to those specifications listed
below and will be examined for suitability by an Editorial Committee including
review by competent specialists outside the Museum.

Authors proposing new taxa in a CONTRIBUTIONS IN SCIENCE must
indicate that the primary type has become the property of a scientific institution of
their choice and cited by name.

MANUSCRIPT FORM.— ( 1 ) The 1964 AIBS Style Manual for Biological
Journals is to be followed in preparation of copy. (2) Double space entire manu-
script. (3) Footnotes should be avoided if possible. Acknowledgments as footnotes
will not be accepted. (4) Place all tables on separate pages. (5) Figure legends and
unavoidable footnotes must be typed on separate sheets. Several of one kind may
be placed on a sheet. (6) An abstract must be included for all papers. This will be
published at the head of each paper. (7) A Spanish summary is required for all
manuscripts dealing with Latin American subjects. Summaries, in other languages
are not required but are strongly recommended. Summaries will be published at
the end of the paper. (8) A diagnosis must accompany any newly proposed taxon.
(9) Submit two copies of manuscript.

ILLUSTRATIONS. — All illustrations, including maps and photographs, will
be referred to as figures. All illustrations should be of sufficient clarity and in the
proper proportions for reduction to CONTRIBUTIONS page size. Consult the 1964
AIBS Style Manual for Biological Journals in preparing illustration and legend copy
for style. Submit only illustrations made with permanent ink and glossy photo-
graphic prints of good contrast. Original illustrations and art work will be returned
after the manuscript has been published.

PROOF.— Authors will be sent galley proof which should be corrected and
returned promptly. Changes in the manuscript after galley proof will be billed to
the author. Unless otherwise requested, page proof also will be sent to the author.
One hundred copies of each paper will be given free to each author or divided
equally among multiple authors. Orders for additional copies must be sent to the
Editor at the time corrected galley proof is returned. Appropriate order forms will
be included with the galley proof.

Virginia D. Miller
Editor

by DOBBS BROS. LIBRARY BINDING CO., INC., St. Augustine, I®ffia 9. 7*5, < 2 8. 2 *

I

NEW SPECIES OF SALAMANDERS (GENUS BOLITOGLOSSA ) FROM
COLOMBIA, ECUADOR AND PANAMA

By Arden H. Brame, Jr.1 and David B. Wake2

Abstract: Several undescribed species of plethodontid sala-
manders are reported from South America and Panama. Boli-
toglossa medemi is a dark species with rather large, extensively
webbed hands and feet and a broad head. It is known from
several sites in northwestern Colombia and extreme southwestern
Panama, where it occurs between 50 and 800 m elevation.
Bolitoglossa ramosi is a somewhat smaller species, with lighter
coloration, small but extensively webbed hands and feet, and a
broad head. It occurs in sympatry with Bolitoglossa vallecula
in the Cordillera Central east of Medellin, Colombia, at eleva-
tions of about 1930 m. Bolitoglossa silverstonei is a slender,
long-tailed species with light pigmentation, broad and extensively
webbed hands and feet, and a head of moderate breadth. It is
known only from a site near Quibdo, in northwestern Colombia,
at an elevation of 400 m. Bolitoglossa walked has less exten-
sively webbed hands and feet and fewer maxillary teeth than the
other species. It has dark dorsal and light ventral coloration.
This species occurs near Cali, Colombia, at elevations near 2000
m. Bolitoglossa equatoriana has more extensively webbed hands
and feet than B. walked, but shares similar low numbers of
teeth and coloration. It occurs sympatrically with B. peruviana
at elevations of about 260 m in Amazonian Ecuador. These new
species permit a re-evaluation of relationships among South
American members of the genus Bolitoglossa. Characters used
for analysis of relationships among the twenty-two species are
discussed, and five species groups are recognized. The palmata
and altamazonica groups are divided. B. silverstonei is assigned
to a new sima group, along with B. sima, B. chica, and B.
biseriata. The other newly described species are the only mem-
bers of the medemi group.

Salamanders from the Neotropics have been known for over 140 years,
but until recently they have been considered to be rare and insignificant.
However, it is increasingly evident that the salamander fauna of the New
World Tropics is extensive and diverse. The plethodontid salamanders, which
range from northern Mexico to Bolivia, account for over 40 percent of all
salamander species. This diverse group is of interest to evolutionary biologists

jjj Research Associate, Section of Herpetology, Natural History Museum of Los
e" Angeles County; and Supervisor, Eaton Canyon Nature Center, 1750 North
^ Altadena Drive, Pasadena, California 91107.

2 Research Associate, Section of Herpetology, Natural History Museum of Los
Angeles County; and Director, Museum of Vertebrate Zoology, University of
California, Berkeley, California 94720.

1

1

2

Contributions in Science

No. 219

in that it provides an opportunity to analyze an adaptive radiation in great
detail. Many generalized salamanders, probably similar to ancestral forms,
survive in extratropical habitats. In addition, many populations intermediate
between adaptive extremes are known to occur in the tropics.

A major hindrance to intensive analysis of evolutionary patterns has
been the absence of basic biological information concerning the tropical
species. Most are poorly known, and are represented in collections by only
a few specimens. For some areas, our knowledge of species composition
is fragmentary and new populations continue to be discovered. One such
area is northwestern South America where the known fauna has increased
from one (Dunn, 1926) to nearly twenty species in recent years. In this paper
we describe five additional species and discuss the relationships of South
American members of the genus Bolitoglossa.

Many specimens used for this study were provided by Philip A. Silver-
stone, who collected salamanders incidental to his work with Colombian
frogs. Brame obtained additional specimens in Colombia in the spring of
1971 and W. R. Heyer collected a fine series in Ecuador in the summer of
1971. These specimens are deposited in the Los Angeles County Museum
of Natural History (LACM). Specimens also have been loaned to us by
the following curators and museums: Charles F. Walker, University of
Michigan, Museum of Zoology (UNMZ); Hobart M. Smith and Dorothy M.
Smith, University of Illinois Museum of Natural History (UIMNH) ; William
E. Duellman, University of Kansas Museum of Natural History (KU).
Additional material is deposited at the Museum of Vertebrate Zoology,
University of California, Berkeley (MVZ). We thank these curators and
institutions for their assistance. We are grateful to William F. Presch for
assistance in providing x-rays, and Les Siemens and Gene Christman for
assistance with illustrations. Carlos Martinez provided the Spanish summary.
Aspects of the work have been supported by NSF grant GB 17112 to David
B. Wake.

The first species occurs in low coastal mountains of northeastern Colom-
bia and neighboring Panama. We are pleased to name it in honor of our good
friend, Professor Federico Medem, of the Universidad Nacional de Colombia
(Villavicencio), who has aided and encouraged us in our studies of Colombian
salamanders.

Bolitoglossa medemi , new species
Figures 1 and 2

Holotype— LACM 42276, an adult female from Finca Chibigui, approx-
imately 76° 30' W, 6° 15' N, on the Rio Arquia, Departamento de Antioquia,
Colombia. The specimen was collected on April 23, 1968, by Philip A. Silver-

1971

New Species of Salamanders

3

stone, Jorge E. Ramos, and Nacianseno Borja. It was active on the ground
during daylight hours. Elevation approximately 300 m (980 ft).

Paratypes -COLOMBIA: LACM 42277-78, same data as holotype; PAS
237 (cleared and stained, LACM 72067), Belen, downstream from Finca
Chibigui, and very near Vegaes on the Rio Arquia, Dept. Antioquia, about
100 m (328 ft) elevation; LACM 42280, along trail between Rio Opogodo
and Rio Napipi, near the latter, approximately 77° 10' W, 6° 43' N, Dept.
Choco, 30 to 80 m (100-260 ft); LACM 42279, along Rio Opogodo at base
of eastern slope of the Serrania de Baudo, approximately 77° 18' W, 6° 50' N,
Dept. Choco, about 60 m (200 ft) ; LACM 70565, N slope Alto de Buey, Dept.
Choco, 400 m (1312 ft); LACM 70566, Camino de Yupe, Dept. Choco,
350 m (1148 ft); LACM 70567, Camino de Yupe, Dept. Choco, 400-500 m
(1312-1640 ft); LACM 70568, Camino de Yupe, Dept. Choco, 605 m (1984
ft). PANAMA: KU 116533-34, Rio Jaque, 1.5 km above Rio Imamado
approximately 77° 57' W, 7° 25' N, Prov. Darien, 50 m ( 164 ft) ; KU 116530,
Jaque -Imam ado divide in Cordillera de Jurado, above Rio Jaque, Prov.
Darien, 730-800 m (2394-2625 ft).

Diagnosis— A moderately small species (5 adult males: 33.7-46.7, mean
40.4 mm SL [standard length, measured from tip of snout to posterior angle
of vent]; 5 adult females: 34.2-48.2, mean 43.3 mm SL) with moderate
numbers of maxillary (mean 41) and vomerine (mean 31) teeth. Distin-
guished from B. ramosi by having fewer maxillary teeth and a darker dorsal
ground color; from B. walkeri by its broader head, more extensively webbed
feet, and longer legs; from B. equatoriana by its more numerous maxillary
teeth and somewhat longer legs. Bolitoglossa medemi is distinguished from
other Panamanian and South American species by the combination of large,
extensively to completely webbed hands and feet, relatively broad head, long
legs, distinctive coloration (very dark, unmarked dorsum, much lighter venter
with a few widely scattered, irregular light pigment spots), and size and
dentitional features (Table 1).

Description of Holotype— Adult female with moderately long, somewhat
pointed snout and small nostrils. Labial protuberances of nasolabial grooves
small, poorly developed. Moderately long canthus rostralis gently arched.
Head broad (SL 5.8 times head width) and moderately long (SL 4.2
times snout- gular fold length). Deep groove below eye extends for almost
full length of orbit, following curvature of eye, but does not commu-
nicate with lip. Large eyes slightly protuberant. Well-defined postorbital
groove extends posteriorly from eye as shallow depression for 2.2 mm, then
sharply ventrad at level of posterior end of mandible and across gular area
as nuchal groove, parallel to, and 4.8 mm anterior to sharply defined gular
fold. Vomerine teeth number 23, arranged in single rows that extend to lateral
borders of internal nares; rows form slightly curved arches that terminate
in center of palate, where they nearly meet. Small maxillary teeth number 4 1 ;
extend posteriorly to point about two-thirds through eye. No premaxillary

4

Contributions in Science

No. 219

teeth can be seen. Relatively short tail (0.75 times SL) has strong lateral
compression and is moderately constricted at base. Postiliac glands poorly
developed. Limbs long with limb interval (costal folds between appressed
limbs) of one; SL 4.0 times right forelimb, 4.0 times right hind limb, 10.0
times right foot width. Webbing of hands and feet nearly complete, but all
digital tips extend beyond thin web. Longest digits with long, pointed tips
(Fig. 2). Large hands and feet; rather narrow and long compared with those
of other extensively webbed species. No subterminal pads. Fingers, in order
of decreasing length, 3, 4, 2, 1 ; toes, in order of decreasing length, 3, 4, 2, 5, 1.

Measurements (in mm) are as follows: Head width, 8.3; snout to gular
fold (head length), 11.5; head depth at posterior angle of jaw, 4.0; eyelid
length, 3.4; eyelid width, 2.2; anterior rim of orbit to snout, 3.7; horizontal
orbital diameter, 2.6; interorbital distance, 3.0; distance between vomerine
teeth and parasphenoid tooth patch, 0.8; snout to forelimb, 15.5; distance
separating internal nares, 2.1; distance separating external nares, 2.7; snout
projection beyond mandible, 0.8; snout to posterior angle of vent (SL), 48.2;

Figure 1. Dorsal and ventral views of holotype of Bolitoglossa medemi (LACM
42276).

1971

New Species of Salamanders

5

snout to anterior angle of vent, 44.9; axilla to groin, 26.6; tail length, 36.3;
tail width at base, 3.0; tail depth at base, 3.7; forelimb length, 12.2; hind
limb length, 12.2; width of right hand, 3.6; width of right foot, 4.8.

Coloration of Holotype (in alcohol) .-This is a very dark salamander
which has the dorsum and upper two-thirds of the lateral sides of the trunk

Figure 2. Outlines of hands and feet of species of Bolitoglossa drawn from cleared
and stained specimens through use of microprojector, a. Right hand of B. ramosi
(LACM 64603). b. Right foot of B. ramosi (LACM 64603). c. Right hand of
B. medemi (LACM 72067). d. Right foot of B. medemi (LACM 72067).

6

Contributions in Science

No. 219

uniformly colored a deep, leaden black. The dorsal surfaces of the head, tail,
and legs are similarly colored. Small, indistinct guanophores are scattered on
the snout, eyelids, and around the insertion of the limbs. Ventral surfaces are
distinctly lighter than dorsal ones, and have a general grayish cast. Scattered
guanophores are conspicuous on the throat and anterior part of the venter.
Posteriorly and on the tail small to moderately sized, irregularly shaped
patches of golden cream to grayish silver pigment occur. The ground color
of the tail is darker than that of the throat and belly. Ventral surfaces of the
limbs are mottled black and light gray, and the hands and feet are medium
gray. The iris is golden with melanic mottling.

Variation — Specimens from the Departamento de Choco have somewhat
lighter venters with fewer light patches that the animals from the Depto.
Antioquia. The Panamanian individuals are similar in coloration to the
holotype, but lack light ventral patches and have indistinct guanophores. The
distinction between the dark dorsal and lateral and lighter ventral coloration
is somewhat sharper than in the Colombian specimens.

A total of ten adult specimens are available, nine of which are from
Colombia. The single adult from Panama will be discussed separately. Males
have small, well-defined, rounded mental hedonic glands. Females are larger
(4 females, 34.2-48.2, mean 42.7 mm SL; 5 males 33.7-46.7 mean 40.4 mm
SL) and have longer legs (SL 3. 5-4.3, mean 3.9 times hind limb length in
females versus 3.4— 4.0, mean 3.7 in males) than males. Limb interval is zero
to 1.5 in males and one to 1.5 in females. Males have premaxillary teeth which
penetrate the upper lip, but females lack them.

The single adult female from Panama differs from the Colombian females
in having premaxillary teeth. It is about the same size as the Colombian speci-
mens but has a narrower head (SL 6.4 times head width) than any Colombian
individuals (SL 5.6-6. 1, mean 5.8, times head width). The feet of the Pana-
manian adult are somewhat narrower than those of the Colombian specimens.

The juveniles are uniformly dark dorsally and lighter ventrally, with no
distinctive markings.

Osteology— Information concerning osteology has been derived from
one cleared and stained female and from stereoscopic radiographs of all adults
available.

The heavily ossified, well sutured skull is as well developed as that of
any member of the genus. The premaxilla bone has a dental part that is well
developed. It is closely aligned with the maxillae on either side. This align-
ment is typical of female Bolitoglossa. However, the toothless condition of
the relatively large premaxilla is unusual. Frontal processes of the premaxilla
are stout. They ascend along the margins of the cartilaginous nasal capsule,
then proceed posteriorly. Near their tips they are dilated and in close apposi-
tion. As a result, the internasal fontanelle is very small and is restricted to the
anterior end of the snout. The ascending part of each frontal process bears
a winglike, flattened process lying against the anteromedial surface of the car-

1971

New Species of Salamanders

7

tilaginous nasal capsule. These processes meet similar enveloping processes
of the nasal. Posteriorly the frontal processes broadly overlap the expanded
anterior part of the frontals in a firm articulation. The processes fall short of
the ends of the nasals, but extend beyond the anterior margin of the orbits.
Nasal bones are very large and protuberant. They extend far anteriorly where
they overlie the enlarged nasal capsules. The area occupied by the nasal and
prefrontal of more primitive species of the genus is included within the area
of the nasal. Medially and anteriorly the nasals overlap the middle parts of
the frontal processes of the premaxilla. The overlapping pieces are in medial
contact for a short distance, an unusual arrangement in this genus. Posteriorly
the nasals overlap the frontals and terminate in rounded borders beyond the
margin of the orbits, approximately at the level of the eyes. Large ventro-
lateral lobes of the nasals overlap the anterior margin of the relatively large
facial processes of the maxillae. The nasal is evacuated posterior and medial
to these lobes. The nasolacrimal duct extends from the nasal capsule through
this evacuated area, then posteriorly through the lower layers of the skin to
the anterior corner of the eye. The route of the duct is free of bone. This
bone-free area extends from the anterior end of the evacuation in the nasals,
between the nasals and the maxillae to the eyes. The maxillae are well-
developed bones that extend posteriorly to the limit of the eyes. Anteriorly
the maxillae are produced into a flattened sheet of bone which partially under-
lies and envelops the cartilaginous nasal capsule. The palatal processes are
small, but the facial processes are moderately large and relatively high.

Vomers are well developed and completely separated from each other,
except posteriorly where the toothed portions are in slight contact. The inter-
vomerine fontanelle is broad. Preorbital processes extend beyond the lateral
margins of the vomerine bodies. Vomerine teeth are in patches which barely
extend onto the preorbital processes.

Frontals are large and stout, with a strong sutural joint along the mid-
line. The facial portions are stout, but are not especially large, in contrast to
more northern species (Wake and Brame, 1969). No marked lobes are present
posteriorly, where the margin is irregular. Parietals are well developed and
have the parietal spurs that are characteristic of the genus. The occipito-otic
bones bear low crests over the anterior vertical and lateral horizontal semi-
circular canals. The ridges over the latter form braces for the relatively well-
developed, vertically oriented squamosals. The large parasphenoid is very
narrow anteriorly. The anterior terminus is blunt, rather than pointed. Where
the parasphenoid is narrowest, the orbitosphenoids nearly contact each other
on the midline. Posterior vomerine teeth are in large patches on the para-
sphenoid. They narrowly fail to come into medial contact. The right patch
bears 91 and the left, 96 ankylosed, bicuspid teeth, in the single cleared indi-
vidual. The operculum has no stilus. Quadrates are stout. They are connected
to the skull by the cartilaginous parts of the suspensorium, and by the rela-
tively large squamosals. The squamosals are very attenuated dorsally where

8

Contributions in Science

No. 219

they fit into a depression in the wall of the otic capsule, and they are broadly
expanded where they overlap the quadrates.

The hyobranchial apparatus is typical of other members of the genus
(Wake, 1966).

Vertebrae are similar to those of other species of Bolitoglossa. The cen-
tra are spool-shaped, the intervertebral cartilages are unmineralized, and no
articular condyles are formed. There are one cervical, fourteen trunk, one
sacral, two caudosacral, and 26 caudal vertebrae in the cleared individual.
Caudal vertebrae in other specimens number 25 (LACM 42276, 42279, both
adults), 24 (LACM 42278, adult, last 5 regenerated), 22 (LACM 42277,
juvenile), 19 (LACM 42280, all regenerated; KU 116534, juvenile), and
17 (KU 116537, all regenerated).

The first caudal vertebra is shorter than the next eleven vertebrae, but is
the same length as the second caudosacral. All but the last two trunk vertebrae
are longer than the longest caudal vertebrae (two to six), but caudal vertebrae
two to ten are longer than the sacral and caudosacral vertebrae. Ribs are
present on all but the last trunk vertebrae. One specimen (LACM 42279)
has a small rib on one side of the last trunk vertebra. Transverse processes
are short on all but the first two or three caudal vertebrae, but they are clearly
present on all but the last vertebra. Transverse processes of the first caudo-
sacral vertebra are long and directed almost perpendicularly to the body axis.
Those of the second are much shorter and are directed somewhat posteriorly.
The large, stout, non-bifurcated processes of the first caudal vertebra arise
from the anterior margin of the vertebra (in contrast to the more central
location of the caudosacral processes). From their anterior origin the proc-
esses extend first anteriorly, then sharply in a lateral direction. They do not
cross those of the more anterior vertebra. This distinctive pattern of processes
on the first three postsacral vertebrae is one not seen in any related or neigh-
boring species. Processes on succeeding vertebrae arise from anterior positions
and are anteriorly directed. They progressively diminish in size posteriorly.
Hypapophyseal keels are absent only on the first and last two caudal ver-
tebrae.

Hands and feet are large and distinctive. They are characterized by rela-
tive narrowness, accentuated by the presence of inordinately long central
digits. Some variation in phalangeal formulae is encountered. The usual
formula is 1, 2, 3, 2 for the hands and 1, 2, 3, 3, 2 for the feet. Two adults
have a foot formula of 1, 2, 3, 2, 2 on one side, and several other individuals
have very small penultimate and terminal phalanges in the fourth digit.
Terminal phalanges are rather well developed but are erratically shaped (Fig.
2). There are seven carpals and eight tarsals, the generalized Bolitoglossa
numbers (Wake, 1966). The tibia bears a prominent, sharp-edged crest, but
has no free spur.

Remarks — PAS 237 (LACM 72067) was captured by Norman J. Scott
on a tree leaf where it was exposed at night. Other Colombian specimens were

1971

New Species of Salamanders

9

collected during daylight hours, exposed on the surface. KU 116530 was
collected in cloud forest (Myers, 1969) where it is sympatric with B.
phalarosoma.

Range— The Choco region of extreme northwestern Colombia, in the
Rio Atrato (Caribbean) drainage, and the Rio Jaque (Pacific) drainage of
extreme southeastern Panama (Fig. 8). The species ranges from about 30 to
800 m ( 1 00 to 2624 ft) in elevation.

Bolitoglossa vallecula has been the only species known from uplands of
the Cordillera Central of Colombia. A second species was collected in 1968
and 1971. We name it after Jorge Eduardo Ramos, who contributed much
toward the success of the Silverstone and Brame— Newcomer trips to
Colombia.

Bolitoglossa ramosi, new species
Figures 2 and 3

Holotype .— LACM 64601, an adult male from near Represa de Santa
Rita ( = Santa Rita Dam Site), Departamento de Antioquia, Colombia. This
site is between Guatape and San Rafael at about 75° 7' W, 6° 17' N, ca. 16
km by road NE of Guatape. The specimen was collected by Brame and Jorge
E. Ramos from the rolled-up base of a large palm frond on the ground of a
forested hill near the dam construction site, April 3, 1971. Elevation about
1930 m (6330ft).

Paratypes — LACM 64600, 64602-03, same data as holotype; LACM
42287-90, collected from bromeliads at the same locality by Philip A. Silver-
stone and Jorge E. Ramos, June 9, 1968.

Diagnosis A moderately small species (5 adult males: 37.1-45.4, mean

40.4 mm SL; 2 adult females: 37.2-46.7, mean 42.0 mm SL) with relatively
high numbers of maxillary (mean 47) and vomerine (mean 32) teeth; distin-
guished from B. medemi by its greater numbers of maxillary teeth and lighter
dorsal ground color; from B. equatoriana and B. walkeri by its broader head
and more numerous teeth. B. ramosi is distinguished from other Panamanian
and South American salamanders by the combination of its extensively
webbed hands and feet with the third digits long and pointed, its broad head,
and distinctive coloration (rich rusty red to medium gray-black dorsally, with
a darker venter, and a sprinkling of orange-red color on the dorsum and bright
yellow patches on the venter) .

Description of Holotype— Adult male with moderately long, truncate
snout and small nostrils. Large mental hedonic gland present (2.9 mm long
and 3.3 mm wide). Labial protuberances of nasolabial grooves large and well
developed, extending beyond margins of jaw. Moderately long canthus ros-
tralis gently arched. Head moderately broad (SL 6.4 times head width) and
moderately long (SL 4.5 times snout-gular fold length). Deep groove below

10

Contributions in Science

No. 219

eye extends for almost full length of orbit, following curvature of eye, but
does not communicate with lip. Large eyes slightly protuberant. Well-defined
postorbital groove extends posteriorly from eye as shallow depression for 1.8
mm; proceeds sharply in ventral direction at level of posterior end of mandible
and across gular area as nuchal groove, parallel to, and 3.3 mm anterior to
well-defined gular fold. Vomerine teeth number 26, arranged in single rows
that become patched laterally. Patches extend slightly beyond lateral margins
of internal nares; then row forms gentle arch to center of palate, where it is
directed posteriorly. Small maxillary teeth number 55; extending posteriorly

Figure 3. Dorsal and ventral views of paratype of Bolitoglossa ramosi (LACM
42289).

1971

New Species of Salamanders

11

to point about three-fourths through eye. Premaxillary teeth (2) well anterior
to projected curvature of maxillary tooth row; piercing lip. Moderately long
tail (0.94 times SL) with strong lateral compression, moderately constricted
at base. No postiliac glands evident. Limbs moderately long (limb interval
one); SL 4.2 times right forelimb and hind limb, and 9.7 times right foot
width. Webbing of hands and feet nearly complete, but tips of longer digits
pointed, extending beyond limits of relatively thick web. Third digit unusually
long and pointed. Hands and feet moderate in size. No subterminal pads.
Fingers, in order of decreasing length, are 3, 2, 4, 1; toes, in order of decreas-
ing length, are 3, 4, 2, 5, 1.

Measurements (in mm) are as follows: Head width, 7.1; snout to gular
fold (head length), 11.2; head depth at posterior angle of jaw, 3.8; eyelid
length, 2.8; eyelid width, 1.8; anterior rim of orbit to snout, 3.2; horizontal
orbital diameter, 2.1; interorbital distance, 2.3; distance between vomerine
teeth and parasphenoid tooth patch, 0.6; snout to forelimb, 13.8; distance
separating external nares, 2.8; distance separating internal nares, 1.8; snout
projection beyond mandible, 1.2; snout to posterior angle of vent (SL), 45.4;
snout to anterior angle of vent, 41.2; axilla to groin, 24.2; tail length, 42.8;
tail width at base, 3.2; tail depth at base, 3.7; forelimb length, 10.8; hind
limb length, 10.8; width of right hand, 3.7; width of right foot, 4.7.

Coloration of Holotype (in life).— This is a brightly colored salamander
with a rich rusty red dorsal color on head, trunk and tail. A few dark black
spots of ground color show through in some areas (especially on the snout).
The venter is a dark gray-black. Ventral surfaces of the throat, trunk, and tail
have widely scattered, irregularly shaped small spots and patches. These are
bright pale yellow. The head is mottled rusty red and black except for the
white-tipped nasolabial protuberances. The borders of the mouth are dark-
ened. Dorsal surfaces of the upper arm and leg are light red, but lower parts
of the limbs and the entire ventral side match the respective surfaces of the
trunk. Dorsal and ventral surfaces of the webbed pad are relatively dark, and
the phalanges tend to be outlined by some darker pigment dorsally. The eyes
are dark, with heavy concentrations of melanin.

Variation— Pertinent data are presented in Table 1. The males have
longer snouts than the females. No hedonic glands are evident on the two
small males but they are large and prominent on the three large ones. The
largest male (the holotype) and the three largest male paratypes have pre-
maxillary teeth which protrude from the lip. One of the small males and also
LACM 64603 differ in coloration from the remaining paratypes and the holo-
type in having a pair of broad, white stripes extending from the tips of the
nasolabial protuberances to the eyelids. The paratypes (except LACM 64603)
differ from the holotype in having a lighter gray dorsal color and light orange-
red patches about the base and first one-third of the tail, and also by having
larger pale yellow patches ventrally.

12

Contributions in Science

No. 219

Table I. j

X

CJ

Cfi

Measurements and data for specimens of new
species of Bolitoglossa

Xi

■*-» +•»
U 0)

CS _ <u o

s g £ * 5 oH -gH

> 6 1 .§ -g g’ss&eS

A-S ^ *->£ la 3 Js JTg

P M 3 M ^ ^ 68 a Oli ” fi’Scid

oa-sc $ e a ^ c ng SsSg

CW % <L> a « O d) cd 5 5 5 .9

| Limb Interval

Foot Width

B. silver stonei
LACM 42283®

3

49.3

27.0

7.8

10.8

10.6

55.7

51

24

3

5.3

B. medemi
LACM 70565

$

46.7

24.5

8.1

12.5

12.5

47.5

59

34

0.5

5.3

LACM 42278

S

41.7

21.4

7.4

12.4

12.0

34.66

43

22

0

4.3

LACM 42280

$

41.3

21.6

6.9

10.6

10.6

17.8&

33

32

0.5

3.7

LACM 70567

$

38.5

20.1

6.3

9.6

9.2

36.2

35

27

1.5

3.9

LACM 42279

$

33.7

16.7

6.0

9.3

9.4

33.2

41

29

0.5

3.3

LACM 42276®

$

48.2

26.6

8.3

12.2

12.2

36.3

41

23

1

4.8

LACM 72067c

$

47.3

26.0

7.9

11.2

11.3

36.5

50

26

1

4.8

KU 116533

$

47.0

26.3

7.3

12.0

11.8

16.P

38

50

1.5

4.4

LACM 70568

$

39.7

19.7

7.0

11.4

10.2

32.3

45

42

1

4.1

LACM 70566

$

34.2

17.8

6.1

8.9

8.0

28.7

28

27

1

3.3

LACM 42277

juv.

30.0

14.1

5.3

7.3

7.1

20.6

14

20

0.5

2.7

KU 116530

juv.

28.9

16.1

5.5

7.1

7.0

7.06

23

28

1

2,7

B. ramosi
LACM 64601®

$

45.4

24.2

7.1

10.8

10.8

42.8

55

26

1

4.7

LACM 64602

$

41.4

21.4

7.1

10.4

10.3

32.26

50

31

1.5

4.1

LACM 64603c

S

40.0

21.3

6.9

10.7

10.7

29.56

38

25

1

4.0

LACM 42290

$

37.9

19.0

6.7

9.6

9.2

29.8

47

26

1

3.2

LACM 42289

$

37.1

20.4

6.6

9.4

9.3

33.9

41

31

2

3.2

LACM 64600

$

46.7

24.7

7.8

11.2

10.8

36.0

50

51

2.5

4.3

LACM 42288

$

37.2

19.5

6.6

8.8

8.6

29.7

45

37

1.5

3.2

B. walkeri
UMMZ 128833®

3

40.2

21.3

6.4

10.2

9.8

32.86

29

22

1.5

4.5

MVZ 68628

$

41.4

23.1

6.3

9.0

8.6

19.26

33

36

2.5

3.7

MVZ 68627

$

38.9

22.0

6.2

8.7

7.9

37.0

18

28

2.5

3.2

B. equatoriana
LACM 70561

$

42.8

22.7

7.1

10.6 '

10.0

35.8

23

18

2

4.3

LACM 70562

$

40.2

21.2

7.1

10.7

10.4

33.6

26

24

1.5

4.2

LACM 70550®

$

57.9

32.8

9.1

13.2

13.2

49.6

48

11

3

5.7

UIMNH 54296

$

45.9

24.4

7.4

11.2

10.2

41.0

27

28

2

4.2

UIMNH 86692

$

44.0

23.0

7.0

10.6

10.5

35.2

27

26

2

4.1

KU 98951

9

43.0

22.8

7.5

10.6

10.1

b

23

24

2

4.3

LACM 70552

9

42.6

23.5

6.7

10.1

9.7

16.36

24

18

3

3.8

LACM 70551

9

42.4

23.1

7.0

10.3

10.0

34.9

27

21

2

4.0

UIMNH 86694

9

41.4

22.1

6.7

9.8

9.8

28.76

19

24

2

3.9

LACM 70553

9

40.3

20.9

6.8

9.4

9.6

33.7

20

19

3

3.9

LACM 70555

9

39.7

21.8

6.4

9.1

8.9

35.0

27

23

3

3.6

LACM 70556

9

39.3

21.2

6.7

9.7

9.0

29.9

14

21

1.5

4.0

LACM 70554

9

39.0

21.7

6.6

10.1

9.9

29.0

24

17

2.5

3.6

UIMNH 86696

9

38.3

20.6

6.2

9.2

9.1

22.16

23

20

2

3.9

LACM 70557

9

37.5

21.2

6.3

9.0

8.8

28.2

22

17

3

3.4

LACM 70558

9

36.8

19.7

6.1

8.9

9.0

29.3

19

17

2.5

3.5

UIMNH 86695

9

36.5

19.3

6.3

8.7

8.6

29.2

27

23

2

4.0

LACM 70559

juv.

34.0

18.9

6.2

8.6

8.3

28.3

18

16

2

3.3

LACM 70560

juv.

32.4

16.7

5.9

7.9

7.3

25.0

4

16

2

3.0

°holotype; ^regenerated tails or tails missing; Ccleared and stained.

1971

New Species of Salamanders

13

Osteology— Information has been derived from one cleared and stained
adult male (LACM 64603) and from stereoscopic radiographs of all adults
available.

The skull is well formed and bones in the posterior portion are closely
sutured. The snout is short and anterior cranial elements are small, with
slight or no articulations. In comparison with B. medemi the snout region is
poorly developed. The premaxilla is small and slender, with short, distally
expanded frontal processes. The processes are well separated for their entire
length, but the internasal fontanelle is very small. The irregularly expanded
terminal parts of the processes are small and barely overlap the anterior ends
of the frontals. Lateral parts of the frontals extend anteriorly so that the tips
of the processes lie more or less enclosed by the frontals. The processes extend
beyond both the anterior border of the orbit and the posterior margin of the
nasals. Nasal bones are of moderate size and, relative to the premaxilla, they
are strongly protuberant. Their only articulation is by means of a ventrolateral
lobe which barely contacts the facial process of the maxilla. The pointed pos-
terior tips of the nasal bones extend to the anterior border of the orbits. The
separation between the nasals is great, approximating their length. The pos-
terolateral margins of the nasals and the anterodorsal margins of the facial
process of the maxilla are evacuated for the passage of the nasolacrimal duct.
Prefrontal bones are very erratic in shape, and they are very small. In the one
cleared specimen the prefrontal of one side is an elongate bone with about
one-quarter the area of the facial process of the maxilla and less than one-
tenth the area of the nasal. On the other side the bone is reduced to a tiny dot
that is less than one-tenth the size of its pair. Prefrontals have no contacts
with other bones in this species. The maxillae extend about three-quarters
through the eye. They are very slender, with well developed facial processes.
The anterior ends of the maxillae are flattened, terminating in narrowed
points. Palatal processes are poorly developed.

Vomers are of moderate size and are completely separated from each
other. The toothed portions are drawn into processes medially, where they
converge, but remain well separated. The intervomerine fontanelle is very
large. Preorbital processes extend beyond the lateral margins of the vomerine
bodies. Vomerine teeth are in a single row that extends beyond the lateral
margin of the internal nares.

Frontals are large and well sutured to each other. Facial portions are
relatively smaller than B. medemi. Posteriorly the bones are well sutured to
the parietals. There are slight lateral lobes on the posterior margins of the
frontals. Parietals are well developed and closely articulated with each other.
There are no crests on the occipito-otics. The nearly vertical squamosals rest
in depressions in ridges on the lateral margins of the occipito-otics. The large
parasphenoid has a rather narrow, blunt-tipped anterior end. The orbitosphe-
noids are well separated at their ventral margins. Posterior vomerine tooth
patches are not in contact medially. The right patch bears 74 and the left, 82

14

Contributions in Science

No. 219

ankylosed, bicuspid teeth in the cleared specimen. The operculum has no
stilus. Quadrates and squamosals are moderately developed.

There are one cervical, fourteen trunk, one sacral, two caudosacral and
28 (LACM 42289, 42290, 64601), 27 (LACM 64600), 25 (LACM 42288),
23 (LACM 6460, tip regenerated), or 19 (LACM 42287, juvenile) caudal
vertebrae; the tail of LACM 64603 is regenerated beyond caudal vertebra
seventeen. Ribs are present on all but the last trunk vertebra. The transverse
processes on the first caudosacral vertebra are very long and are oriented
nearly perpendicularly to the body axis. Those of the second caudosacral ver-
tebra are shorter and extend sharply in an anterior direction. The very long,
unbranched processes of the first caudal vertebra are slanted anteriorly. The
slant is sharper than on the preceding vertebra. Their tips extend nearly to a
level equivalent to the anterior end of the second caudosacral vertebra. The
processes of these adjacent vertebrae do not overlap. Processes on the second
caudal vertebra are much smaller than those on the first, and they become
progressively smaller on the remaining vertebrae. The last vertebra to have
distinct processes varies from the eighth to the eighteenth. The last caudo-
sacral and first caudal vertebrae are shorter than neighboring vertebrae. Ver-
tebrae in the anterior one-half of the tail are as long as any but the first three
trunk vertebrae. The fourteenth caudal is the first vertebra that is shorter
than the first caudal.

The tibia has a distinct crest but no spur. Phalangeal formulae are 1, 2,
3, 2 (or 1), and 1, 2, 3, 3 (or 2), 2 (or 1). The more distal phalanges are
poorly developed, but there is a tendency for reduction and loss (Fig. 2).
Terminal phalanges are extremely small and poorly ossified, with erratic
shapes. Penultimate phalanges are reduced in the longer toes. Proximal pha-
langes are short and stout, often as broad as long. They are somewhat flat-
tened. The distance between bony areas of a given digit is great, and often the
cartilage between elements is longer than the adjacent bones. Metatarsals and
metacarpals are flattened, with lateral bony webs. The outermost metapodials
have characteristic shapes resulting from a large, rounded web along the
margin of the bones. There are seven carpals and seven or eight tarsals. In
one tarsus, D 4-5 is fused with D 3, and D 1-2 is partly mineralized in several
tarsi (Fig. 2).

Remarks. —All specimens were collected either in bromeliads located
within a few feet of the surface, or in the rolled bases of palm fronds on the
surface of a forested hill. In 1968 the specimens were collected in sympatry
with Bolitoglossa vallecula, a species that is widespread in the northern part
of the Cordillera Central of Colombia (Brame and Wake, 1963). No B.
vallecula were found associated with B. ramosi during the 1971 visit.

Range — Known only from the type locality in the Cordillera Central of
Colombia (Fig. 8).

The following most distinctive of the new species is named in honor of

1971

New Species of Salamanders

15

Philip A. Silverstone, in appreciation of his assistance to us and in recognition
of his important contributions to Neotropical herpetology.

Bolitoglossa silverstonei, new species
Figures 4 and 5

Holotype .— LACM 42283, an adult male from Quebrada Bochorama,
Loma de Encarnacion, Departamento de Choco, Colombia, about 51 km (32
mi) SE Quibdo at approximately 76° 23' W, 5° 20' N. This site is a “one-
hour walk” SE Playa de Oro. The specimen was collected in a rolled planta-
nillo leaf on a steep hillside near a stream at about 400 m (1312 ft) elevation
by Philip A. Silverstone and Jorge E. Ramos on May 31, 1968. The species
is known only from the holotype.

Diagnosis— A moderate-sized species (49.3 mm SL) with moderate
numbers of maxillary (51) and vomerine (24) teeth; distinguished from B.
biseriata by its larger feet and more numerous maxillary teeth; from B. sima
by its shorter legs and more numerous maxillary teeth. Bolitoglossa silver-
stonei is distinguished from other Panamanian and South American species
by the combination of its extensively webbed hands and feet, distinctive ven-
tral coloration (cream with a light peppering of small brownish spots), and
size and dentitional features.

Description of Holotype— Adult male with moderately long, somewhat
truncate snout and small (2.0 mm wide), nearly circular mental hedonic
gland and small nostrils. Labial protuberances of nasolabial grooves mod-
erately large, extending below lower jaw margin. Head moderately broad
(SL 6.4 times head width) and long (SL 4.3 times snout-gular fold length).
Deep groove below eye extends for almost full length of orbit, following
curvature of eye, but does not communicate with lip. Eyes moderately small,
slightly protuberant. Well-defined postorbital groove extends posteriorly from
eye as shallow depression for 2.2 mm, then sharply ventrad at level of pos-
terior end of mandible and across gular area as nuchal groove, parallel to,
and 4.2 mm anterior to sharply defined gular fold. Vomerine teeth number
24, in moderately patchy rows that extend slightly beyond lateral borders of
internal nares. From lateral terminus, rows extend medially in nearly straight
line to near center of palate, then bend sharply posteriad and closely approach
(1.0 mm separation) parasphenoid tooth patch. Small maxillary teeth num-
ber 51; extending posteriorly to point about three-fourths through eye. Large
premaxillary teeth (two) pierce lip. Long tail (1.1 times SL) rounded and
moderately constricted at base. Postiliac glands indistinct. Limbs moderately
short (limb interval three). Standard length 4.6 times right forelimb, 4.5
times right hind limb, and 9.3 times width of right foot. Webbing of hands
and feet extensive, nearly complete, with only tips of longer digits extending
slightly beyond web (Fig. 5). No subterminal pads present. Fingers, in order

16

Contributions in Science

No. 219

of decreasing length, are 3, 2, 4, 1; toes, in order of decreasing length, are
3,4, 2,5, 1.

Measurements (in mm) are as follows: Head width, 7.8; snout to gular
fold (head length), 11.5; head depth at posterior angle of jaw, 4.8; eyelid
length, 3.2; eyelid width, 2.0; anterior rim of orbit to snout, 3.8; horizontal
orbital diameter, 2.7; interorbital distance, 3.3; distance between vomerine
teeth and parasphenoid tooth patch, 1.0; snout to forelimb, 14.5; distance
separating internal nares, 2.4; distance separating external nares, 3.0; snout
projection beyond mandible, 0.9; snout to posterior angle of vent (SL), 49.3;
snout to anterior angle of vent, 45.1; axilla to groin, 27.0; tail length, 55.7;
tail width at base, 3.9; tail depth at base, 3.9; forelimb length, 10.6; hind limb
length, 10.8; width of right hand, 4.1; width of right foot, 5.3.

Coloration of Holotype (in alcohol).— This is a rather light-colored sala-
mander. A reddish brown dorsal mottling overlies the blackish purple ground
color. The dorsal pigmentation is distinctly darker than that of the ventral
surfaces. Lateral surfaces of the trunk and tail are light reddish brown with
some scattered melanophores. An indistinct ventrolateral stripe of blackish
purple sharply separates the dark dorsal and lateral from the light ventral
pigmentation. The broad stripe of the trunk becomes narrow and discontinu-
ous on the tail. All ventral surfaces are light golden cream to grayish white,
peppered with minute, widely scattered melanophores that are clearly visible
over the entire surface. Some coalescence of melanophores occurs laterally,
producing larger spots of pigment. The head is colored like the trunk, dark
dorsally and light ventrally. The whitish ventral coloration of the throat ex-
tends along the upper lip region and in front of the eyes. The small eyes have
a reddish brown iris, with a gold ring surrounding the horizontally elliptical
pupil. Limbs are dark dorsally and light ventrally, with other markings similar
to the respective parts of the trunk. The hands and feet are light dorsally and
ventrally, and there are no obvious ventral melanophores.

Osteology— Stereoscopic radiographs have provided all of the following
information. The skull is comprised of well-articulated bones and is generally
well developed. The premaxilla has a very small dental process which is
placed well ahead of the maxillae. Frontal processes of the premaxilla are
separated for their entire lengths. The processes are large and expanded near
their tips. Nasal bones are large and strongly protuberant. No prefrontal
bones can be seen. Vomers are well separated on the midline. The preorbital
processes of the vomers extend laterally well beyond the limits of the internal
nares. No stilus is present on the middle ear bone. Ribs are present on all but
the last trunk vertebra. There are one cervical, fourteen trunk, one sacral,
two caudosacral, and 38 caudal vertebrae. The transverse processes on the
first caudosacral vertebra are very long and slant posteriorly. Those on the
second caudosacral vertebra are much shorter and slant slightly in an anterior
direction. The long, unbranched processes of the first caudal vertebra arise
near its anterior end and extend sharply in an anterior direction. They termi-

1971

New Species of Salamanders

17

Figure 4. Dorsal and ventral views of paratype of Bolitoglossa silverstonei (LACM
42283).

nate beyond the point of attachment of the processes of the second caudo-
sacral vertebra. The tips of the processes of the last caudosacral and first
caudal vertebrae do not cross. Transverse processes of the remaining caudal
vertebrae are progressively smaller. All are located near the anterior end of
the vertebrae and are oriented sharply forward. They are discrete on the first
30 vertebrae. As is usual in species with constricted tail bases, the last caudo-
sacral and first caudal vertebrae are shortened, relative to neighboring verte-
brae. Posterior to this region the vertebrae are longer. The fourth through
seventh caudal vertebrae are as long as the longest trunk vertebrae (two and
three). The first vertebra shorter than the first caudal vertebra is the seven-
teenth caudal. From that point the vertebrae are progressively shorter to the
tail tip. No tibial spurs are present. Phalangeal formulae are 1, 2, 3, 2 and
1, 2, 3, 3, 2. All phalangeal elements are small and poorly developed, and the
abrupt decrease in size from the proximal to the distal elements in the longest
digits is striking (Fig. 5). Terminal phalanges are all minute and unexpanded;
most are tiny points of bone. Much cartilage is present at the ends of the

18

Contributions in Science

No. 219

metapodials and phalanges, and the distance from one bony area to another
is always greater than the length of the distal bony element. Metapodials are
dumbbell-shaped with only slight lateral expansion.

Remarks — Playa de Oro is located in the Choco forest area of Colombia,
in a region identified as wet tropical forest (Holdridge System) by Espinal
and Montenegro (1963). Previously only B. biseriata has been known from
this area, but discovery of B. silverstonei, B. medemi, and B. phalarosoma
from northwestern Colombia and from Panama suggests that this has been
a region of lowland diversification. Two additional species, B. sima and B.
chica, occur in the Ecuadorian portion of the Choco.

Range— Known only from the type locality in the lowlands of north-
western Colombia (Fig. 8).

During the past ten years we have been generously aided in our efforts
by the cooperation and encouragement of Professor Charles F. Walker of the
Museum of Zoology, University of Michigan. It is a pleasure to name the
following Colombian species in his honor.

Bolitoglossa walkeri, new species
Figures 5 and 6

Holotype — UMMZ 128833, an adult male from “Television Tower
Mountain,” 15 km WNW Cali and 0.9 km S El Jordan, Departamento de
Valle, Colombia. The specimen was collected from a bromeliad in cloud
forest at an elevation of 2050 m (6724 ft) by Walter Moberly and Kraig K.
Adler on July 17, 1965.

Paratypes — MVZ 68627-28, 4 km NW San Antonio, Depto.Valle, Colom-
bia, 1982 m (6500 ft) elevation.

Diagnosis— An apparently small species (3’ adults 38.9-41.4, mean 40.2
mm SL) with low numbers of maxillary (mean 27) and moderate numbers
of vomerine (mean 28) teeth. Distinguished from B. equatoriana by its nar-
rower head, less extensively webbed, slightly smaller hands and feet, and
ventral coloration (dirty white to gray, with some streaks of darker pigment
and an overlay of brassy pigment, but without the encroachment of dark pig-
ment which leaves the large, whitish spots characteristic of B. equatoriana)',
from B. medemi by its narrower head, less extensively webbed feet, and
shorter legs; from B. ramosi by its narrower head and less numerous teeth.
Bolitoglossa walkeri differs from other Panamanian and South American
Bolitoglossa by the combination of its extensively webbed hands and feet,
color, and its size and dentitional features (Table 1).

Description of Holotype — Adult male with moderately short, truncate
snout and pronounced, rounded, mental hedonic gland; small nostrils. Labial
protuberances of nasolabial grooves well developed, extending below lower
jaw margin. Strongly arched canthus rostralis moderately long. Head mod-

1971

New Species of Salamanders

19

erately broad (SL 6.3 times head width) and long (SL 4.4 times snout-gular
fold length). Deep groove below eye extends for almost full length of orbit,
following curvature of eye, but does not communicate with lip. Eyes relatively
large, moderately protuberant. Well-defined postorbital groove extends pos-
teriorly from eye as shallow depression for 1.8 mm, then sharply ventrad at

Figure 5. Outlines of hands and feet of three species of Bolitoglossa, drawn from
radiographs through use of microprojector. Bony parts of digits are outlined, a. Right
hand of holotype of B. walkeri (UMMZ 128833). b. Right foot of holotype of
B. walkeri. c. Right foot of holotype of B. silverstonei (LACM 42283). The left
side of the drawing is distorted as a result of fixation artifact, d. Right foot of
B. equatoriana (KU 98951). The foot is slightly distorted on the left and slightly
foreshortened as a result of fixation artifact.

20

Contributions in Science

No. 219

Figure 6. Dorsal and ventral views of holotype of Bolitoglossa walked (UMMZ
128833).

level of posterior end of mandible and across gular area as nuchal groove,
parallel to, and 3.2 mm anterior to sharply defined gular fold. Vomerine teeth
number 22, arranged in single rows extending from one-half to two-thirds
diameter of internal nares; from lateral terminus, rows form moderately strong
arches to center of palate, where two rows nearly meet. Small maxillary teeth
number 29; extending posteriorly to point about one-half through eye. Single
premaxillary tooth pierces lip. Relatively short tail (0.82 times SL) has slight
lateral compression and is slightly constricted at base. Postiliac glands small,
indistinct. Limb length moderate with limb interval of 1.5. Standard length

1971

New Species of Salamanders

21

4.1 times right forelimb, 3.9 times right hind limb, and 9.0 times width of
right foot. Webbing of hands and feet extensive, thin. Tips of all digits dis-
cernible, longer digits protruding substantially from web. Digital tips broadly
rounded. Hands and feet relatively large. No subterminal pads. Fingers, in
order of decreasing length, are 3, 4, 2, 1; toes, in order of decreasing length,
are 3, 4, 2, 5, 1.

Measurements (in mm) are as follows: Head width, 6.4; snout to gular
fold (head length), 9.2; head depth at posterior angle of jaw, 3.6; eyelid
length, 3.2; eyelid width, 1.8; anterior rim of orbit to snout, 3.0; horizontal
orbital diameter, 2.3; interorbital distance, 2.8; distance between vomerine
and parasphenoid teeth, 0.5; snout to forelimb, 11.8; distance separating in-
ternal nares, 2.0; distance separating external nares, 2.7; snout projection
beyond mandible, 1.1; snout to anterior angle of vent, 35.8; snout to posterior
angle of vent (SL), 40.2; axilla to groin, 21.3; tail length, 32.8; tail width at
base, 2.8; tail depth at base, 3.2; forelimb length, 9.8; hind limb length, 10.2;
width of right hand, 3.6; width of right foot, 4.5.

Coloration of Holotype In life (from field notes of Kraig Adler) :
“Golden brown above in different shades, with blackish spots and blotches.
Cream white streaks running lengthwise, especially over vent and on tail;
black “V’s” on neck, pointing outwards; some faint reddish pigment on
dorsum, especially in midline. Dark golden below, light tan golden between
eye and nasolabial groove, nose region speckled with various shades of golden.
Belly dirty white overlaid with much brassy pigment; some few black streaks,
also at posterior end of anus [sic]; throat heavily flecked with golden, espe-
cially at anterior end; mental gland bright golden; soles of hands and feet
pinkish ( = blood) and golden.” After several years in alcohol the brighter
pigments have faded, but the pattern remains distinct. The impression is of a
rather dark tannish brown animal with a much lighter venter. The whitish
ventral pigment is more sharply demarcated from the lateral dark pigment of
the tail than of the trunk. The mental gland is light and prominent on the
relatively dark throat.

Variation. —Pertinent data are presented in Table 1. The holotype is a
male and the two paratypes are females. The paratypes have proportionally
shorter limbs (limb interval 2.5 rather than 1.5) and narrower feet (SL
11.2-12.2 times right foot width, rather than 9.0) than the holotype. Both
features are sexually dimorphic in similar ways in most species of Bolito-
glossa. Premaxillary teeth are absent in one paratype and fail to protrude
from the lip in the other; these are also female characteristics.

One specimen, MVZ 68628, is colored like the holotype, but has a some-
what darker venter which lacks dark streaks, whereas MVZ 68627 has a
lighter dorsal ground color than the holotype, but has a similar ventral colora-
tion. This animal is briefly described in the field notes of the collector, A. H.
Miller, as follows: “The light areas of the back, belly and undertail surface

22

Contributions in Science

No. 219

were yellow or bronze in life so that the whole animal was distinctly light and
bright.”

Osteology All information has been derived from stereoscopic radio-

graphs. The skull is well developed with well articulated bones. The small,
slender premaxilla has relatively short, divergent frontal processes. The dilated
tips of these processes fall short of the posterior margin of the nasals. The
large, protuberant nasals have distinct lateral lobes that articulate firmly with
the maxillae. No prefrontals are evident. Vomers are well separated for their
entire lengths, but the toothed parts approach the midline posteriorly. Pre-
orbital processes of the vomers extend well beyond the lateral margins of the
internal nares and bear teeth for most of their lengths. Maxillae extend about
to the posterior margin of the eyes. The operculum has no stilus. Ribs are
present on all but the last trunk vertebra, but those on the next to last vertebra
are very small in the holotype. There are one cervical, fourteen trunk, one
sacral, two caudosacral and 28 caudal vertebrae in the single specimen that
has a complete tail.

The long, stout, transverse processes on the first caudosacral vertebra
are nearly perpendicular in orientation, but have a slight posterior slant. The
shorter and more slender processes on the second caudosacral vertebra have
a sharp anterior slant. These processes are stouter and less slanted in the holo-
type than in the paratypes. Their tips reach to a point about one-third through
the preceding vertebra. The very long processes of the first caudal vertebra
are long and sinuous. They slant strongly in an anterior direction. Tips of the
processes extend beyond the bases of the processes on the second caudosacral
vertebra, but the processes of the adjacent vertebrae do not overlap. The
processes are not branched. Processes on succeeding vertebrae are progres-
sively shorter. They are visible to about the nineteenth vertebra, but are minute
beyond the seventh. All lie at the anterior end of the vertebrae and slant
anteriorly.

The second caudosacral and first caudal vertebrae are shorter than all
but the first trunk vertebra, which equals them in length, and the seventeenth
and succeeding caudal vertebrae. The second through eighth caudal vertebrae
are as long as the longest trunk vertebra (the seventh), and the third caudal
vertebra is the longest in the entire column.

No tibial spur is present, but a small ridge is present in mid-shank on the
left tibia in the holotype. Phalangeal formulae are 1, 2, 3, 2 and 1, 2, 3, 3, 2.
Digits are well developed. Phalangeal elements are increasingly shortened
toward the digital tip. Most are dumbbell-shaped. Terminal phalanges are
expanded at their tips. Distance between the bony parts of the digits is always
less than the length of the shortest phalanx of the digit. Lateral weblike proc-
esses of the metatarsals extend into the fleshy web (Fig. 5).

Remarks— All of the specimens were collected in cloud forest at inter-
mediate elevations. The holotype was taken from a bromeliad. Alden and
Virginia Miller collected MVZ 68627 during the day (March 9, 1958) while

1971

New Species of Salamanders

23

it was exposed on the surface of a large (five inch) leaf hanging from a flower-
ing epiphyte that was not noticeably moist. Dr. Miller collected the other
paratype (MVZ 68628) during the day (September 9, 1958) in a brushy,
thick part of the forest. It was apparently dislodged from its position in the
foliage, since it was found on the ground as a path was retraced. The indi-
vidual had, as yet, not righted itself.

Range.— This species is known only from neighboring localities in cloud
forest of intermediate elevation (about 2000 m) WNW of Cali, Depto. de
Valle, Colombia (Fig. 8).

Examples of an undescribed species of salamander have been collected
in sympatry with Bolitoglossa peruviana on several occasions. This species,
named for its geographic location, is the sixth form discovered in Ecuador.

Bolitoglossa equatoriana, new species
Figures 5 and 7

Holotype — LACM 70550, an adult female from Limon Cocha, 0° 24' S,
76° 37' W, Provinicia de Napo, Ecuador. The specimen was collected at a
secondary-primary growth border, 1 m above the ground, on August 5, 1971
by W. Ronald Heyer. Elevation 260 m (850 ft) .

Paratypes — LACM 70551-64 (14 specimens) collected by W. Ronald
Heyer between June 11 and August 5, 1971; KU 98951, UIMNH 86692,
UIMNH 86694-96 collected by different collectors between July 1 8 and July
28, 1965 at the type locality.

Diagnosis— A moderate-sized species (15 females: 36.5-57.9, mean 41.7
mm SL; two males: 40.2-42.8, mean 41.5 mm SL) with low numbers of
maxillary (mean 25) and moderate numbers of vomerine (mean 21) teeth.
Distinguished from B. walked by its broader head, more extensively webbed
and slightly longer hands and feet, and spotted ventral color pattern; from
B. medemi by its less numerous maxillary teeth and somewhat shorter legs;
from B. ramosi by its narrower head and less numerous maxillary teeth; from
B. peruviana by its broader head, larger hands and feet, and less numerous
maxillary teeth, as well as by its spotted ventral color pattern; and from
B. altamazonica by its broader head, larger hands and feet, and spotted ventral
color pattern. Bolitoglossa equatoriana is distinguished from all other Pana-
manian and South American species of Bolitoglossa by the combination of
its extensively webbed hands and feet, coloration, and its size and dentitional
features (Table 1).

Description of Holotype— Adult female with moderately short, relatively
broad, truncated snout. Nostrils rather small, nasolabial protuberances mod-
erately developed. Slightly arched canthus rostralis of moderate length. Head
of moderate width (SL 6.4 times head width) and length (SL 4.3 times snout-
gular fold length). Deep groove below eye extends for almost full length of
orbit, following curvature of eye, but does not communicate with lip. Mod-

24

Contributions in Science

No. 219

erately large eyes only slightly protuberant. Well-defined postorbital groove
extends posteriorly from eye as shallow depression for 2.8 mm; then sharply
ventrad at level of posterior end of mandible and across gular area as nuchal
groove, parallel to, and 5. 1 mm anterior to sharply defined gular fold. Vomerine
teeth number 1 1 , arranged in single rows that extend to center or to lateral
margin of internal nares. Slightly arched rows extend nearly to midline on
palate, but have no posterior extension. Small maxillary teeth number 48;
extending to point about one-half through eye. Three premaxillary teeth.
Relatively* short tail (0.86 times SL) is laterally compressed, with strong
basal constriction. Postiliac glands not evident. Limbs are of moderate length
(limb interval 3); SL 4.4 times right forelimb, 4.4 times right hind limb, and
10.2 times right foot width. Webbing of hands and feet extensive, moderately
thick. Finger and toe tips, especially of third digits, protrude substantially
from webbed pad. Tips of third fingers and toes pointed, others rounded. No
subterminal pads. Hands and feet moderately large. Fingers, in order of
decreasing length, are 3, 4, 2, 1; toes, in order of decreasing length, are
3, 4, 2, 5,1.

Measurements (in mm) are as follows: Head width, 9.1; snout to gular
fold (head length), 13.4; head depth at posterior angle of jaw, 4.6; eyelid
length, 3.8; eyelid width, 2.0; anterior rim of orbit to snout, 3.7; horizontal
orbital diameter, 2.2; interorbital distance, 3.3; distance between vomerine
teeth and parasphenoid tooth patch, 0.8; snout to forelimb, 16.7; distance
separating internal nares, 2.4; distance separating external nares, 3.1; snout
projection beyond mandible, 1.0; snout to posterior angle of vent (SL), 57.9;
snout to anterior angle of vent, 53.4; axilla to groin, 32.8; tail length, 49.6;
tail width at base, 3.7; tail depth at base, 4.4; forelimb length, 13.2; hind limb
length, 13.2; width of right hand, 4.2; width of right foot, 5.7.

Coloration of Holotype (in alcohol).— The dorsal color consists of a dis-
tinct though irregular beige to gray dorsal band divided down the middle by
a dark blackish brown thin stripe of ground color. The ground color of the
lateral surfaces is much darker than that of the dorsal and ventral surfaces.
The venter appears somewhat light because of the many tiny bluish white
cells covering much of the blackish ground color. The tiny spots are grouped
together as patches on the last three-fourths of the tail venter. The front of
the head is a medium brown and the back of the head is covered by the
anterior end of the dorsal band. The hind limbs have a considerable amount
of beige and gray to brown mottling dorsally, but the dorsal area of the front
limbs is mostly a blackish brown ground color. The inside half of the hands
and feet are covered dorsally with many tiny light colored spots overlying the
ground color. The limbs all have some of these spots ventrally but the ventral
surfaces of the hands and feet are an immaculate gray-black.

Variation— Pertinent data are presented in Table 1. The female holotype
is considerably larger (57.9 mm SL) than the largest paratype (45.9 mm SL).
Most of the paratypes are adult females except for two males and four juve-

1971

New Species of Salamanders

25

niles. Standard length is from 5.7 to 6.4 (mean 6.1) times head width in the
entire sample. Limb length is somewhat variable in the series, and the holo-
type has relatively broad hands and feet (SL 10.2 times right foot width in
holotype, 9.1 to 11.2, mean 10.4 in paratypes). The holotype has the third
longest tail (0.86 times SL, versus 0.58 to 0.89, mean 0.79 in paratypes).

1 Cm

Figure 7. Dorsal and ventral views of a paratype of Bolitoglossa equatoriana
(UIMNH 54296).

26

Contributions in Science

No. 219

Digital tips of the holotype are like most of the paratypes. The tip of digit
three is pointed with the other digits having more rounded tips. There is
considerable variation in dorsal coloration. The holotype has a distinct though
irregular beige to gray dorsal band divided by a dark blackish brown stripe of
ground color down the middle; one other specimen has a uniform broad beige
dorsal band; six specimens have weak or partial dorsal bands of beige to gray;
two have a few dorsal light brown patches; two are uniform blackish brown
dorsally. The type series varies in ventral coloration as follows : five specimens
have a few to moderate numbers of mostly small, bluish silver to white spots
or patches; thirteen have an irregular row of moderate-sized, bluish silver to
white patches on either side of the midline (see paratype ventral view, Fig. 7) ;
and the holotype is covered ventrally by hundreds of tiny iridophores.

Osteology— All information has been derived from stereoscopic radio-
graphs of the type series. The skull is well ossified and the bones are well
articulated. The premaxilla is small and slender but frontal processes may be
well developed. The frontal processes of UIMNH 54296 are slender and are
not expanded at their tips. Those of some other specimens are expanded at
their tips. The processes are separated for their entire length in all specimens.
The large, protuberant nasals have a strong, extensive articulation with the
maxillae. Prefrontals are definitely present in some individuals but absent in
others. Vomers are well separated for their entire lengths. Preorbital processes
of the vomers extend laterally well beyond the limits of the internal nares.
Maxillae are of moderate size and extend about to the posterior margin of the
eyeball. There is no stilus on the operculum. All but the last trunk vertebrae
bear ribs. There are one cervical, fourteen trunk, one sacral, two caudosacral
and from 23 to 30 caudal vertebrae in those specimens with complete tails.
The long, stout transverse processes of the first caudosacral vertebra are
directed nearly perpendicularly to the body axis, but with a slight posterior
slant. The shorter processes on the second caudosacral vertebra slant in an
anterior direction. Very long, unbranched processes are present on the first
caudal vertebra, and these slant sharply toward the head. They do not cross
the processes of the second caudosacral vertebra even though they extend in
front of the base of the latter. Processes of succeeding vertebrae are progres-
sively smaller. They lie at the anterior end of each vertebra and slant anteri-
orly. Caudal transverse processes are visible as far as the seventeenth vertebra
in one adult, but they are small and highly variable in degree of development
past the tenth vertebra. In the basal part of the tail the vertebrae increase in
length, and the fourth and fifth caudal are as long as the longest (anterior two
to six) trunk vertebrae in some, but a little shorter in others. About the thir-
teenth caudal is the first that is shorter than the three vertebrae immediately
behind the sacrum. Vertebrae are progressively shorter from that point to the
tail tip. No tibial spurs are present. Phalangeal formulae in some individuals
are 1, 2, 3, 2; 1, 2, 3, 3, 2. A tendency toward phalangeal reduction is apparent
and formulae may be 1, 2, 3, 1 and 1, 2, 3, 2, 1 in extreme instances. Terminal

1971

New Species of Salamanders

27

phalanges are usually small and short, often being broader than long. They
are rounded at their tips, and usually neither pointed nor expanded. Penulti-
mate phalanges of the longest digits are small and often broader than long.
Distance between the bony parts equals or surpasses the length of the penulti-
mate phalanges in the longest digits. Terminal phalanges are shorter and
smaller than penultimate ones in most instances. Metapodials are flat and
broad, with some lateral bony growth extending into the fleshy web.

Remarks— Found between 7:30 and 9:30 pm, from 0.5 to 2 m (IV2 to
6V2 ft) above the ground on broad leaves, palm leaves and stems, along the
stream banks in secondary growth, secondary-primary border, and in agricul-

Figure 8. Distribution of five new species of Bolitoglossa in Panama, Colombia,
and Ecuador. Symbols: • Bolitoglossa medemi\ O ramosi\ A B. silver stonei\
A B. walkeri; * B. equatoriana.

28

Contributions in Science

No. 219

tural clearings. It occurs in sympatry with the smaller and more slender B.
peruviana.

Range— Known only from the type locality in the Amazonian lowlands
of Ecuador (Fig. 8).

Discussion

In 1963 we discussed the relationships of the South American members
of Bolitoglossa, and our views have recently been elaborated (Wake and
Brame, 1966; Wake, Brame and Myers, 1970). The continued discovery of
new populations and undescribed species points up the tentative nature of
such discussions and the need for continuing revision. Nevertheless, it is use-
ful to present our current views concerning species relationships, even in a
developmental state, for they may aid in planning research projects and in
zoogeographic work. Small samples and incomplete knowledge make full
documentation impossible. However, we can present the basis for our char-
acter analysis and the kind of reasoning used.

Characters subject to interspecific variation are divided into discrete
states for analytical purposes. Direction of character state change is deter-
mined in several instances. Usually this is based on out-group comparisons,
with conditions that are present in more generalized relatives (such as the
extratropical plethodontids) considered to be primitive. In some instances it
is possible to identify one or more highly specialized states, but operationally
the primitive state is identified by following the trend in specialization back to
the simplest or most generalized condition by phenetic methods. Correlation
of character state trends with other trends, for example, geographic patterns,
is sometimes used in initial analysis. We have come to expect more ancestral
states among northern and upland members of a given group, and derived
states in species that are southern, lowland, or both. Finally, as a working
hypothesis we expect the more derived state of a given character to be present
in species in which derived states of many other characters are present. This
last criterion involves some circularity in reasoning, and is used only tenta-
tively and when other criteria are not applicable. Our knowledge of the neo-
tropical salamanders is not sufficient to detect all of the parallelism and
convergence which we suspect are present, and the fact that few characters
are used in our analysis increases our chance of error. Hopefully future work
will improve our ability to detect these phenomena, and will also increase the
number of characters, thus diminishing the chance of error. Larger series will
permit quantification and the use of continuously variable characters.

Characters :

Size— Average adult size is small (ca. 40 mm SL), moderate (50-60
mm SL), or large (ca. 70 mm SL). Intermediate conditions (e.g., moderately
small) are recognized. Moderate size is characteristic of many generalized
neotropical salamanders, and is probably close to the ancestral condition.

1971

New Species of Salamanders

29

Either extreme is considered derived, but examples of parallelism are sus-
pected. The character has low reliability on a genus-wide basis, but may be
of use within a species group established on other grounds.

Structure of Hands and Feet— This is one of the most complex and use-
ful sets of characters. Much information can be derived from detailed consid-
eration of both external and internal structure of the appendages. Categories
of foot-webbing have been outlined previously (Wake and Brame, 1969).
The primitive hand and foot has little webbing, large and discrete digits and a
full complement of phalangeal and mesopodial elements. Terminal phalanges
are primitively large and well developed, and cutaneous subterminal pads are
well developed. Derived characters include increase in webbing, decrease in
number and size of phalangeal and mesopodial elements (through loss and
fusions) and loss of digital integrity. Many types of reduction trends, all con-
sidered to be derived, are found. These include reduction in size, or loss, of
the subterminal pads; reduction in size and degree of development of certain
phalanges, for example, the terminals; reduction in total phalangeal bone
relative to metapodial bone; disproportionate digital reduction, for example
the central relative to the first digit. Also important are the shape of the toe
tips, the cutaneous outline, the degree of flattening, and the proportions of
limbs, feet and digits.

Numbers of Maxillary and Vomerine Teeth— Numbers of maxillary
teeth in adults are low (mean 0-30), moderate (30-60), or high (above 60).
Similar categories for vomerine teeth in adults are low (0-20), moderate
(20-30), and high (above 30). Moderate numbers characterize generalized
relatives and are considered ancestral; both extremes are derived. Teeth in-
crease in number with increasing size, but at different rates in different species.
The values given here are not absolutes, but must be considered relative to
size of the species. Thus the number of teeth in adults of a small species may
be considered to be high, while the same number for a large species might be
considered moderate or even low. Eventually we hope to deal with such onto-
genetically variable characters in a more satisfactory manner.

Head Width— Heads are narrow (greater than 6.7 times SL), moderate
(6. 3-6.6), or broad (less than 6.3). The character must be used with caution,
since the proportion changes with age and size, to a degree. Moderate heads
are closest to the ancestral condition and either extreme is considered to be
derived.

Coloration.— We are unable to break the color continuum into discrete
states. Nevertheless, certain features, such as unusual pigmentation, bands,
stripes, spotting and streaking patterns, etc., are frequently used when com-
paring species within groups.

Behavioral Attributes— Terrestrial habits are considered to be primitive
for Bolitoglossa. Arboreal habits, varying from a tendency toward arboreality
to complete arboreality, are considered to be derived.

Comparative Osteology Large numbers of osteological features have

30

Contributions in Science

No. 219

potential value in systematic work (see Wake, 1966), but unfortunately the
small samples available for most South American species preclude extensive
use of such characters. In general, any reduction, fusion, loss or elaboration
of the presumed ancestral condition is considered derived. Osteological fea-
tures that are used for these species include presence or absence of prefrontal
bones and tibial spurs, shape of the nasals, premaxillae, maxillae, and vomers,
and arrangement of the transverse processes of the caudosacral and caudal
vertebrae. Features associated with the hands and feet have been discussed
above.

Certain other characters are used within species groups for the purposes
of comparing close relatives in some detail. These include some relatively
subjective features, such as snout shape and degree of protuberance of the
eyes, as well as proportional relationships, such as relative leg, tail, and head
lengths.

Most characters have been used in a phenetic manner, and those species
which have high similarity are considered to be more closely related than
those with low similarity. All of the following groups have been recognized
on the basis of total similarity, with group borders recognized by discontinui-
ties. In a fluid situation, such as obtains in the genus Bolitoglossa in South
America, undescribed species might easily fill one of these discontinuities,
necessitating changes in this arrangement in the future. Within the species
groups, attention is focused on direction of change in characters, and relative
degree of derivation of the various species. Attention is also given to the
degree of derivation of one group relative to others.

The genus Bolitoglossa is by far the largest in the Order Caudata, with over
60 species. It is convenient to recognize informal species groups, which in
turn form major assemblages. The species groups are not of equivalent rank,
but are comprised of from one to many species. Most have discrete geographic
patterns, and close relatives are not usually sympatric. Many of the species
groups appear to have resulted from the fragmentation and diversification of
what once were more or less continuously distributed populations. This pat-
tern is apparent in the helmrichi group of Nuclear Central America (Wake
and Brame, 1969) and in the adspersa group of northern South America
(Brame and Wake, 1963; Wake, Brame and Myers, 1970). Because of our
fragmentary knowledge of South American species we defer characterization
of these groups to a later date.

The following species groups and subgroups occur in South America and
adjacent Panama:

A. The adspersa group (subgroup 1. hypacra, adspersa, vallecula, sava-
gei, taylori, borburata, orestes; subgroup 2. palmata; subgroup 3.
nicefori, capitana, pandi ).

B. The sima group ( sima , chica, biseriata, silverstonei) .

C. The medemi group ( medemi , ramosi, walkeri, equatoriana) .

1971

New Species of Salamanders

31

D. The altamazonica group ( altamazonica , peruviana) .

E. The phalarosoma group (phalarosoma) .

The major departures from our arrangement of 1963 are: 1 ) the descrip-
tion of B. taylori and its addition to the ads per sa group; 2) the dissolution of
the palmata group and the assignment of B. orestes and B. palmata to different
subgroups of the adspersa group; 3) the division of the altamazonica group
and the uniting of the coastal species (B. sima, B. chicd) with B. biseriata,
formerly of the adspersa group, to form the sima group; 4) the description of
B. silver stonei as a member of the sima group; 5) the description of B. ramosi,
B. medemi, B. walkeri, and B. equatoriana, members of the medemi group.

The major division is between the relatively primitive adspersa group
and the other, more derived groups. The sima and altamazonica groups share
numerous derived features in proportions and foot structure, all perhaps
related to lowland, arboreal existence. They are separated by coloration and
osteological differences. The medemi group is more similar to these two
groups than to any other, although it also has some similarity in coloration,
proportions and foot structure to B. phalarosoma. The medemi group is the
only one of the four derived groups that contains some relatively primitive,
upland species. These species have slight similarities to members of the
adspersa group, but such species as B. biseriata also are similar to members
of the adspersa group in some features. The adspersa group contains the most
generalized South American species ( B . hypacra, B. vallecula, B. adspersa)
which resemble highland Middle American species (B. marmorea, B. cer-
roensis) in many features, mostly primitive states. The adspersa group con-
tains several highly derived species, both in the lowlands ( B . borburata, B.
capitana) and the highlands ( B . orestes, B. palmata).

The revised organizational scheme for South American species presented
here is based in large part on our expanded knowledge of many species as the
result of recent collection. Since our last survey of South American sala-
manders we have seen, in addition to specimens already reported, good series
of specimens that were living, preserved, or both, of the following species:
B. altamazonica, B. peruviana, B. sima, B. chica, B. vallecula, B. adspersa,
B. orestes, B. savagei, and B. biseriata. Additionally we have seen a few
recently collected specimens of B. phalarosoma and both living and preserved
specimens of B. capitana. Recently many specimens of B. hypacra and B.
nicefori, previously known from their holotypes, have been collected, and
living and preserved specimens have been studied. Species which remain
poorly known include B. palmata and B. pandi, the latter known only from
the holotype. Further comments in this paper will be focused on the newly
described species and their relatives.

Members of the sima group share similarity in size, webbing and other
features of their hands and feet, head proportions, and coloration. Bolitoglossa
silverstonei has more teeth than the other three members of the group, and
has a broader foot than either B. chica or B. biseriata. It is larger than B.

32

Contributions in Science

No. 219

chica. In South America the group is restricted to the wet forest west of the
Cordillera Occidental, but B. biseriata is widely distributed in Panama. All
species of the group are restricted to the lowlands, below 1000 m.

Members of the medemi group share similarities in proportions, denti-
tion and coloration. Bolitoglossa medemi and B. ramosi form one subgroup,
and B. walked and B. equatodana another. The former pair are similarly pro-
portioned and have generally similar color patterns. Both have extensively
webbed feet with reduced phalangeal numbers and flattened digits. The feet
of B. ramosi are smaller and much less well developed than those of B.
medemi. Bolitoglossa walked and B. equatoriana are somewhat more gen-
eralized than the other species pair. Bolitoglossa walked has the least web-
bing, the most discrete digits, and the most highly developed phalanges of any
species of the group. It has a somewhat narrower head than B. equatoriana,
and there are some color differences, but otherwise the species are similar.
While all members of the medemi group are allopatric, only B. walked lacks
sympatric associates. Bolitoglossa walked and B. ramosi occur at about 2000
m elevation, and the other species are lowland forms of the Choco and the
Amazonian basin.

The description of these five species brings the total number of species of
Bolitoglossa known from South America to 21, and another, B. taylori, occurs
nearly on the Colombian border in Panama. Of these, three ( equatoriana ,
peruviana, altamazonica) are extensively webbed, lowland Amazonian species,
six ( medemi , silver stonei, phalarosoma, biseriata, chica, sima) are extensively
webbed, lowland Chocoan species, and three ( hypacra , vallecula, adspersa)
are generalized, slightly webbed upland species from the Cordillera Occi-
dental, Cordillera Central, and Cordillera Oriental, respectively. The remain-
ing species range from diminutive, specialized highland species ( orestes ) to
giant species of intermediate elevation ( capitana ), and the degree of diversity
is relatively great. Many species inhabit cloud forest formations, and it is
these areas that are likely to produce additional populations. While some of
the generalized species are terrestrial, most species are occasionally to almost
exclusively arboreal. Species known to occur in bromeliads include B. nicefori,
B. savagei, B. ramosi, B. vallecula, B. borburata, and B. walked, and most, if
not all, of the lowland species are arboreal.

Recent field work has disclosed that sympatry, unknown in 1963, occurs
in the following combinations: B. medemi-B. phalarosoma, B. vallecula-B.
ramosi, B. peruviana-B . equatoriana, and B. sima-B. chica. We can expect
future field work to yield much additional information concerning ecology
and distribution, and, doubtless, new populations and undescribed species will
be found.

The five groups of South American Bolitoglossa present a rather broad
array of species. The adspersa group is diverse and its species are allopatric,
distributed broadly across Colombia to Panama, Venezuela, and Ecuador.
The sima and phalarosoma groups are specialized lowland forms of the wet

1971

New Species of Salamanders

33

northwestern forests. The medemi group is rather broadly distributed, eco-
logically and geographically, with species in the uplands in areas of Caribbean
and Pacific drainage, in the Choco, and in the Amazonian Basin. Finally, the
altamazonica group has the most peripheral distribution within the genus,
mostly within the Amazonian Basin.

Resumen

En el presente reporte se describen nuevas especies de salamandras ple-
todontidas para America del Sur y Panama. Bolitoglossa medemi es una
especie de color oscuro con manos y pies grandes y extensivamente palmeados
y con la cabeza ancha. Se le conoce en varias localidades en el noroeste de
Colombia y en el sudoeste de Panama, donde se le encuentra entre 50 y 800
m. de elevacion. Bolitoglossa ramosi es una especie de menor tamano, de
color mas claro, con manos y pies pequenos, pero tambien extensivamente
palmeados, y con la cabeza ancha. Esta especie es simpatrica con Bolito-
glossa vallecula en la Cordillera Central al este de Medellin, Colombia, a
altitudes de aproximadamente 1930 m. Bolitoglossa silver stonei es una especie
delgada, de larga cola y color claro, las manos y los pies son anchos y extensi-
vamente palmeados y la cabeza es moderadamente ancha. Ha sido encontrada
solo en una localidad cerca de Quibdo, a una altura de 400 m., en el noroeste
de Colombia. Bolitoglossa walkeri no tiene las extremidades tan palmeadas y
generalmente posee menos dientes maxilares que las otras especies. El color
es oscuro en el dorso y claro en el vientre. Se le encuentra a elevaciones de
cerca de 2000 m., cerca de Cali, Colombia. Bolitoglossa equatoriana tiene
extremidades mas palmeadas que B. walkeri, pero tiene como esta ultima, un
numero bajo de dientes y la misma coloracion. Es simpatrica con B. peruviana
en localidades de una elevacion de aproximadamente 260 m. en la Amazonia
ecuatoriana. La descripcion de estas nuevas especies permite una reevaluacion
de las relaciones sistematicas entre los miembros sudamericanos del genero
Bolitoglossa. Se discuten ademas los caracteres usados en el analisis siste-
matico de veintidos especies ye se reconocen cinco grupos de ellas. Los grupos
palmata y altamazonica son divididos. B. silver stonei es incluida en el nuevo
grupo sima, junto con B. sima, B. chica y B. biseriata. Las otras especies
nuevas son los unicos miembros del grupo medemi.

Literature Cited

Brame, A. H., Jr., and D. B. Wake. 1963. The salamanders of South America.

Los Angeles Co. Mus., Contrib. Sci. 69: 1-72.

Dunn, E. R. 1926. The salamanders of the family Plethodontidae. Northampton,
Mass. Smith College Publ. 441 p.

Espinal, L. S., and E. Montenegro. 1963. Formaciones Vegetales de Colombia.
Instituto Geografico, Bogota, Colombia, 201 p.

34

Contributions in Science

No. 219

Wake, D. B. 1966. Comparative osteology and evolution of the lungless sala-
manders, family Plethodontidae. So. Calif. Acad. Sci., Mem. 4: 1-111.

Wake, D. B., and A. H. Brame, Jr. 1966. Notes on South American salamanders
of the genus Bolitoglossa. Copeia 1966, 360-363.

1969. Systematics and evolution of Neotropical salamanders of the

Bolitoglossa helmrichi group. Los Angeles Co. Mus., Contrib. Sci. 175: 1-40.

Wake, D. B., A. H. Brame, Jr., and C. W. Myers. 1970. Bolitoglossa taylori, a
new salamander from cloud forest of the Serrania de Pirre, Eastern Panama.
Amer. Mus. Nov. 2430: 1-18.

Accepted for publication August 30, 197 1

Printed in Los Angeles, California by Continental Graphics

NUMBER 220
FEBRUARY 8, 1972

(! zL f<rr

A SYNOPSIS OF THE BURROWING
LAND CRABS OF THE WORLD and LIST
OF THEIR ARTHROPOD SYMBIONTS
AND BURROW ASSOCIATES

By Donald B. Bright and Charles L. Hogue

I

CONTRIBUTIONS IN SC1CNGE

0

NATURAL HISTORY MUSEUM • LOS ANGELES COUNTY

CONTRIBUTIONS IN SCIENCE is a series of miscellaneous technical papers
in the fields of Biology, Geology and Anthropology, published at irregular intervals
by the Natural History Museum of Los Angeles County. Issues are numbered sep-
arately, and numbers run consecutively regardless of subject matter. Number 1 was
issued January 23, 1957. The series is available to scientific institutions and scien-
tists on an exchange basis. Copies may also be purchased at a nominal price. Inquiries
should be directed to Virginia D. Miller, Natural History Museum of Los Angeles
County, 900 Exposition Boulevard, Los Angeles, California 90007.

INSTRUCTIONS FOR AUTHORS

Manuscripts for CONTRIBUTIONS IN SCIENCE may be in any field of
Life or Earth Sciences. Acceptance of papers will be determined by the amount
and character of new information. Although priority will be given to manuscripts
by staff members, or to papers dealing largely with specimens in the collections of
the Museum, other technical papers will be considered. All manuscripts must be
recommended for consideration by the curator in charge of the proper section or
by the editorial board. Manuscripts must conform to those specifications listed
below and will be examined for suitability by an Editorial Committee including
review by competent specialists outside the Museum.

Authors proposing new taxa in a CONTRIBUTIONS IN SCIENCE must
indicate that the primary type has become the property of a scientific institution of
their choice and cited by name.

MANUSCRIPT FORM.— (1 ) The 1964 AIBS Style Manual for Biological
Journals is to be followed in preparation of copy. (2) Double space entire manu-
script. (3) Footnotes should be avoided if possible. Acknowledgments as footnotes
will not be accepted. (4) Place all tables on separate pages. (5) Figure legends and
unavoidable footnotes must be typed on separate sheets. Several of one kind may
be placed on a sheet. (6) An abstract must be included for all papers. This will be
published at the head of each paper. (7) A Spanish summary is required for all
manuscripts dealing with Latin American subjects. Summaries in other languages
are not required but are strongly recommended. Summaries will be published at
the end of the paper. (8) A diagnosis must accompany any newly proposed taxon.
(9) Submit two copies of manuscript.

ILLUSTRATIONS. — All illustrations, including maps and photographs, will
be referred to as figures. All illustrations should be of sufficient clarity and in the
proper proportions for reduction to CONTRIBUTIONS page size. Consult the 1964
AIBS Style Manual for Biological Journals in preparing illustration and legend copy
for style. Submit only illustrations made with permanent ink and glossy photo-
graphic prints of good contrast. Original illustrations and art work will be returned
after the manuscript has been published.

PROOF.— Authors will be sent galley proof which should be corrected and
returned promptly. Changes in the manuscript after galley proof will be billed to
the author. Unless otherwise requested, page proof also will be sent to the author.
One hundred copies of each paper will be given free to each author or divided
equally among multiple authors. Orders for additional copies must be sent to the
Editor at the time corrected galley proof is returned. Appropriate order forms will
be included with the galley proof.

Virginia D. Miller
Editor

A SYNOPSIS OF THE BURROWING LAND CRABS OF THE WORLD
AND LIST OF THEIR ARTHROPOD SYMBIONTS
AND BURROW ASSOCIATES

By Donald B. Bright1 and Charles L. Hogue2

Abstract: The burrowing land crabs of the world are
defined as an ecological group and the burrow or crabhole
faunal community is recognized and discussed as such.

Introductory remarks on terminology, relationship of the
crabhole habitat to other habitat types, general physical nature
of the crabhole, and the major ecological structure of the com-
munity as now known are presented. The remainder of the paper
consists of two parts: 1) A list of all the species of borrowing
land crabs of the world, including notations on distribution,
recognition, and ecology. Twenty-four species in the genera
Sesarma, Ocypode, Uca, Ucides, Gecarcoidea, Car disoma, and
Gecarcinus are given. 2) A list of all published records of arthro-
pods found in crab burrows either associated with the crab as a
burrow coinhabitant or having symbiotic relationships with it.

The vast majority of these are insects, primarily mosquitoes, of
which 140 species are noted. For each burrow associate or sym-
biont, the distribution, recorded crab host, type of relationship
(specific, semispecific, transient or accidental) are given.

INTRODUCTION

The present paper represents a literature survey to establish the present
state of knowledge on the unique ecological relationship existing between
burrowing land crabs and a variety of associated organisms. From our own
field studies it is evident that there are many unrecorded species of arthropods
occurring in crabholes and undescribed ecological phenomena to be dis-
covered and analyzed. We hope that from this beginning other workers will
recognize the land crab burrow as a special habitat and respond to the
need for inquiring further into its natural history.

Published data on land crabs consists primarily of species accounts and
selected aspects of behavior and natural history; no broad coverage of the
basic ecology of any species exists. Likewise, with regard to the burrow
associates, no general ecological treatment is available, only taxonomic notes
and fragmentary collecting data.

We are presently engaged in a project to study the biology of land
crabs and their burrow associates (Hogue and Bright, 1969). One preliminary
field survey in Kenya, East Africa (Hogue and Bright, 1971) has been
reported. Field studies in Costa Rica, Baja California, Pacific mainland

1Donald B. Bright, Associate Professor of Biology, California State Colege, Ful-
lerton, California 92631.

2Charles L. Hogue, Senior Curator of Entomology, Natural History Museum
of Los Angeles County, Los Angeles, California 90007.

1

2

Contributions in Science

No. 220

Mexico, Panama, Peru, Ecuador, Western Caribbean (Islas San Andres and
Providencia) , and Australia will be reported upon in forthcoming papers.
All specimens and data collected on this project are given the code LCBA
(Land Crab Burrow Associates). Entomological materials are deposited in
the Los Angeles County Museum of Natural History, and crustaceans in
the Department of Biology, California State College, Fullerton.

DEFINITIONS

Since we make use of terms and concepts originally devised for very
different community types, we find it necessary to define certain of our present
usages :

Crabhole and crab burrow: Used synonymously.

Burrowing land crab: This term refers to a group of tropical species which
dig well-defined burrows above the normal flooding and flushing action of
the tides. These species belong to the families Gecarcinidae, Ocypodidae,
and Grapsidae. The family Coenobitidae is excluded since the terrestrial
hermit crabs are non-burrowers. Thus our attention will be directed to the
following taxa:

Gecarcinidae (all 3 genera of the family)

Gecarcoidea, all species
Cardisoma, all species
Gecarcinus, all species

Grapsidae (only 1 of 26 genera in the family)

Sesarma, certain species only

Ocypodidae (3 of 4 genera in the family)

Ocypode , certain species only
Uca, certain species only
Ucides, certain species only

Several species of land dwelling lobsters (e.g., Thalassina) , crayfish
( Procambarus , Cambarus) and freshwater crabs ( Sudanonautes ) also con-
struct burrows supporting an associated fauna. Though we consider these
outside our specified limits with burrowing land crabs (and they are omitted
from the section on burrowing land crabs), we mention them in our survey
of associates because they occur sympatrically with burrowing land crabs
and confusion frequently arises in identifying the true owner of a burrow
(Scharlf and Tweedie, 1942).

Ecological structure of crabhole community: All of the organisms found in
the crabhole and associated ecologically with land crabs we refer to as the
crabhole community. It is presently possible to evaluate only the gross com-
munity structure of arthropod symbionts and burrow associates of land
crabs in terms of the general levels of interrelationships and the most con-
spicuous variations in niches displayed by the arthropod fraction. By niche
we mean the total ecological role a species plays in the community, both
in regard to its habitual location of occurrence (place niche— microhabitat)

1972

Burrowing Land Crabs of the World

3

and its inter-dependency with other members of the community (functional
niche).

Within the crabhole community we find two general levels of inter-
specific reaction: The first of these, simple association, is shown by the
assemblage of species whose common occurrence is dictated directly by the
spectrum of indigenous limiting factors (physical and biotic) encountered
in the crabhole. The second, symbiosis, refers to those species not only tied
together by these factors but which also depend directly upon some form
of intimate (often or usually also involving prolonged physical contact)
interaction with one other member of the community, i.e., parasitism or
commensalism (no example of mutualism yet having been found). A sum-
mary of the general ecological structure of the arthropod fraction of the
crabhole community is given in Table 1.

The extent to which a particular organism is dependent on the crab-
hole as a suitable habitat and to which it is an obligate member of the
crabhole community may be further classed:

I. Specific (or obligatory)

The species is narrowly adapted to conditions in the crabhole or lives
symbiotically with the host or other community member. Such species have
specific adaptations to the physical and biological stresses encountered in
the burrow (the precise nature of and adaptive significance of which are

Table I

SUMMARY OF THE KNOWN ECOLOGICAL STRUCTURE OF THE
CRABHOLE COMMUNITY (ARTHROPOD FRACTION)

Level of
Interspecies
Reaction

Niche

Place Functional

Major Examples

ASSOCIATION

Burrow water

Developing;
feeding, breed-
ing, etc.

Immature mosquitoes

Diving beetles
( Bidessus )

Cyclops

Burrow chambers
and surface of
burrow water

Resting, mating,
etc.

Adult mosquitoes

Adult biting gnats
(Culicoides)

SYMBIOSIS

Parasitism

Gill chamber of
Gecarcinus
lateralis

Attaching and
feeding

Mite ( Laelaps cancer )

Commensalism

Peribuccal cavity
and renal grooves
of Gecarcinus
ruricola

Attaching to host
and feeding on
food debris

Drosophila

carcinophila

4

Contributions in Science

No. 220

presently little known, such as prolonged developmental period, impermeable
cuticle, reduced salt absorbing organs, etc.). Examples occur among the
following genera: Deinocerites, Aedes ( Cancraedes and Geoskusea) , and
Drosophila.

II. Semispecific (or semiobligatory)

The species usually inhabits the crabhole, being adapted at least in
part to certain of its conditions but survives well in other habitats. An
example is the mosquito Aedes ( Skusea ) pembaensis which habitually breeds
in crabholes but also develops commonly in various other types of coastal
ground water accumulations (pools, swamps and even in artificial containers).

III. Transient (or facultative)

The species usually inhabits other sites but may take up temporary
residence or breed in the crabhole because of its similarity or proximity to
the normal habitat. Examples are the many mosquitoes occurring along
the seashore such as Aedes ( Neomacleaya ) panayensis and Aedes ( Ochlero -
tatus ) taeniorhynchus', insular treehole breeders, Aedes ( Stegomyia ) poly-
nesiensis; and indiscriminate breeders, Culex (Culex) annulirostris.

IV. Accidental

The species is adapted to another habitat and only rarely occurs in and
about crabholes for some anomalous reason (e.g., larvae flushed from ground
pools during heavy rains, wind blown adults, etc.). For example, adults of
Mansonia mosquitoes are sometimes found in crabholes but do not develop
from larvae and pupae living in the burrow; these mosquitoes require certain
aquatic plants to which they attach with special respiratory structures for
extracting vascular oxygen. These plants live only in open freshwater pools
and ponds.

These four categories of habitat dependence, of course, are provisional.
Unfortunately, for no species do we yet know the complete story of its
functional niche in the crabhole community.

GENERAL PHYSICAL NATURE OF THE HABITAT

While there is considerable specific variation, in general the land crab
burrow is a gently sloping or near vertical tubular excavation ranging from
depths of 1.5-3. 4 m. The diameter generally is equivalent to the carapace
width of the host crab and the depth is determined by the level of the water
table.

Normally, the bottom of the burrow is filled with water derived indirectly
from ground seepage from nearby sources (streams, estuaries, ponds, open
sea, etc.) or directly from rainfall. Thus the water may vary considerably
in solute concentrations even from day to day or hour to hour. Like estuarine
organisms in general, which are able to accommodate to such changes
physiologically, crabhole water dwellers have wide osmoregulatory capacities.
Some species, in genera such as Gecarcinus, are so well adapted to terres-

1972

Burrowing Land Crabs of the World

5

triality that their burrows often are for physical protection only and penetra-
tion to the water table to maintain a supply of water for physiological func-
tions is not necessary. The community of these shallow burrows, lacking the
aquatic fraction, is depauperate.

The burrows are located most often in compact alluvial soil well above
high tide lines but still close enough to the sea to permit migration for spawn-
ing and close enough to a ground water source to maintain a reservoir. Bur-
rows also are often found along large rivers far inland where the fresh
water affords the crab’s hydrobiotic needs.

LAND CRABS OF THE WORLD

The following genera and species accounts are for those burrowing
land crabs listed as hosts in the arthropod portion of this paper or those,
based on our field experience, that are likely to be additional hosts.

In several of the species accounts there is question regarding reliable
taxonomic and zoogeographic data. This is particularly true for distributional
patterns in the Indo-Pacific, e.g., the species of Car disoma and Sesarma.
(See Tweedie (1950) for a discussion of the problems associated with Cardi-
soma.) Correspondence with a number of workers indicates that revisions
are in preparation for Sesarma, Uca and Gecarcinus. These works should aid
future studies on the distribution of land crabs and their burrow associates.

No attempt has been made to provide a complete synonymy. Where
there is considerable taxonomic confusion a note to clarify our usage has
been included in the species accounts.

In listing the various species we have assumed that the published reports
on burrow ownership are correct. Where authors were unable to identify the
crustacean but provide descriptions, specific localities and/or habitats, we
have sometimes made a provisional determination. It is hoped that in future
accounts authors will attempt to determine hosts specifically and include addi-
tional remarks on general ecology of the crab and its associates.

The species we define as land crabs are enumerated below followed by
a synopsis of important information on each species.

Family: Grapsidae

1. Sesarma {Sesarma) sulcatum

2. Sesarma {Sesarma) meinerti

3. Sesarma {Chiromantes) africanum

4. Sesarma {Holometopus) ortmanni

5. Sesarma {Holometopus) eulimene

Family: Ocypodidae

6. Ocypode gaudichaudii

7. Ocypode occidentalis

8. Ocypode quadrata

9. Ocypode ceratophthalma

6

Contributions in Science

No. 220

10. Uca pugilator

11. Uca subcylindrica

12. Uca t anger i

13. Ucides cordatus

14. Ucides occidentalis

Family: Gecarcinidae

15. Gecarcoidea humei

16. Cardisoma guanhumi

17. Cardisoma crassum

18. Cardisoma armatum

19. Cardisoma carnifex

20. Cardisoma hirtipes

21. Gecarcinus planatus

22. Gecarcinus ruricola

23. Gecarcinus quadratus

24. Gecarcinus lateralis

Family GRAPSIDAE
Genus Sesarma Say, 1817

Characters: Carapace squarish; sides generally straight and parallel;
orbits of eye deep, oval and occupy only slightly less than half of the anterior
border of the carapace; antennules transverse; epistome well defined; chelipeds
thick, and subequal in male, and third pair of legs longest.

Distribution: Tropical and subtropical coastal areas of the world.

Habitat: Coastal marshes, mud flats, banks of drying streams, gravelly
mud along lagoons and mangroves.

Habits: Generally these have well defined burrows which are similar to
those of most Cardisoma juveniles. These burrows are in muddy, mud-gravel
areas extending from the surface down 1 m to the water table. Some do not
construct burrows but live under debris (rocks and roots). Most individuals
are solitary. In some areas young live in the same burrow with an adult.

References1: Bott, 1955 (D,T); Campbell, 1967 (D,T); Crane, 1947
(B,D,T,) ; Crosnier, 1965 (D,T) ; Gordon, 1934 (D); Macnae, 1966 (B,D) ;
Miers, 1880 (D); Rathbun, 1914 (T); Tesch, 1917 (D,T); Tweedie, 1940
(D,T).

Note: Considerable taxonomic confusion prevails in this taxon, particu-
larly the validity of the subgeneric groupings. Many synonymies are suspected

!The application of each reference is indicated by a symbol following (T— taxonomy;
B— biology, i.e., habits, habitat, life history, physiology; D—distribution; G— general).

1972

Burrowing Land Crabs of the World

7

because of the inordinate number of species, e.g., 115 in the Indo-Pacific
alone. Campbell (1967) and Crosnier (1965) give recent accounts dealing
with these problems. Because of the above, some species determinations are
questionable.

1. Sesarma ( Sesarma ) sulcatum Smith, 1870

Color: Carapace and legs dark brownish gray; lower portion of male
chelae cream yellow; chelae in females cream with a few maroon striations;
females with conspicuous yellow line across front.

Distribution: Pacific coast of the Americas (San Ignacio Lagoon, Baja
California to southern Panama).

Habitat: Gravelly mud along lagoon shores, tidal marshes, and on the
banks of streams and mangroves.

Habits: They construct straight or slightly sloped burrows or live under
debris characteristic of the habitat. Some spend considerable time climbing
the branches of marsh plants, e.g., Sueda, Salicornia (Baja California) and the
roots and pneumatophores of mangroves, e.g., Rhizophora and Avicennia
(Costa Rica and Panama). Individuals are solitary, but in areas where burrow
structure is not well developed they tend to occur in groups, e.g., three to six
individuals under a rock (Wright, 1966). In drier habitats they commonly use
burrows (occupied and unoccupied) of other crabs, e.g., Cardisoma crassum,
Ucides occidentalis and Uca spp. They are often sympatric also with Goniopsis
pulchra. They are active throughout most of the day (except in drier areas)
feeding primarily on plant materials.

Common Names: Mangrove crab; Marsh crab; Speckled crab.

References: Bott, 1955 (D,T); Crane, 1947 (B,D,T,); Garth, 1960
(D); Wright, 1966 (G).

2. Sesarma ( Sesarma ) meinerti de Man, 1887

Color: Carapace black to gray or purple to deep violet, anterior and
lateral margin bordered with orange to light yellow; underside a dirty yellow;
and chelipeds a striking brilliant red. Cott (1930) gives a good account of this
plus a consideration of the theory of warning colors.

Distribution: Andamans and Madras, Mozambique, Mauritius, Mada-
gascar; east coast of Africa (south to Port St. John’s); across Indo-Pacific to
Australia (Cooktown) and north to the Philippines.

Habitat: Sandy-clay areas and higher, drier, muddy banks associated
with estuaries and mangroves.

Habits: Burrows are well developed and most common in areas where
there is dry, relatively hard mud. The burrows are deep and usually extend to
the water table. Often the mouth of the burrow has a hood built of mud exca-
vated while enlarging the tunnel or cleaning out. These crabs are retiring,
remaining at the mouth of the burrow, and only leave to forage at night. They
apparently feed primarily on plant material, but also act as scavengers where

8

Contributions in Science

No. 220

they occur in high density. There is no indication of colonialism in areas of
high density.

Common Name: Marsh crab?

References: Alcock, 1900 (D,T); Chace, 1953 (D,T); Cott, 1930
(G); Crosnier, 1965 (D,T) ; Hogue & Bright, 1971 (D); Macnae, 1966
(B,D); Millard and Harrison, 1954 (G).

3. Sesarma ( Chiromantes ) africanum H. Milne Edwards, 1837

Color: Carapace reddish brown; transverse patches of stiff hairs over
carapace and limbs; distal portion of chelae violet-red in color.

Distribution: Senegal to Benguela, Angola; also Barbados (?).

Habitat: Mangroves; salt marshes; and mouths of rivers.

Habits: Occurs primarily in dense, well shaded areas of mangroves.
Juveniles and adults are conspicuous, climbing over vegetation, since they
have no burrows or only very small ones. Typically, when threatened, they
hide under debris and roots. In open localities, with soft mud, they construct
shallow (.3 -.6 m) individual burrows. These burrows are not known to inter-
sect. They are presumed to be scavengers.

Common Name: Hairy lagoon crab.

Reference: Rathbun, 1921 (D,T).

4. Sesarma ( Holometopus ) ortmanni Crosnier, 1965

Color: Carapace greenish brown and heavily calcified; chelae a dull to
bright orange.

Distribution: East coast of Africa; Madagascar.

Habitat: Muddy soil along the margins of mangroves.

Habits: Constructs shallow burrows among exposed pneumatophores of
the mangrove, Avicennia.

Common Name: None recorded.

References: Crosnier, 1965 (D,T); Macnae and Kalk, 1969 (B,D).

5. Sesarma ( Holometopus ) eulimene de Man, 1898

Color: Carapace dull brown and with conspicuous pits; underside a
dirty white, chelae of male bright orange-red.

Distribution: East coast of Africa from Malindi to Durban.

Habitat: Mud areas of salt marshes and mangroves.

Habits: Poorly known. Burrows generally well developed, but shallow,
and often associated with the pneumatophores of mangroves. Millard and
Harrison (1954) indicate extensive distribution of this species in Richards
Bay, South Africa, in areas along the mangrove margins where there is deep,
soft mud covered at least once per day by tidal flux.

Common Name: None recorded.

References: Barnard, 1947 (D,T), 1950 (D,T); Crosnier, 1965
(D,T) ; Macnae and Kalk, 1969 (B,D) ; Millard and Harrison, 1954 (G).

1972

Burrowing Land Crabs of the World

9

Family OCYPODIDAE
Genus Ocypode Fabricius, 1798

Characters: Carapace deep, somewhat broader than long but generally
squarish; orbits large and divided into two chambers; eye stalk often prolonged
as a style; antennae small and rudimentary; epistome small; chelipeds shorter
than legs and subequal; 4th pair of legs shorter and thinner than others.

Distribution: Tropical and subtropical coast of American Atlantic
(Rhode Island to Brazil), Mediterranean Sea, coasts of Africa, Red Sea,
Indo-Pacific, and eastern Pacific (Turtle Bay, Baja California to Chile).

Habitat: Sandy beaches with tidal surge; rubble flats; sand-mud areas
adjacent to mangrove swamps.

Habits: Construct simple to complex burrows in the soft substratum of
the habitat. In several areas species occur sympatrically but generally are dis-
tinguishable on the basis of feeding habits or the presence-absence-degree of
development of the style over the eye.

References: Alcock, 1900 (D,T); Chace and Hobbs, 1969 (G); Crane,
1941b (B); Garth, 1960 (D); Rathbun, 1918 (D,T) ; Tweedie, 1950 (D,T).

6. Ocypode gaudichaudii H. Milne Edwards & Lucas, 1843

Color: Highly variable, coral red to dark brown. Individual color associ-
ated with size, sex and the color of substratum of the habitat.

Distribution: Pacific coasts of America. Gulf of Fonseca, El Salvador,
to Chile; Galapagos Islands.

Habitat: Common on sandy beaches of protected bays; occasionally on
exposed sandy areas when the surge is not high; also occurs along the shores
of lagoons.

Habits: Burrows highly variable; Crane (1941) indicates three types:

1) shallow, simple, oblique; 2) straight with a right angle at 1-3 m depth; and
3) straight for 15-20 cm then extending downward in a gradual spiral. The
second type is most common in the center of the range of distribution. All are
diurnal and most active following high tide and before flooding midtide. Occa-
sionally occurs sympatrically with O. occidentalis. These two species are
distinguishable by: 1) O. gaudichaudii : carapace length of 17 mm or more
and with well developed ocular styles; actively manipulate the substratum
feeding on microscopic organic matter (similar to several species of Uca );

2) O. occidentalis : no ocular style; they are confirmed predators and/or
scavengers.

Common Names: Cart-driver; Carretero (Peru).

References: Bott, 1955 (D,T) ; Crane, 1941b (B); Garth, 1948 (D,T) ;
Rathbun, 1918 (D,T).

7. Ocypode occidentalis Stimpson, 1862

Color: Upper surface of body and legs generally darkish gray with white
marbling. Manus of chelipeds, tips of walking legs and underside of body

10

Contributions in Science

No. 220

cream white. Colors tend to vary with substratum except where sand is vol-
canic (dark black). These crabs are, due to color, conspicuous when found
on light colored or dry sand.

Distribution: Turtle Bay, Baja California to Ancon, Peru.

Habitat: Sandy surge-beaten beaches; sandy-silt areas adjacent to rivers
where water flow is fairly rapid, e.g., Playas del Coco, Costa Rica.

Habits: Distinguishable from O. gaudichaudii by absence of ocular styles.
Burrows similar to common mode for O. gaudichaudii. Completely nocturnal
in habits except when very young (see remarks in discussion of O. gaudi-
chaudii). They are scavengers and their rapid movements when disturbed
contribute to their common name, ghost crab. Feeding begins shortly after
onset of ebb until about mid-flood tide. Activity greatest during ebbing of
tide and at slack water.

Common Name: Ghost crab.

References: Bott, 1955 (D,T); Crane, 1941b (B); Garth, 1960 (D);
Rathbun, 1918 (D,T).

8. Ocypode quadrata (Fabricius, 1787)

Color: Upper surface white with small black spots or generally a pale
yellow or a grayish white or a speckled brown. Many show a degree of irides-
cence along the outer areas of the carapace. The general color pattern is appar-
ently associated with the color of the substratum of the habitat, e.g., dark
brown at Tortuguero, Costa Rica, while pale yellow at Punta Cahuita, Costa
Rica. Chace and Hobbs (1969) give a detailed account of the two color phases
found on Dominica.

Distribution: Atlantic coasts of America. Rhode Island to Estado do
Santa Catharina, Brazil, including most of the islands in the greater and lesser
Antilles.

Habitat: Sandy beach areas from upper tidal level to well beyond supra-
littoral area. Distance from sea generally associated with distribution of same
and associated vegetation.

Habits: Burrows deep, .45-.75 m. Chace and Hobbs (1969) note two
types of burrows: 1 ) vertical or nearly so; and 2) U-shaped. Throughout most
of Central America, the U-shaped are most common with the bottom of the
U about .45 m. Burrow construction is initiated by a general scratching using
the chelipeds followed by removal of sand for the entrance using the dactyli
of the walking legs. Once the burrow is sufficiently deep, sand is then trans-
ported in either the right or left cheliped to the surface and then dumped, or
it may be tossed out of the burrow using either cheliped. These crabs are
scavengers and are probably a geminate species of O. occidentalis in the Pacific.

Common Name: Ghost crab. (In the past also called a sand crab, but
this name is more commonly applied to members of the unrelated genus
Emerita. )

References: Bott, 1955 (D,T); Chace and Hobbs, 1969 (G); Chace

1972

Burrowing Land Crabs of the World

11

and Holthuis, 1948 (T); Pearse, 1916 (G); Rathbun, 1918 (D,T), 1933
(D,T).

Note: This has until recently been cited by many authors as O. albicans
Latreille, 1802 (see Chace and Hobbs, 1969).

9. Ocypode ceratophthalma (Pallas, 1772)

Color: Generally the dorsal appearance is from sage green and yellow
to grayish white. Tweedie ( 1950) gives the color for those found on the Cocos-
Keeling Islands as uniform gray, but sometimes olive, and with a splash of
yellow on the chelae. The color is highly variable, and associated with the
character of the substratum (see Green ( 1964) for discussion of color changes
in Hawaiian forms) .

Distribution: East coast of Africa (Port St. John), Red Sea, Xndo-
Pacific North to Kii Peninsula, Japan, including many islands (Mauritius,
Maidive and Laccadive Islands, Caroline and Marshall Islands, Tuamotu
Islands, Guam and Hawaiian Islands).

Habitat: Sandy beaches.

Habits: Burrows are common from the high tide mark to 6-10 m above;
less frequently they occur in areas beyond the edge of the sea. They have also
been reported along the margins of lagoons. Burrows are generally 37-50 cm
deep. In those sites affected by tidal flux the crabs emerge from their burrows
during the ebbing of the tide to feed. They are scavengers feeding on debris
deposited by the tidal exchange; in the areas above the tidal influence they feed
on small organisms, e.g., crickets (Gressitt, 1954). Well defined ocular styles
are present in adults of this species.

Common Names: Ghost crab; Kepiting Mata Panjang (Longeyed crab)
(Christmas Island, Indian Ocean).

References: Alcock, 1900 (D,T); Barnard, 1950 (D,T); Borradaile,
1902 (B); Dakin et al., 1952 (G); Day et al., 1954 (B,D); George and Knott,
1963 (D,T) ; Gordon, 1934 (D); Green, 1964 (B); Holthuis, 1953 (D);
Miers, 1880 (D); Millard and Harrison, 1954 (G); Sakai, 1940 (D); Tesch,
1918 (D,T) ; Tweedie, 1950 (D,T) ; Ward, 1934 (D).

Genus Uca Leach, 1814

Characters: Carapace deep, somewhat broader than long and with
antero-lateral area pronounced and/or projected, orbits deep and oblique;
antennae large and pronounced; epistome short and distinct; chelipeds of males
extraordinarily unequal and large while in female equal and small; ambulatory
legs longer than small cheliped of male and both chelipeds of female, last
pair of ambulatory legs shorter than rest.

Distribution: American Atlantic (Boston to Uruguay), Mediterranean,
west coast of Africa (Portugal to Angola), east coast of Africa to Indo-Pacific,
including Maury Islands, and eastern Pacific (British Columbia to Valparaiso,
Chile).

12

Contributions in Science

No. 220

Habitat: Associated with a wide variety of mud-sand habitats, e.g., clay
tidal flats, salt marshes, mangroves, low tide muddy areas, etc.

Habits: Most construct well-defined burrows where substratum is moist
enough to maintain burrow configuration. Area may be exposed or associated
with dense vegetation (e.g., mangroves, Avicennia , or pickle weed, Sali-
cornia ) . Many species show a correlation of burrow site with tidal flux whereas
others occur widely throughout tidal zone even when burrows may become
quite dry. Most species show considerable social behavior. Crane (1941a,
1943b, and 1957) gives a thorough account of display, breeding and rela-
tionships of a number of sympatric species of the genus. See Salmon (1965)
for an additional account of courtship behavior. There is considerable tax-
onomic discord associated with this genus due to voids in collections from
certain areas and the high degree of variability in color.

References: Barnard, 1950 (D,T); Bott, 1954 (D,T); Crane, 1941a
(G), b (B), 1943a (D), b (B), 1957 (B); Hagen, 1968 (D); Salmon, 1965
(B); Schmitt, 1921 (D).

10. Uca pugilator (Bose, 1801)

Color: Carapace cream white; large cheliped of male buff with apricot
at base of movable finger; chelipeds of small males and females white with a
grayish cast.

Distribution: Boston Harbor, Massachusetts, to Brownsville, Texas.

Habitat: Sandy areas where sand content is generally more than 40 per

cent.

Habits: They construct well defined, moderately deep burrows showing
considerable variation with age and location. See Pearse (1914a) and Dem-
bowski (1925) for details of burrow construction. Feeding, burrow repair-
construction and social behavior occur during the low tide period, generally
ceasing when the tidal level reaches the burrow entrances; however, some of
these events are correlated with sunset and sunrise as well (see Salmon, 1965).
This species is restricted by substratum preference. In a typical situation the
sand content may increase down a bank from 10 per cent at the top to 60-70
per cent at the low tide level, with the crabs feeding at the lower levels and
living near the top. See Teal (1958) for an account of feeding habits as related
to sand content.

Common Name: Sand fiddler crab.

References: Burkenroad, 1947 (B); Crane, 1943b (B); Dembowski,
1925 (B); Pearse, 1914a (B), b (B); Salmon, 1965 (B); Salmon and Stout,
1962 (B); Teal, 1958 (B).

11. Uca subcylindrica (Stimpson, 1859)

Color: No record.

Distribution: Corpus Christi, Texas to northern Mexico.

Habitat: Mud-sandy areas in and adjacent to estuarine situations.

1972

Burrowing Land Crabs of the World

13

Habits: Burrows are small, slightly twisted, and constructed in loose
sandy soil with high moisture content or in muddy areas. Burrows seldom
deeper than .75 m. Little other ecological data recorded for this species.

Common Name: Puffed fiddler crab.

Reference: Rathbun, 1918 (D,T).

12. Uca tangeri (Eydoux, 1835)

Color: Carapace reddish brown to dirty yellow; male large cheliped
reddish brown to pale blue; female chelae dirty cream with some pinkish areas.

Distribution: Portugal, north and west coasts of Africa to Angola.

Habitat: Sandy areas adjacent to brackish water; salt marshes; not
common on open coasts.

Habits: Burrows which extend downward for about 30 cm or so are
located from the mean tide to highest high tide level. The top of the burrow is
generally plugged with 7.5-10 cm of excavated substratum by the time the
burrow is covered by the tide. As the tide recedes the plug is removed and
the crab emerges to feed, etc. Habits vary from solitary to colonial. In
colonial situations, e.g., salt marsh flats, the burrows are still simple, but they
often intersect. See Hagen (1961) and Hediger (1934) for additional details.

Common Names: Fiddler crab; Calling crab.

References: Altevogt, 1959 (B), 1962 (B); Barnard, 1950 (D,T);
Hagen, 1961 (D), 1962 (B); Hediger, 1933 (B), 1934 (B); Rathbun, 1921
(D,T).

Note: This was described by Eydoux in 1835 as Gelasimus tangeri, and is
recorded as such by some authors.

Genus U aides Rathbun, 1897

Characters: Interorbital distance a little more than one-half the greatest
width of the carapace; orbits deep but not much larger than the eyes; anten-
nules oblique; epistome small but prominent; legs stout.

Distribution: East and west coasts of the Americas.

Habitat: Common inhabitants of muddy shores and mangrove swamps
where there is a moderate degree of tidal flux.

Habits: Burrows constructed below the highest high tide level most
commonly at midtide level so that burrows covered daily by tidal surge. Bur-
rows not uniform, generally shallow, relatively straight, and frequently with
multiple entrances; almost always filled with ground water to the level of the
burrow mouth. Juveniles occur in dense numbers per unit area within an
interchange of intersecting burrows or in burrows of less density per unit area
but where the burrow is directly adjacent and/or attached to those of adults.

References: Bright, 1966 (G); Rathbun, 1918 (D,T).

Note: Until recently this genus was placed among the members of the
family Gecarcinidae but Chace and Hobbs (1969) placed it among the mem-
bers of the family Ocypodidae.

14

Contributions in Science

No. 220

13. Ucides cordatus (Linnaeus, 1763)

Color: Carapace pale yellow with cervical groove and urogastric lobe
rusty brown; walking legs red-violet; tips of chelae cream. Young with a tend-
ency to have a dark gray area along the median anterior margin of the cara-
pace. Color is variable throughout the range of distribution primarily asso-
ciated with the nature of the habitat substratum.

Distribution: Atlantic coasts of America. Southern Florida (Biscayne
Bay) to Santos, Brazil, including the West Indies.

Habitat: Areas, frequently flooded by tidal surge, in mangroves along
mouths of rivers and brackish water marshes adjacent to the sea.

Habits: This crab constructs burrows in very soft mud in areas where
there is an absence of low ground vegetation (shrubs). Primary sites are
mangroves with burrows concentrated along the upper edge of distribution
of the red mangrove ( Rhizophora mangle). Burrows are wide, mostly straight
and relatively shallow. Young and adults construct their burrows in close
proximity. At both ends of the range there is some indication that burrows
are shallower and often nothing more than depressions. This is a twin species
of U. occidentalis in the Pacific.

Common Names: Pagurus; Kaburi (Cuba); Uga (Brazil).

References: Bott, 1955 (D,T); Bright, 1966 (G); Chace and Hobbs,
1969 (G); Garth, 1960 (D); Manning and Provenzano, 1961 (D); de
Oliveira, 1946 (B); Rathbun, 1918 (D,T), 1933 (D,T).

14. Ucides occidentalis (Ortmann, 1897)

Color: Carapace reddish gray with orange-red on the lateral margin;
however, older forms tend to become rust red due to staining from the mud
in the burrow. The last three ambulatory legs and most of the chelipeds
dark red; dactyli of chelipeds reddish white; underside brownish white. Molt
condition not predictable from color change.

Distribution: Pacific coast of America (Espiritu Santo Island, Baja
California to Rio Tumbes, Peru).

Habitat: Mud of mangrove areas, mouths of rivers and brackish water
marshes.

Habits: This species maintains burrows in areas which are generally
covered by high tide at least once per month. Burrows are most common
along the mud-water margin of mangroves; may be associated with salt marsh
vegetation, e.g., Salicornia, Sueda, etc. Burrows are shallow and often with
more than one entrance. Typically, there is a small side chamber or tunnel
paralleling the surface just inside the mouth of the burrow. Juveniles are
most often found in small pockets connecting to the burrows occupied by
the adults. Burrows do not generally extend more than 50 cm below the
surface. This is a geminate species of U. cordatus which occurs in the Atlantic.

Common Names: Wide red land crab, Cangrejo amarillo (Peru).

1972

Burrowing Land Crabs of the World

15

References: Bright, 1966 (G); Garth, 1960 (D); Rathbun, 1918
(D,T).

Family GECARCINIDAE
Genus Gecarcoidea Milne Edwards, 1837

Characters: Fronto-orbital border less than half the greatest breadth
of carapace; orbits deep; antennae very small and excluded from the orbit;
epistome sunken and quite hairy; chelipeds equal or nearly so in both sexes;
legs stout.

Distribution: Indo-Pacific Islands.

Habitat: Moist soil or muddy areas in the jungle areas adjacent to
the sea.

Habits: Burrows shallow and not well developed.

References: Caiman, 1911 (B); Gibson-Hill, 1947 (B); Keilin, 1921
(D); Rathbun, 1918 (D,T) ; Tweedie, 1947 (D,T); Webb, 1922 (B).

15. Gecarcoidea humei (Wood-Mason, 1873)

Color: Dorsal surface a relatively uniform red-violet with some indi-
cation of red near the base of the chelae; claws white-brown with reddish
violet tinge; scars on carapace yellow or yellowish white.

Distribution: Indo-Pacific Islands (Nicobars, Andamans, New Britain,
Celebes, Christmas (Indian Ocean), Philippines, Loyalties, Formosa, New
Guinea, Pulu Weh, and Talauts.

Habitat: Moist areas along the coasts of islands, typically within the
jungle.

Habits: Burrows not well developed; generally shallow 15-60 cm in
length; mostly parallel with the surface; burrows principally for retreat.
Found only where there is a constant source of water to keep soil moist.
Andrews (1900) gives a good account of selected aspects of the life history
of this species. Gibson-Hill (1947) reports extensive migrations to the sea
during spawning together with an account of the aspects of development
of the young.

Common Name: None recorded.

References: Andrews, 1900 (B); Caiman, 1909 (B); A. Milne Ed-
wards, 1879 (T); H. Milne Edwards, 1834 (T), 1837 (T); Ortmann, 1894
(D,T); Pocock, 1888 (D); Rathbun, 1918 (D,T); Sakai, 1940 (D);Tweedie,
1947 (D,T), 1954 (B) ; Wood-Mason, 1873 (T).

Note: This is frequently listed as Gecarcoidea lalandii H. Milne Ed-
wards, 1837. The genus was erected by H. Milne Edwards (1837) for G.
lalandii which was erroneously recorded from Brazil, and then renamed
by the same author in 1853 as the genus Pelocarcinus. Wood-Mason (1873),
de Man (1879), and Pocock (1888) all described this under a number of
genera and species. Ortmann (1890) reduced all these species to synonymy
with Gecarcoidea lalandii except Pocock’s natalis. Tweedie (1947) gives
a good review of this taxonomic snarl and concludes that G. lalandii must

16

Contributions in Science

No. 220

be regarded as indeterminable on the basis of an incorrect type locality.
We use the name humei on the advice of Wood-Mason (1873). Tweedie
(1947) also supports the position that because these are generally island
forms there are several variants which can be recognized as subspecies, e.g.,
the crab found on Christmas Island (Indian Ocean) is G. humei natalis
(Pocock).

Genus Cardisoma Latreille, 1825

Characters: Fronto-orbital distance much more than half the greatest
width of the carapace; orbits deep with eyes filling half of the orbit; anten-
nules folded; epistome short and well defined; legs stout.

Distribution: Tropical America, Cape Verde Islands, west coast of
Africa, Indo-Pacific from Port St. Johns, Africa to Hawaiian Islands.

Habitat: Commonly inhabits muddy shores, mangrove swamps and
saline lowland soils near the coast.

Habits : Constructs well defined deep burrows in soft soils where ground
water is available during the dry season. Often they plug the burrow mouth
with mud during the dry season to keep the lower portions of the burrow
moist. Burrow sites are always above the mean high tide level. They return
to the sea to spawn and to introduce the pre-zoea to the required sea water
environment. All are primarily herbivorous but feed on carrion also.

References: Barnard 1950 (D,T); Behre, 1949 (B); Bright, 1966
(G); Gifford, 1963 (G); Herreid and Gifford, 1963 (B) ; Holthuis, 1959 (D);
Rathbun, 1918 (D,T);Tesch, 1918 (D,T).

16. Cardisoma guanhumi Latreille, 1825

Color: Carapace deep violet in young, but tends to become bluish
gray with age or approach to molt; ambulatory legs deep blue with larger
cheliped dirty white. Local variation may be due to the type of soil charac-
teristic of habitat.

Distribution: Atlantic coasts of America, central east coast of Florida,
Louisiana, Texas to Florianopolis, Brazil, including the West Indies.

Habitat: Open fields, margins of mangrove swamps, along margins of
rivers, in forests, along roads and under houses. In all known situations the
soil is saline.

Habits: All are typically found within a few hundred yards of a brackish
or saltwater source. The young are found under debris, e.g., coconut husks,
palm fronds, flotsum, etc., directly adjacent to salt water (generally above
the highest high tide level) or in the burrows of adults. Adults construct
burrows of varying depth and structural complexity, depending upon their
age and the location of the burrow with respect to available ground water.
Large aggregations of adult crabs are often found in areas where the sub-
stratum is soft yet still suitable for burrows, and where there is little ground
cover. Although adults occur several hundred meters from the sea, they are
most common within about 200 meters of the tidal zone. Inland dwelling

1972

Burrowing Land Crabs of the World

17

individuals migrate to the sea in great numbers during the breeding season
to shed their eggs. This is a twin species of C. crassum which occurs in the
Pacific.

Common Names: Great land crab; White land crab, Juey, Cangrejo;
Guanhumi; Mulatto land crab; Guaiamu, guarani, guayamu (Brazil).

References: Behre, 1949 (B); Bott, 1955 (D,T); Bright, 1966 (G);
Chace and Holthuis, 1948 (T); Feliciano, 1962 (B); Gifford, 1963 (G);
Herreid, 1963 (B), 1967 (B); Herreid and Gifford, 1963 (B); de Oliveira,
1946 (B); Pearse, 1916 (G); Peyton et al., 1964 (D); Rathbun, 1933 (D,T).

17. Cardisoma crassum Smith, 1870

Color: Carapace deep blue; dactyli of walking legs red; large cheliped
pale yellow to dirty white; underside cream-white.

Distribution: Pacific coasts of America. Todos Santos, Baja California
to the Rio Chira, Peru.

Habitat: Open fields, margins of mangrove swamps, along roads and
fence-rows, margins of rivers and streams, under houses and in cultivated
fields; generally in saline lowland soils near the coast.

Habits: These show habits similar to those of C. guanhumi in areas
adjacent to brackish or salt water sources. In contrast to C. guanhumi the
young construct separate shallow burrows along river banks and edges of
mangroves. Adult burrow construction parallels that of C. guanhumi. During
the dry season, adults with burrows in open, exposed areas close the top of
the burrow with a plug of mud. Some reports have indicated that closure
of the burrow also occurs prior to the onset of ecdysis (shedding). Adult
migrations to the sea during the spawning period are common in Mexico,
Costa Rica, Panama, and Ecuador. This is a geminate species of C. guanhumi
which occurs in the Atlantic.

Common Names: Mouthless crab, Cangrejo sin boca (Peru), Cajo
(Mexico).

References: Bott, 1955 (D,T); Bright, 1966 (G); Garth, 1948 (D,T),
1960 (D); Murphy, 1944; Pesta, 1931 (D,T) ; Rathbun, 1918 (D,T).

18. Cardisoma armatum Herklots, 1851

Color: Young, newly molted individuals with violaceous carapace; tips
of chelae and walking legs bright red; with age and approach to molt carapace
turns dirty yellow with occasionally slight reddish spots dorsally.

Distribution: Western coast of Africa from St. Louis to Baia dos
Tigres, Angola, Africa and Cape Verde Islands.

Habitat: Moist sandy areas above the mean high tide level; mangroves,
mouth of rivers, under houses, in cultivated areas adjacent to permanent
sources of brackish or sea water; and inland areas of larger islands.

Habits: Youngest juveniles are in small depressions or newly dug
shallow burrows directly adjacent to water; older juveniles found in smaller
compartments within the burrows of adults. Adults construct deep burrows,

18

Contributions in Science

No. 220

and often these are part of a large colony where the burrows intersect. Both
juveniles and adults are nocturnal scavengers, often moving considerable
distance from their burrows to feed on palmnuts, coconuts, dead fish and
scraps of vegetation. Spawning activities have not been recorded in the
literature for this species.

Common Name: Edible land crab.

References: Barnard, 1950 (D,T); Cheesman, 1922 (B), 1923 (B);
Dalziel, 1920 (D); Rathbun, 1921 (D,T) ; Wanson, 1935 (D).

19. Cardisoma carnifex (Herbst, 1794)

Color: Carapace dark purple; chelipeds light purple to dark cream.

Distribution : East coast of Africa, whole of Indo-Pacific, north eastern
Australia and north toward Japanese Mainland (Loo Choo), including
Mauritius, Madagascar, Andaman Islands, Malay Archipelago, Polynesia and
Melanesia.

Habitat: Common inhabitants of muddy shores, mangrove swamps or
Kuli and saline lowland soils near the coast. Not uncommon in the jungle
adjacent to the sea. All around the Indian Ocean it is most commonly found
between the high tide mark and just beyond the extreme highest high tide
line.

Habits: Constructs well defined burrows in soft soils where ground
water is available during the dry season. Habits generally parallel other
members of the genus. On coral atolls it is common among coconut husks,
under rubble piles and in mixed forest areas adjacent to plantations.

Common Names: Land crab; Kepiting Balong (Cocos Island), Papaka
Tupa (Tuamotu Islands).

References: Alcock, 1900 (D,T); Barnard, 1950 (D,T); Borradaile,
1902 (B); Forest and Guinot, 1961 (D); Hogue and Bright, 1971 (B) ; Hol-
thuis, 1953 (D); Macnae, 1963 (B,D), 1966 (B,D) ; Miyake, 1939 (D,T) ;
Silas and Sankarankutty, 1960 (B); Stebbing, 1910 (D); Tesch, 1918 (D,T);
Tweedie, 1950 (D,T).

Note: Over much of the range this occurs in sympatry with C. hirtipes
(see Miyake, 1939).

20. Cardisoma hirtipes Dana, 1851

Color: Carapace generally dark violet and chelae bright cinnamon
red. There is considerable color variation throughout the Indo-Pacific, e.g.,
Tweedie (1947) notes that on Christmas Island (Indian Ocean) the carapace
is light bluish gray and the chelae are dirty white.

Distribution: Occurs in sympatry with C. carnifex from east coast
of Africa throughout whole of Indo-Pacific. Miyake (1939) gives a good
account of the distinguishing species characteristics.

Habitat: Moist saline soils; mud banks in the immediate neighborhood
of fresh-water. Where soil is dense or crusted they frequently scratch-out a
space under a treeroot or rock.

1972

Burrowing Land Crabs of the World

19

Habits: Normally the crab digs burrows in the soft mud directly adjacent
to streams. The burrows are only 50-75 cm below ground level. Considering
our preliminary studies on related species it seems likely that this species
can abide in areas where there is more ground water than is characteristic
for C. carnifex. Perhaps it is the ecological equivalent of the American species
of the genus Ucides. A few weeks after the onset of the rainy season in
January or February spawning occurs. Copulation occurs at the edge of the
sea just prior to the time the female sheds the previous batch of eggs. They
are primarily plant feeders, occasionally causing crop damage to melons and
pumpkins (Esaki, 1940), but in areas associated with human habitation
they are carrion feeders as well.

Common Name: Land crab.

References: Alcock, 1900 (D,T); Esaki, 1940 (D); Gibson-Hill,
1947 (B); Gordon, 1934 (D); Miyake, 1939 (D,T) ; Sakai, 1940 (D);
Tesch, 1918 (D,T) ; Tweedie, 1947 (D,T).

Note: Tweedie (1950) restored Cardisoma frontalis H. Milne Edwards,
1853, to specific status from synonymy with Cardisoma hirtipes. He gives
the distribution of C. frontalis as Loyalty Islands, northern Daitozima, Japan,
and Cocos-Keeling Islands, and further states that examination of series
presently considered as C. hirtipes would probably result in extension of this
distribution. However, until there is an extensive revision of the genus with
clarification of the number of island endemics we will herein consider these
still to be C. hirtipes.

Genus Gecarcinus Leach, 1814

Characters : Fronto-orbital distance half or less than half of the greatest
width of the carapace; orbits deep with eyes nearly filling the orbits; anten-
nae very short; epistome linear; legs stout, the second pair being longest.

Distribution: Tropical America, Bermudas, Ascension Island, West
and South Africa, Australasia.

Habitat: Drier areas above the tidal margins of mangroves; river mouths
and adjacent coastal sandy and saline soil areas.

Habits: Burrows always shallow and devoid of ground water, except
during rain storms. Many utilize debris as a source of protection in lieu of
a burrow. In the extreme northern and southern portions of the distribution
the burrows are deep 1.2 m and often with mouth plugged during the dry
season.

References: Bright, 1966 (G); Finnegan, 1931 (D); Garth, 1948
(D,T) ; Chace and Hobbs, 1969 (G); Villalobos and Cabrera, 1964 (B).

21. Gecarcinus planatus Stimpson, 1860

Color: Body and legs generally an orange-red; tips of walking legs
often dark red; tips of chelae cream with small flecks of brown.

Distribution: Pacific coasts of America. Restricted to islands from
west coast of Mexico to Colombia.

20

Contributions in Science

No. 220

Habitat: Rocky areas, under roots and in soft soils above highest high
tide mark on slopes up to 120 m, often associated with beach strand vege-
tation.

Habits: Adults and older juveniles scratch out shallow burrows under
rocks, roots or debris. Burrow serves primarily as a hiding place. Young hide
in natural crevices and small spaces providing natural protection. Burrows
have no standing water. Newly metamorphosed juveniles hide along the
shore under debris, often gregariously. They are nocturnal feeders, and com-
monly move considerable distance from their burrows. Bold when on feeding
excursions, ambling over almost anything in their path, e.g., sleeping scien-
tists, food lockers, young birds, etc. This is a geminate species of G. ruricola
occurring in the Atlantic.

Common Names: Island crab; Big red crab; Cangrejo rojo (Panama).

References: Garth, 1948 (D,T), 1960 (D); Rathbun, 1918 (D,T).

22. Gecarcinus ruricola (Linnaeus, 1758)

Color: Body and legs generally black with purplish tinge; small light
yellowish spot on the posterior margin of the carapace; last two joints of
legs red; red and yellow patch below the orbit of the eye; abdomen light
yellow with violet hue; older individuals or those undergoing late preecdysial
changes are overall much lighter in color.

Distribution: Atlantic coasts of America. Restricted to islands: Ba-
hamas; southern Florida; greater and lesser Antilles; Curasao, and Cayman
Islands.

Habitat: Low and marshy areas not far from the sea; lower slopes of
island mountains up to 500 meters.

Habits: They hollow out obliquely inclined shallow burrows, which
are quite frequently under a tree or the edge of a large rock. After meta-
morphosis, the young are found in large numbers just above the high tide
level, however, very shortly after the second or third molt they move to
areas well above the highest high tide. Along the shore edge, they are often
found sympatrically with Gecarcinus lateralis. On larger islands, e.g., Isla
Providencia, Colombia, they are common along mountain slopes and cliffs
adjacent to the beach and to heights of 500 meters and as far as a thousand
meters from the shore. They are more secretive than most of the gecarcinids.
During the rainy season they are reported to move in large numbers to the
sea to breed. This is a geminate species of G. planatus in the Pacific.

Common Names: Black crab; Mountain crab; Blue land-crab. Red
tourlourou.

References: Chace and Hobbs, 1969 (G); Chace and Holthuis, 1948
(T); Rathbun, 1918 (D,T).

23. Gecarcinus quadratus Saussure, 1853

Color: Carapace brownish red with two white spots in the cardiac
region, intestinal region orange-red; large chelipeds with light purple tinge;

1972

Burrowing Land Crabs of the World

21

merus of maxilliped light yellow; underside sooty white.

Distribution: Primarily Pacific coasts of America. (Atlantic side of
Colombia). Not known to occur on permanently isolated islands. Acapulco,
Mexico to La Libertad, Ecuador; Turbo.

Habitat: Above the highest tide zones of sandy beaches in moist forest
and mangrove areas where there is low growing vegetation or debris.

Habits: All occur in drier areas directly adjacent to mangroves or the
ocean. They are most common along the uppermost areas of sandy beaches
from 1.5-15 m above the supratidal area where there is a dense covering
of debris, e.g., coconut husks and fronds, or low growing beach strand vege-
tation, e.g., lpomoea (family Convolvulaceae). Burrow construction is
correlated with the length of the dry season. Crabs at both extremes of the
distribution tend to construct simple burrows, 7-50 cm deep, while at mid-
range, burrows are not common. Non-burrowers tend to occupy small de-
pressions under vegetation debris, houses, etc. This is a geminate species of
G. lateralis in the Atlantic.

Common Names: Red land crab; Whitespot crab.

References: Bright, 1966 (G); Finnegan, 1931 (D); Garth, 1948
(D,T); Pesta, 1931 (D,T).

24. Gecarcinus lateralis (Freminville, 1835)

Color: Carapace dark red with small white spots just posterior to the
eyes and a pair of white spots in the cardiac region; underside cream-white;
chelipeds reddish gray; dactyli sooty gray. Pattern of dark red carapace
highly variable throughout range of distribution and in distinct (isolated)
populations.

Distribution: Atlantic coasts of America. Bahamas; Florida Keys;
South Padre Island, Texas; Yucatan; to Macuto, Venezuela. Also occurs
on islands in the West Indies.

Habitat: Along the upper dry zone of sandy beaches and adjacent low
hills, 6 to 9 m above highest high tide level; associated with a variety of
beach strand vegetation, e.g., coconuts, and low growing vines, e.g., lpomoea.

Habits: All occur in nearly dry areas, i.e., where there is no standing
water but a good bit of interstitial soil moisture. Burrow construction is as
for G. quadratus. There is also a tendency for the depth of the burrow to
be correlated with the length of the dry season. Burrows are deeper on the
extremes of the range of distribution.

Common Names: Black land-crab; Common land-crab.

References: Bliss, 1964 (B); Bott, 1955 (D,T); Bright, 1966 (G);
Cabrera, 1965 (B); Chace and Hobbs 1969 (G); Chace and Holthuis, 1948
(T); Pearse, 1916 (G); Rathbun, 1918 (D,T), 1933 (D,T); Ray, 1967 (D).

ARTHROPOD INHABITANTS OF LAND CRAB BURROWS

The following list represents an attempt to cite all published records
of arthropods found in land crab burrows. Because the host was not identi-

22

Contributions in Science

No. 220

fied in all cases and there is frequent confusion in usage of the terms land
crab, crab, lobster, mud lobster, crayfish, etc., references to all are included.
Accounts of dubious validity are also included for completeness and to
establish the need for verification.

Some explanation is in order regarding certain assumptions made and
conventions used in compiling the list. 1) Identifications and associations
with the host crab, i.e., the determinations of the real crab owner of the
burrow from which the associates were collected, are assumed to be correct.
The list has been read by various specialists and it is hoped that errors of
identification, erroneous records and synonymies have been largely detected.
2) The type of association and degree of dependence (see Table I and dis-
cussion of the Crabhole Community above) of the species on the crabhole
were ascertained or inferred from all available information on the biology
of the species. Dubious decisions and the criteria for allocation to a category
are explained where relevant.

INSECTS
Order DIPTERA

Family CULICIDAE

We found it impossible to scour the voluminous literature on mosquitoes
for all records of species utilizing crabholes. Fortunately, for most regions,
comprehensive (though not always current) reviews including ecological
data are available (Belkin, 1962: South Pacific; Steffan, 1966: Papuan Region;
Hopkins, 1952: Ethiopian Region; Dyar, 1928: tropical America; Delfinado,
1966: Philippines; Barraud, 1934: India; Mattingly, 1958, 1959: Indo-
malayan Region).

Presently 140 species of mosquitoes are recorded as either resting as
adults or breeding in crabholes. To this list could be added several more
from unpublished works known to us and no doubt others from other
studies now under way. Surprisingly no culicidologist has attempted previously
an exclusive investigation of this habitat. Most authors, with a few notable
exceptions, seem to regard the crabhole as an aberrant breeding site being
utilized by only a handful of specially adapted species. Our bibliographic
research would indicate that, whereas those taxa specifically adapted to
crabholes are indeed few, the number of transient species is much larger
than previously suspected. We feel that much more attention should be paid
to crabholes in general mosquito surveys than has been customary in the
past (see remarks of Peyton, 1970:2-3). Crabholes are easily sampled using
the crabhole pump and collection methods described by Belkin et al. (1965:
37-38).

Most of these transient crabhole breeders are salt water adapted or
tolerant species which are general ground pool breeders along the coast.

1972

Burrowing Land Crabs of the World

23

Their usual normal habitats consist of salt marshes, mangrove pot holes, tidal
pools, puddles, coral rock pools, etc. A considerable number of container
breeding species (tree holes, especially in mangroves, and artificial containers)
also are found in the crabhole. From this it seems probable that the specific
crabhole mosquito fauna is of mixed origin derived from both of these
more primitive categories through convergent adaptations (van den Assem,
1961:19).

The nature of these adaptations is virtually unstudied. Some charac-
teristics frequently observed in crabhole species are as follows: 1. Stubby or
vestigial anal papillae on the larvae. This condition is common, but by no
means universal, among larvae which develop in waters with high salt con-
tent. 2. Short head hair 1-C, the adaptive significance of which is totally
obscure. 3. Prolonged developmental period. 4. Very specialized and aberrant
reproductive behavior such as pupal attendance and lack of swarming in
Deinocerites (Provost and Haeger, 1967) and oviposition directly on the
host crab in Aedes pembaensis (Goiny et al., 1957).

At least a few crabhole mosquitoes are of known public health im-
portance. Two primary vectors of serious diseases in Africa, while not specific
or even semispecific members of the crabhole community, nevertheless may
develop in tremendous numbers in this habitat and may even find refuge
there during eradication programs designed to treat only the more usual
breeding sites. These species are Aedes aegypti and Anopheles gambiae (?
melas, merus ) for both of which there are several well authenticated records
of breeding in crabholes (see list). Vectors of filariasis on the east African
coast, Aedes pembaensis, and in the south Pacific, Aedes polynesiensis, are
semispecific members of the crabhole community. The former species has
also been found to harbor several kinds of viruses of unknown but possible
pathogenicity (Heisch et al., 1956), as have various other mosquitoes which
develop in crabholes, including Deinocerites. The eastern Equine, Venezuelan
and St. Louis encephalitis viruses have all recently been isolated from D.
pseudes in Panama (Galindo, 1967; Templis and Galindo, 1970:175; Gray-
son, 1967). Trypanosome organisms have also been recently isolated from
wild adults of this species in Panama (Gorgas Mem. Lab., 1970). While
Deinocerites do not appear to be strongly anthropophilic they may act as
important agents in maintaining virus reservoirs in coastal animal popu-
lations (silent cycles) and which may enter the human population via other
vectors.

Genus Aedes
Subgenus Aedimorphus

A. abnormalis (Theobald, 1910)

Distribution: Tropical Africa; interior and coastal.
Crab Host: Cardisoma armatum.

24

Contributions in Science

No. 220

Type: Transient. Adults only. Probably breeds in ground and rock
pools.

Reference1: Wanson, 1935:576.

A. albocephcilus (Theobald, 1903)

Distribution: Tropical Africa, Madagascar, Seychelles; interior and
coastal.

Crab Host: Not recorded.

Type: Accidental. Normally breeds in grassy swamps, pools, etc., in
interior; saline seepage pools on coast.

References: Hopkins, 1952:182; Ingram and Macfie, 1917:142 (and
as minutus ).

A. caliginosus (Graham, 1910)

Distribution: Nigeria; coastal.

Crab Host: Cardisoma armatum.

Type: Transient. Adults only. Recorded also from borrow pits and
stream pools.

Reference: Dalziel, 1920:253.

Note: Identification dubious.

A. centropunctatus Theobald, 1913

Distribution: Sudan, British West Africa; interior by watercourses (?).
Crab Host: Sudanonautes africanus.

Type : ? Bionomics insufficiently known.

Reference: Hanney, 1960:99.

A. domesticus (Theobald, 1901)

Distribution: Tropical Africa; interior and coastal.

Crab Host: Cardisoma armatum.

Type: Transient. Usually breeds in grassy swamps, borrow pits, etc.
Reference: Wanson, 1935:576-577, 579.

A. durbanensis (Theobald, 1903)

Distribution: Africa, Arabia; primarily coastal.

Crab Host: Cardisoma armatum.

Type: Transient. Usually breeds in fresh water ground pools.
Reference: Wanson, 1935:576-578.

A. fowleri (Charmoy, 1908)

Distribution: Africa; interior and coastal.

Crab Host: Cardisoma armatum.

Unless followed by symbols indicating additional significant information (T-tax-
onomy; B-biology; D-distribution; G-general), the references cite only records of
the occurrence of the species in crabholes.

1972

Burrowing Land Crabs of the World

25

Type: Transient. Usually breeds in rock pools and grassy ground pools.

Reference: Wanson, 1935:576-577, 579 (as nigeriensis) .

A. irritans (Theobald, 1901)

Distribution: West and Central Africa; primarily coastal.

Crab Host: Cardisoma armatum.

Type: Semispecific. Breeds also in small brackish pools along the coast.
Strongly anthropophilic.

References: Bruce-Chwatt and Fitz-John, 1951:119-120 (B); Dalziel,
1920:251-253; Dunn, 1928:249; Ingram & Macfie, 1917:135; Kumm,
1931:65; Wanson, 1935.

A. nigricephalus (Theobald, 1901)

Distribution: West Africa; coastal.

Crab Host: Cardisoma armatum.

Type: Semispecific. Also breeds in ground pools.

References: Bruce-Chwatt and Fitz-John, 1951:119; Dalziel, 1920:
251; Dunn, 1928:249; Kumm, 1931:65; Wanson, 1935.

A. punctothoracis (Theobald, 1910)

Distribution: West tropical Africa; primarily coastal.

Crab Host: Cardisoma armatum.

Type: Accidental. Adults only. Normally breeds only in ground pools.

References: Dalziel, 1920:253; Wanson, 1935:576.

A. tarsalis (Newstead, 1907)

Distribution: Tropical Africa; interior and coastal.

Crab Host: Cardisoma armatum.

Type: Transient. Usually breeds in rock and ground pools.

References: Kumm, 1931:65 (as sudanensis)\ Macfie and Ingram,
1916:7 (as sudanensis) ; Wanson, 1935:576-577.

Note: May be confused with centropunctatus.

Subgenus Cancraedes

General Reference: Mattingly, 1958.

All species for which the immatures are known (*) are found breeding
primarily in crabholes from which all have been taken as adults. The entire
subgenus appears to be adapted to this habitat and comprises a specific or
semispecific member of the crabhole community though mangrove pot holes
and coastal ground pools may serve as secondary breeding sites. All species
have a coastal distribution or are found on small islands. None of the crab
hosts has been identified.

A. cancricomes Edwards, 1922

Distribution : Andaman Islands.

*A. curtipes Edwards, 1915

Distribution : Borneo, Philippines, Malaya, ? Thailand.

26

Contributions in Science

No. 220

*A. indonesiae Mattingly, 1958
Distribution: Java, Sumatra, east Gulf of Siam.

A. kohkutensis Mattingly, 1958
Distribution : Thailand.

A. mamoedjoensis Mattingly, 1958
Distribution: Celebes.

*A. masculinus Mattingly, 1958
Distribution: Malaya, ? Philippines.

A. palawanicus Mattingly, 1958
Distribution: Philippines.

A. penghuensis Lien, 1968
Distribution: Taiwan.

A. simplex (Theobald, 1903)

Distribution: Ceylon.

A. thurmanae Mattingly, 1958
Distribution: Celebes.

Subgenus Geoskusea

General References: Mattingly, 1959; Belkin, 1962:332-339.

As with the preceding, crabholes in coastal areas are the primary breed-
ing places of all species in this subgenus for which the immatures are known
( * ) ; the adults of all species have been taken from this habitat. Thus all can
probably be classified as specific or semispecific members of the crabhole
community. None of the crab hosts has been identified.

A. baisasi Knight and Hull, 1951
Distribution: Philippines.

A. becki Belkin, 1962
Distribution : Solomons.

*A. daggyi Stone and Bohart, 1944
Distribution: New Hebrides, Solomons.

A. daliensis (Taylor, 1916)

Distribution: Australia.

A. fimbripes Edwards, 1924
Distribution: Bismark Archipelago, New Guinea.

*A. kabaenensis Brug, 1939
Distribution: Celebes.

*A. longijorceps Edwards, 1929
Distribution : Solomons.

1972

Burrowing Land Crabs of the World

27

Note: Adults have been observed feeding on mud skippers (Perioph-
thalmus musgravei ) resting on mangrove roots in the Solomon Islands (Sloof
and Marks, 1965).

A. perry i Belkin, 1962
Distribution : Solomons.

A. tonsus Edwards, 1924
Distribution : Moluccas.

Subgenus Howardina
A. inaequalis (Grabham, 1907)

Distribution: Jamaica; interior and coastal.

Crab Host: Not recorded.

Type: Transient. Most commonly breeds in treeholes and broken
bamboo.

Reference: Berlin, 1969:48.

A. walkeri Theobald, 1901
Distribution: Jamaica, interior and coastal.

Crab Host: Not recorded.

Type: Transient. Normally breeds in bromeliads.

Reference: Berlin, 1969:35.

Subgenus Levua
A. suvae Stone and Bohart, 1944
Distribution: Fiji; coastal.

Crab Host: Not recorded, “crab and lobster holes.”

Type: Specific. No other recorded breeding site.

References: Amos, 1944: 32; Belkin, 1962:400 (G).

Subgenus Mucidus

A. aurantius chrysogaster (Taylor, 1927)

Distribution: Australia, New Guinea; coastal.

Crab Host: Not recorded, “crab pot hole.”

Type: Transient. Usually breeds in various types of ground pools in
coastal areas. Larva predaceous.

Reference: Steffan, 1966:206.

A. scatophagoides (Theobald, 1901)

Distribution: India, Ceylon, Burma, China, tropical Africa; interior
and coastal.

Crab Host: Cardisoma armatum.

Type: Transient. Usually breeds in transient ground pools and marshes.
Larva predaceous.

Reference: Wanson, 1935:576-577, 579.

28

Contributions in Science

No. 220

Subgenus Neomacleaya
A. dux Dyar and Shannon, 1925
Distribution: Southeast Asia; coastal.

Crab Host: Not recorded.

Type: Transient. Adults only. Usually breeds in puddles and hoof prints
near the coast; prefers saline water?

Reference: Delfinado, 1967:20.

A. panayensis Ludlow, 1914

Distribution: Philippines, Moluccas, New Guinea; coastal.

Crab Host: Not recorded.

Type: Transient. Adults only. Usually breeds in marine littoral ground
pools.

References: Delfinado, 1968:33; Steffan, 1966:214.

Subgenus N eomelanoc onion
A. lineatopennis (Ludlow, 1905)

Distribution: Oriental Region, tropical Africa, Australia; interior and
coastal.

Crab Host: Not recorded.

Type: Transient. Adults only. Breeds usually in vegetated ground pools.
Reference: Wanson, 1935:576.

Subgenus Ochlerotatus
A. perventor Cerqueira and Costa, 1946
Distribution: Brazil; interior and coastal.

Crab Host: Cardisoma guanhumi.

Type: Probably transient. Bionomics poorly known. Only breeding
records from crabholes.

References: Forattini, 1958: 177-178; Forattini et al., 1958:37 (B).

A. taeniorhynchus (Wiedemann, 1821)

Distribution: American coasts and interior saline areas.

Crab Host: Cardisoma guanhumi.

Type: Transient. Salt water breeder, usually found in coastal salt marshes,
tide pools, etc., and inland saline sinks. Strongly anthropophilic.

References: Belkin et al., 1970:49 (G); Forattini, 1958:175-177 (B);
Lutz, 1912:19 (as Culex taeniorhynchus) (B); Montchadsky and Garcia,
1966:42; de Oliveira, 1946:297.

Subgenus Paraedes
A. bonnae Mattingly, 1958
Distribution: Malaya; coastal.

Crab Host: Not recorded.

Type: Transient. Usually breeds in ground pools (palm fronds also
recorded).

Reference: Mattingly, 1958:34.

1972

Burrowing Land Crabs of the World

29

Subgenus Pseudarmigeres
A. albomarginatus (Newstead, 1907)

Distribution: Central tropical Africa; interior and coastal.

Crab Host: Car disoma armatum.

Type: ? Single breeding record, from crabhole.

Reference: Wanson, 1935:577.

Subgenus Pseudoskusea
A. lunulatus King and Hoogstraal, 1946

Distribution: New Guinea; coastal.

Crab Host: Not recorded, “crayfish hole.”

Type: ? Single breeding record, from crayfish hole in shaded rain forest,
250 feet elevation.

Reference: King and Hoogstraal, 1946a:97.

Subgenus Rhinoskusea
A. longirostris (Leicester, 1908)

Distribution: Indomalayan, north Australian and Papuan Regions;
coastal.

Crab Host: Not recorded.

Type: Transient. Usually breeds in numerous other marine littoral
ground pool habitats, especially mangrove swamp pools, and artificial
containers.

References: Colless, 1957:144; Leicester, 1908:8 (as Ficalbia longi-
rostris); Mattingly, 1958:39-40.

A. pillaii Mattingly, 1958

Distribution: Malaya; coastal.

Crab Host: Not recorded.

Type: Transient. Usually breeds in numerous other marine littoral ground
pool habitats.

Reference: Reid in litt. after Mattingly, 1958:40.

Subgenus Skusea
A. pembaensis Theobald, 1901

Distribution: East Africa, Madagascar, Seychelles; coastal.

Crab Hosts: Sesarma meinerti, S. eulimene.

Type: Semispecific. Predominantly breeds in crabholes; also commonly
utilizes ground pools and swamps and rarely natural and artificial containers.
The females deposit their eggs on the legs and body of the host. (Goiny et al.,
1957; Hogue and Bright, 1971). Females are strongly anthropophilic and
vectors of filiariasis in east Africa (Heisch, Goiny and Ikata, 1957).

References: Brook Worth et al., 1961 (B); Hopkins, 1952:224; Lums-
den, 1955: 170-171 (B).

30

Contributions in Science

No. 220

Subgenus Stegomyia
A. aegypti (Linneaus, 1962)

Distribution: Cosmopolitan; interior and coastal.

Crab Hosts: Cardisoma armatum, Sesarma africanum.

Type: Transient. Usually breeds in various artificial container habitats.
Strongly anthropophilic and vector of yellow fever.

References: Bruce-Chwatt and Fitz-John, 1951:120; Cheneveau,
1934:590-593; Dalziel, 1920:248, 251-252; Dunn, 1928:249; Riqueau, 1929;
Symes, 1960:5, 8; Wanson, 1935:576 (as argenteus).

A. africanus (Theobald, 1901)

Distribution: Tropical Africa; interior and coastal.

Crab Host: Cardisoma armatum.

Type: Accidental. Normally breeds in treeholes. Anthropophilic and
vector of yellow fever.

Reference: Wanson, 1935:576-577.

A. luteocephalus (Newstead, 1907)

Distribution: Tropical Africa; interior and coastal.

Crab Host: Cardisoma armatum.

Type: Transient or accidental. Normally breeds in treeholes (also in
cut bamboo, rockholes and temporary ground pools).

References: Dalziel, 1920:251-252; Wanson, 1935:576.

A. polynesiensis Marks, 1951
Distribution: South Pacific islands; coastal.

Crab Hosts: Cardisoma carnifex, C. hirtipes.

Type: Semispecific. Usually breeds in containers of various sorts. Strongly
anthropophilic and vector of filariasis and dengue. Crabholes important
breeding sites when other habitats absent as on low coral islands.

References: Belkin, 1962:468, pi. 2; Burnett, 1960; Symes, 1960:5, 8;
Tamashiro, 1964:10-11 (B).

A. pseudoscutellaris (Theobald, 1910)

Distribution: Fiji; coastal.

Crab Host: Not recorded.

Type: Transient. Usually breeds in containers of various sorts.
Reference: Belkin, 1962:470, pi. 2.

Subgenus Verralina
A. butleri Theobald, 1901
Distribution: Indomalayan Region; coastal.

Crab Host: Not recorded.

Type: Probably semispecific. Most commonly known from ground pools
in mangroves and nipa palm axils.

References: Edwards, 1928:346 (as umbrosus ); Leicester, 1908:8.

1972

Burrowing Land Crabs of the World

31

A. parasimilis King and Hoogstraal, 1947

Distribution: New Guinea; primarily coastal?

Crab Host: Not recorded, “crayfish hole.”

Type: Transient. Usually breeds in various types of ground pools.

References: King and Hoogstraal, 1947: 125; van den Assem, 1961 :25.

Genus Anopheles
Subgenus Anopheles
A. tigertti Scanlon and Peyton, 1967

Distribution: Thailand; interior.

Crab Host: Not recorded. Fresh water species.

Type: Semispecific or specific.

Reference: Scanlon and Peyton, 1967.

Subgenus Cellia

A. gambiae Giles, 1902, complex.

Distribution: Africa; interior and coastal.

Crab Hosts: Cardisoma armatum, Sesarma africanum.

Type: Transient. Usually breeds in a wide variety of artificial and natural
ground habitats. Anthropophilic and a vector of malaria.

References: Aders, 1917:393-394; Bruce-Chwatt and Fitz-John,
1951:120; Cheneveau, 1934:590-593 (as costalis ); Dalziel, 1920:251-253;
Dunn, 1928:249; Ingram and Macfie, 1917:135 (as costalis)', Macfie and
Ingram, 1916:7; Wanson, 1935:576, 578.

Note: The salt water species (?) merus (Donitz, 1902) or melas (Theo-
bald, 1903) may be found ultimately to be those associated with crabholes
(see Coluzzi, 1964).

Subgenus Nyssorhynchus
A. albimanus Wiedemann, 1821

Distribution: Tropical America, South America; interior and coastal.

Crab Host: Not recorded.

Type: Accidental. Flushed into crabhole by heavy rains; normally breeds
in vegetated ground pools and sluggish streams.

References: Shropshire and Zetek, 1927:338 (as tarsimaculata in
part); Belkin et al., 1970:49.

Genus Armigeres
A. breinli (Taylor, 1914)

Distribution: New Guinea, Bismark Archipelago, Solomons; primarily
coastal.

Crab Host: Not recorded.

Type: Accidental. Adults only. Species breeds in plant containers and
rarely ground pools.

References: Peters, 1963:10; Steffan, 1966:215.

32

Contributions in Science

No. 220

Genus Culex
Subgenus Aedinus
C. bisulcatus (Coquillett, 1906)

Distribution: Guadeloupe Island, Lesser Antilles; interior and coastal.
Crab Host: Not recorded.

Type: Accidental. Normally breeds in bromeliads.

Reference: Floch and Abonnenc, 1945:39.

Note: Probably adults only in crabholes or burrows.

C. corrigani Dyar and Knab, 1907
Distribution: Panama; coastal and interior.

Crab Host: Not recorded.

Type: ? Bionomics poorly known. Type series from bamboo joints.
Reference: Dyar, 1928:347.

C. latisquama (Coquillett, 1906)

Distribution: Tropical America; coastal (Atlantic only?)

Crab Host: Not recorded.

Type: Specific. Multiple collections, all from crabholes.

Reference: Howard, Dyar and Knab, 1915:305.

Subgenus Culex

C. annulioris prob. ssp. consimilis Newstead, 1907
Distribution: Tropical Africa; interior and coastal.

Crab Host: Cardisoma armatum.

Type: Transient. Usually breeds in ground pools.

References: Bruce-Chwatt and Fitz-John, 1951:119; Dalziel, 1920:253;
Wanson, 1935:576, 578.

C. annulirostris Skuse, 1889

Distribution: Southern and western Australasian Region, Indonesia,
Philippines; interior and coastal.

Crab Host: Not recorded.

Type: Transient. Indiscriminate breeder; usually breeds in ground pools
but also occurs commonly in almost all other habitats.

Reference: Belkin, 1962:pl. 2.

C. carcinoxenus Castro, 1932
Distribution: Brazil; coastal.

Crab Hosts: Cardisoma guanhumi, Ucides cordatus.

Type: Specific or semispecific. Known only from crabholes.
References: Castro, 1932:97; Forattini, Rabello and Heredia, 1956:85.

C. corniger Theobald, 1903
Distribution: American Mediterranean Region.

Crab Host: Ucides cordatus.

Type: Transient. Indiscriminate breeder.

References: Howard, Dyar and Knab, 1915:246; Lutz, 1912:19.

1972

Burrowing Land Crabs of the World

33

C. decens Theobald, 1901

Distribution: Ethiopian Region; interior and coastal.

Crab Host: Cardisoma armatum.

Type: Transient. Indiscriminate breeder.

References: Bruce-Chwatt and Fitz- John, 1951 : 1 19; Dalziel, 1920:251,
253; Dunn, 1928:249.

Note: May be confused with invidiosus.

C. duttoni Theobald, 1901

Distribution: Ethiopian Region; interior and coastal.

Crab Host: Cardisoma armatum.

Type: Transient. Usually breeds in a wide variety of ground habitats as
well as in containers.

Reference: Wanson, 1935:576, 578.

C. foliaceus Lane, 1 945
Distribution: Brazil; interior and coastal.

Crab Host: Not recorded.

Type: Probably transient. Bionomics poorly known.

Reference: Stone, 1950:239.

C. habilitator Dyar and Knab, 1906
Distribution: Antilles and Trinidad; coastal.

Crab Host: Not recorded.

Type: Semispecific. Most records from crabholes; also breeds in ground
pools and pot holes.

References: Bonne and Bonne-Webster, 1925:189; Dyar, 1928:362;
Howard, Dyar and Knab, 1915:262 (as eremita ); Pratt and Seabrook,
1952:27.

C. inflictus Theobald, 1901
Distribution: Tropical America; coastal.

Crab Host: Not recorded.

Type: Specific or semispecific. All records from crabholes.

References: Dyar, 1928:391; Hogue and Wirth, 1968:6; Howard,
Dyar and Knab, 1915:327 (as extricator ); Knab, 1910:868-869 (as extri-
cator) .

Note: A complex of species.

C. invidiosus Theobald, 1901
Distribution: Tropical Africa; interior and coastal.

Crab Host: Cardisoma armatum.

Type: Transient. Indiscriminate breeder.

Reference: Wanson, 1935:576, 578-579.

Note: May be confused with decens.

34

Contributions in Science

No. 220

C. janitor Theobald, 1903

Distribution: Greater Antilles; coastal.

Crab Host: Not recorded.

Type: Specific. Not recorded from other sites.

Reference: Belkin et al., 1970:49; Grabham, 1905:406-407; Hill and
Hill, 1948:55.

C. nigripalpus Theobald, 1901

Distribution: Tropical America; coastal (Atlantic).

Crab Host: Not recorded.

Type: Transient. General ground pool breeder, sometimes found in
crabholes.

References: Belkin et al., 1970:49; Branch and Seabrook, 1959:216;
Martini, 1914:70 (as prasinop\eurus)\ Pratt et al., 1945:246.

C. perfuscus Edwards, 1914

Distribution: Tropical Africa; interior and coastal.

Crab Host: Cardisoma armatum.

Type: Transient. Usually breeds in ground pools.

Reference: Wanson, 1935:576, 578-579.

C. philipi Edwards, 1929

Distribution: Western tropical Africa; interior? and coastal.

Crab Host: Not recorded.

Type: Transient. Type series bred from larvae found in crabholes. Also
found in vegetated pools.

Reference: Edwards, 1929:327.

C. pipiens quinquefasciatus Say, 1823

Distribution: Cosmopolitan; interior and coastal.

Crab Hosts: Cardisoma armatum , Sudanonautes africanus.

Type: Transient. Usually breeds in foul ground pools and ditches and
large artificial containers. Vector of filariasis over wide areas of world.

References: Dunn, 1928:249; Hanney, 1960:99; Wanson, 1935:576,
578-579.

C. pruina Theobald, 1901

Distribution: Western and central Africa; interior and coastal.

Crab Host: Not recorded.

Type: Transient. Usually breeds in ground pools with decaying leaves.
Reference: Dunn, 1928:249.

C. scimitar Branch and Seabrook, 1959

Distribution: Bahamas, small islands.

Crab Host: Not recorded.

1972

Burrowing Land Crabs of the World

35

Type: Transient? Bionomics poorly known; recorded also from ground
pools.

Reference: Branch and Seabrook, 1959:216.

C. sitiens Wiedmann, 1828

Distribution: Oriental Region, east Africa and western Pacific; coastal.
Crab Host: Not recorded.

Type: Transient. Normally breeds in saline or brackish coastal ground
waters. Anthropophilic.

Reference: van Someren et al., 1955:487.

C. thalassius Theobald, 1903
Distribution: Tropical Ethiopian Region; coastal.

Crab Host: Cardisoma armatum, Uca tangeri, Sesarma africanum.
Type: Semispecific. Multiple records from crabholes. Also breeds com-
monly in saline ground pools and artificial containers. Strongly anthropophilic.

References: Bruce-Chwatt and Fitz-John, 1951:119-120; Dalziel,
1920:251-253; Hopkins, 1952:286; Ingram and Macfie, 1917:147-149;
Wanson, 1935:576, 578-579.

Subgenus Culiciomyia
C. cinerellus Edwards, 1922
Distribution: Ethiopian Region; interior and coastal.

Crab Host: Cardisoma armatum.

Type: Semispecific or transient. Numerous records from crabholes.
References: Dalziel, 1920:251-253 (as nebulosus ); Dunn, 1928:249;
Edwards, 1929:327; Wanson, 1935:578.

C. nailoni King and Hoogstraal, 1946
Distribution: New Guinea; interior and coastal.

Crab Host: Not recorded, “crabhole in rain forest.”

Type: ? Bionomics insufficiently known. One record from crabhole.
Reference: King and Hoogstraal, 1946b.

C. ruthi Peters, 1958
Distribution: New Guinea; coastal.

Crab Host: Not recorded.

Type: ? Bionomics insufficiently known. Adults only, captured at the
entrance of small crabholes in partial shade on the beach. Immatures unknown.
Reference: Steffan, 1966:219.

C. spathifurca (Edwards, 1915)

Distribution: Oriental and Indomalayan Regions; interior and coastal.
Crab Host: Not recorded.

Type: Transient. Adults only. A general ground pool breeder.
Reference: Carter and Wijesundara, 1948:145 (as stylifurcatus) .

36

Contributions in Science

No. 220

Subgenus Lophoceraomyia
C. becki Belkin, 1962

Distribution: Solomons; small islands.

Crab Host: Not recorded.

Type: Specific. Known only from crabholes.

Reference: Belkin, 1962:266, pi. 2.

C. infantulus Edwards, 1922

Distribution: Oriental (including Japan) and Indomalayan Regions;
primarily coastal (?)

Crab Host: Not recorded.

Type: Transient. A general ground pool and container breeder.
Reference: Bram, 1967:61.

C. pholeter Bram and Rattanarithikul, 1967

Distribution: Thailand; interior.

Crab Host: Not recorded.

Type : Specific. Collected repeatedly and exclusively from small crabholes
in secondary rain forests in mountainous terrain.

Reference: Bram and Rattanarithikul, 1967:13.

C. reidi, Colless, 1965

Distribution: Singapore, Selangor; coastal.

Crab Host: Not recorded.

Type: Transient. Usually breeds in shaded pools at margin of tidal zone.
Reference: Colless, 1965:280.

C. rubithoracis (Leicester, 1908)

Distribution: Indomalayan Region, Japan; interior and coastal.

Crab Host: Not recorded.

Type: Transient. Ground pool breeder.

Reference: Macdonald, 1957:29.

C. variatus (Leicester, 1908)

Distribution: Indomalayan Region; coastal.

Crab Host: Not recorded.

Type: Transient. Usually breeds in ground pools but also utilizes con-
tainers near the ground.

Reference: Colless, 1965:273.

Subgenus Lutzia

C. tigripes Grandpre and Charmoy, 1901

Distribution: Ethiopian Region; interior and coastal.

Crab Host: Cardisoma armatum.

1972

Burrowing Land Crabs of the World

37

Type: Transient. Indiscriminate breeder. Larva predaceous.

Reference: Wanson, 1935:578.

Subgenus Melanoconion
C. opisthopus Komp, 1926

Distribution: Tropical America; coastal (Atlantic).

Crab Host: Cardisoma guanhumi.

Type: Accidental. Probably normal breeding site deep seepage channels
or solution holes in coral.

References: Pratt et al., 1945:246; Stone and Hair, 1968:41 (as
cedecei) \ Belkin, 1969 (T).

C. carcinophilus Dyar and Knab, 1906

Distribution: Dominican Republic, ? Guatemala, Cuba; coastal.

Crab Host: Cardisoma guanhumi.

Type: Specific.

References: Dyar and Knab, 1906:220; Montchadsky and Garcia,
1966:46.

C. iolambdis Dyar, 1918

Distribution: Tropical America; primarily coastal (Atlantic).

Crab Host: Not recorded.

Type: Transient. Usually breeds in coastal ground pools shaded by
mangroves.

Reference: Pratt and Seabrook, 1952:27.

C. nicaroensis Duret, 1967

Distribution: Cuba; coastal.

Crab Host: Not recorded.

Type: ?

Reference: Duret, 1967:80.

Subgenus Mochthogenes
C. inconspicuosus (Theobald, 1908)

Distribution: Ethiopian Region; interior and coastal.

Crab Host: Cardisoma armatum.

Type: Transient. Usually breeds in nearly stagnant pools, in streams
and in ground pools.

Reference: Dalziel, 1920:251-252, 254.

C. laureli Baisas, 1935

Distribution: Philippines; interior and coastal.

Crab Host: Not recorded.

Type: Transient. Usually breeds in small vegetated ground pools.
Reference: Delfinado, 1966:133.

38

Contributions in Science

No. 220

Subgenus Neoculex
C. insignis (Carter, 1911)

Distribution: Africa; interior and coastal.

Crab Host: Sudanonautes africanus, Car disoma armatum.

Type: Transient. Usually breeds in foul water in pooled streams.

References: Dalziel, 1920:251-253; Dunn, 1928:249; Hanney, 1960:
99; Macfie and Ingram, 1916:11.

C. rima Theobald, 1901

Distribution: West and central Africa; interior and coastal.

Crab Host: Cardisoma armatum.

Type: Specific or semispecific. Multiple records only from crabholes.

References: Bruce-Chwatt and Fitz-John, 1951:119; Dalziel, 1920:
251-253; Philip, 1931:192; Surtees, 1958:90; Wanson, 1935:576, 578.

C. salisburiensis Theobald, 1901

Distribution: Ethiopian Region; interior and coastal.

Crab Host: Cardisoma armatum.

Type: Transient. Usually breeds in forest ground pools and streams.

Reference: Dalziel, 1920:251-253.

Genus Deinocerites

General References: Adames, 1971; Belkin and Hogue, 1959.

Deinocerites normally breeds in crabholes. Referred to as “Crabhole
Mosquitoes” in the literature they unquestionably are specific members of the
crabhole community, being found breeding outside this habitat only very
rarely in such related or proximate places as mangrove treeholes, mangrove
pot holes and coastal ground pools.

The genus ranges throughout the American tropics, each species having
a completely Atlantic or Pacific (with a few exceptions) distribution. Curi-
ously, in spite of the occurrence of suitable habitats and hosts ( Cardisoma
and Ucides) along the entire Brazilian coast to Sao Paulo, the most southernly
Atlantic record is on the coast of the State of Maranhao (Cerqueira, 1938:
291). This apparent truncation in the distribution may be due only to lack
of collecting.

All stages exhibit unique characteristics among the Culicidae; some are
definitely functional in their relationship to the crabhole habitat and com-
munity. The peculiar pupal attendance and mating behavior first described
in detail in Deinocerites cancer (Downes, 1966; Provost and Haeger, 1967)
represents an adaptation correlated with a non-dispersing evolutionary trend
in this line of mosquitoes. Among the immatures the larvae of all species
have lateral head pouches of unknown function and the pupae of some have
three float hairs (these and others discussed by Belkin and Hogue, 1959:421).

Only occasional specimens are observed biting man. The normal food
is probably reptile, amphibian or bird blood (Templis and Galindo, 1970).

1972

Burrowing Land Crabs of the World

39

Deinocerites cancer, at least in Florida (Haeger and Phinizee, 1959), is known
to be autogenous.

D. atlanticus Adames, 1971

Crab Host: Not recorded. Known from small crabholes.

D. barretoi Adames, 1971
Crab Host: Not recorded.

D. belkini Adames, 1971
Crab Host: Not recorded, Uca (?)

D. cancer Theobald, 1901
Crab Host: Cardisoma guanhumi.

References: Downes, 1966 (B); Haeger and Phinizee, 1959 (B);
Komp, 1956; Provost and Haeger, 1967 (B).

D. colombianus Adames, 1971
Crab Host: Not recorded.

D. costaricensis Adames and Hogue, 1970
Crab Host: Cardisoma crassum.

D. curiche Adames, 1971
Crab Host: Not recorded.

D. dyari Belkin and Hogue, 1959
Crab Host: Not recorded.

D. epitedeus (Knab, 1907)

Crab Host: Not recorded.

D. howardi Belkin and Hogue, 1959
Crab Host: Not recorded.

D. mathesoni Belkin and Hogue, 1959
Crab Hosts: Uca pugilator, U. subcylindrica, Gecarcinus lateralis.
References: Fisk, 1941 (as spanius) (B); Peyton et al., 1964 (B).

D. magnus (Theobald, 1901)

Crab Host: Not recorded.

D. melanophyllum Dyar and Knab, 1907
Crab Host: Not recorded.

D. mcdonaldi Belkin and Hogue, 1959
Crab Host: Not recorded.

D. nicoyae Adames and Hogue, 1970
Crab Host: Ucides occidentalis.

40

Contributions in Science

No. 220

D. panamensis Adames, 1971
Crab Host: Not recorded.

D. pseudes Dyar and Knab, 1909

Crab Hosts: Cardisoma crassum, C. guanhumi, Uca subcylindrica,
Gecarcinus lateralis.

References: Galindo, 1967 (B); Hogue and Wirth, 1968; Peyton
etal., 1964 (B).

D. spanius (Dyar and Knab, 1909)

Crab Host: Not recorded.

Genus Eretmapodites

E. quinquevittatus Theobald, 1901
Distribution: Tropical Africa; interior and coastal.

Crab Host: Cardisoma armatum.

Type: Accidental. Adults only. Apparently normally breeds in empty
Achatina shells (land snails).

Reference: Wanson, 1935:576.

Genus Galindomyia
G. leei Stone and Barreto, 1969
Distribution: Colombia; coastal.

Crab Host: Not recorded.

Type: Probably specific, though adults only known from crabholes.
Reference: Stone and Barreto, 1969.

Genus Hodgesia
H. nigeriae Edwards, 1930

Distribution: West tropical Africa; interior and coastal.

Crab Host: Cardisoma armatum.

Type: Accidental. Biology not known; probably breeds in vegetated
jungle pools like its relatives.

Reference: Wanson, 1935:576-577, 579.

Genera Mansonia and Coquillettidia
All occurrences of Mansonia and Coquillettidia are resting adults only
and constitute accidental utilization of the crabhole habitat. The larvae of
all members of these genera are associated strictly with floating and emergent
water plants from which they obtain oxygen with a specially modified, piercing
siphon.

C. aurites (Theobald, 1901)

Distribution: Tropical Africa; interior and coastal.

Crab Host: Cardisoma armatum.

Reference: Wanson, 1935:576.

Note: Anthropophilic.

1972

Burrowing Land Crabs of the World

41

M. africana (Theobald, 1901)

Distribution: Tropical Africa; interior and coastal.

Crab Host: Cardisoma armatum.

Reference: Dalziel, 1920:253.

Note: Anthropophilic and vector of yellow fever and filariasis.

M. uniformis (Theobald, 1901)

Distribution: Widespread throughout Old World Tropics; interior and
coastal.

Crab Host: Cardisoma armatum.

Reference: Wanson, 1935:576.

Note: Anthropophilic and important vector of filariasis.

Genus Psorophora

P. confinnis (Lynch Arribalzaga, 1891)

Distribution: Eastern and southern United States, Caribbean, eastern
South America to Argentina; interior and coastal.

Crab Host: Cambarus diogenes ludovicianus.

Type: Transient. Usually breeds in shallow ground pools.

Reference: Evan, 1962.

Note: A complex of species.

Genus Uranotaenia

Nearly all the records in this genus are from the burrows of freshwater
crabs. See Peyton, 1970.

U. alboabdominalis Theobald, 1910
Distribution: Tropical Africa; interior and coastal.

Crab Host: Sudanonautes africanus.

Type: Transient. Adults only. A ground pool breeder.

Reference: Hanney, 1960:99.

U. annulata Theobald, 1901

Distribution: Western tropical Africa; interior and coastal.

Crab Hosts: Cardisoma armatum, Sudanonautes africanus.

Type: Specific or semispecific. Many records and collections, all from
crabholes.

References: Bruce-Chwatt and Fitz-John, 1951:119; Dalziel, 1920:
251, 253 (as fasciata)\ Dunn, 1928:249; Hanney, 1960:99; Hopkins,
1952:59; Macfie and Ingram, 1916:7; Philip, 1931:192; Surtees, 1958:91.

U. atra Theobald, 1905

Distribution: New Guinea, Bismark Archipelago; coastal.

Crab Host: Not recorded.

Type: Transient. Usually breeds in plant and artificial containers and
ground pools and in sago palm swamps.

References: van den Assem, 1961:25; Steffan, 1966:203.

42

Contributions in Science

No. 220

U. bilineata var. fraseri Edwards, 1912
Distribution: Tropical Africa; interior and coastal.

Crab Host: Not recorded.

Type: Accidental? Probably normally breeds in grass-grown ground
pools.

Reference: Dalziel, 1920:251-253.

U. bicolor Leicester, 1908
Distribution: Southeast Asia; interior.

Crab Host: Not recorded.

Type: Transient. Usually breeds in a wide variety of ground pools.
Reference: Peyton, 1970:3 and personal communication.

U. caliginosa Philip, 1931
Distribution: Nigeria; interior.

Crab Host: Not recorded.

Type: ?

Reference: Philip, 1931:190.

U. husaini Qutubuddin, 1946 (1947)

Distribution: India; interior.

Crab Host: Not recorded.

Type: ? Adults only. Bionomics insufficiently known.

Reference: Qutubuddin, 1946 (1947): 118.

U. koli Peyton, 1970

Distribution: Cambodia, Thailand; interior.

Crab Host: Not recorded. Fresh water species.

Type: Specific or semispecific.

Reference: Peyton and Klein, 1970:248.

U. lateralis Leicester, 1908

Distribution: Indomalayan, north Australian and Papuan Regions;
coastal.

Crab Host: Not recorded.

Type: Transient. Usually breeds in slightly brackish open pools behind
beaches, although may be very common in crabholes.

Reference: Leicester, 1908:8, 217 (as cancer ).

U . mashonaensis Theobald, 1901

Distribution: Tropical Africa, Madagascar; interior and coastal.

Crab Host: Sudanonautes africanus.

Type: Accidental. Adults only. Breeds in various ground habitats
(swamps, rockpools, pools, etc.).

Reference: Hanney, 1960:99.

1972

Burrowing Land Crabs of the World

43

U. mattinglyi Qutubuddin, 1951
Distribution: India; interior.

Crab Host: Not recorded.

Type: ? Adults only. Bionomics insufficiently known. “Habitat: caught
from crabholes in an old pond in the Public Garden, Hyderabad (Deccan)
City, India, in October, 1943.”

Reference: Qutubuddin, 1951:107.

U. montana Ingram and de Meillon, 1927
Distribution: Natal, Transvaal, Cape Province; coastal?

Crab Host: Not recorded.

Type: ? Bionomics insufficiently known. A single record from crabholes.
The larva recorded by Hopkins was collected from a crabhole and lived in
captivity for AV2 months before pupating.

Reference: Hopkins, 1952:58.

U. nivipous Theobald, 1912

Distribution: Tropical Africa; coastal? or interior by large rivers?
Crab Host: Not recorded.

Type: ? Bionomics insufficiently known. Two records only from crab-
holes.

References: Ingram and de Meillon, 1927 (as candidipes) ; Surtees,
1958:91 (as candidipes) .

U. philip pinensis Delfinado, 1966
Distribution: Philippines; interior.

Crab Host: Not recorded.

Type: ? Bionomics insufficiently known.

Reference: Peyton, 1970:3 and personal communication.

U. rossi Delfinado, 1966
Distribution: Philippines; interior.

Crab Host: Not recorded.

Type: Specific or semispecific. Numerous collections all from crabholes.
Reference: Peyton, 1970:3 and personal communication.

Family CHAOBORIDAE

The larvae of phantom midges are all aquatic, usually being encountered
in fresh water ponds and lakes. Those of two species in the genus Corethrella
are recorded from crabholes.

Genus Corethrella
C. stonei Lane, 1942
Distribution: Panama; coastal.

Crab Host: Cardisoma sp. ( guanhumi ?)

Type: ? Bionomics insufficiently known. Adults only, taken at mouths
of crabholes.

Reference: Lane, 1942:119.

44

Contributions in Science

No. 220

C. tripunctata Lane, 1942

Distribution: Trinidad, Puerto Rico, Brazil; coastal?

Crab Host: Not recorded.

Type: ? Bionomics insufficiently known. Larvae found in crabholes be-
neath rocks, 100 feet elevation.

Reference: Lane, 1942:120.

Family CERATOPOGONIDAE

A few species of this family, all in the genus Culicoides, are recorded
as breeding in crabholes. Certainly many more of these coastal, salt marsh
and tidal flat-loving flies will be found utilizing the crabhole habitat.

Genus Culicoides

C. arubae Fox and Hoffman, 1944
Distribution: American Tropics; coastal.

Crab Host: Not recorded.

Type: ? Bionomics insufficiently known.

Reference: Fox and Hoffman, 1944:109.

C. cancer Hogue and Wirth, 1968
Distribution: Costa Rica; coastal (Pacific).

Crab Hosts: Cardisoma crassum, Ucides occidentalis.

Type: Specific. The species has been collected in all stages numerous
times and in very large numbers only from crabholes.

Reference: Hogue and Wirth, 1968 (G).

C. cancrisocius Macfie, 1946
Distribution: Fiji Islands; coastal.

Crab Host: Not recorded.

Type: ? Bionomics insufficiently known.

References: Macfie, 1946; Wirth and Arnaud, 1969:517-518.

C. distinctipennis Austen, 1912

Distribution: Nigeria, Uganda, Gold Coast, Senegal; interior and
coastal.

Crab Host: Cardisoma armatum.

Type:? Bionomics insufficiently known.

Reference: Wanson, 1935:579, 584 (as wansoni).

C. insignis Lutz, 1913
Distribution: Mesoamerica; coastal.

Crab Host: Ucides cordatus.

Type: Transient. Usually breeds in coastal marshes.

References: Forattini et al., 1956:197 (B,D); Forattini et al., 1958:
37 (B, D).

1972

Burrowing Land Crabs of the World

45

C. reticulatus Lutz, 1913

Distribution: Brazil, Panama; coastal.

Crab Host: Cardisoma guanhumi.

Type: Specific or semispecific. Numerous collections, all from crabholes.
References: Forattini et al., 1957:312 (B,D); Lutz, 1912:19, 1913:
50 (B).

Family DROSOPHILIDAE

Two species of Drosophila have developed a symbiotic association with
land crabs of the genus Gecarcinus. The larvae live on the crabs and even
pupate (in one species) on the third maxilliped. The adult flies have been
observed to remain on the crab, running over and hovering about the carapace.
Though the food of the larvae is not known with certainty, it probably con-
sists of food leavings of the host in one case and the host’s tissues in the other.

Genus Drosophila
D. carcinophila Wheeler, 1960

Distribution: Greater and Lesser Antilles, Bahamas, Providencia;
islands only (coastal on large islands).

Crab Host: Gecarcinus ruricola.

Type: Commensal, larva in renal grooves and peribuccal region.

References: Carson, 1967 (B,D) ; Wheeler, 1960 (B,D).

D. endobranchia Carson and Wheeler, 1968

Distribution: Cayman Islands; small islands.

Crab Hosts: Gecarcinus ruricola and lateralis.

Type: Parasitic, larva on host’s gills.

Reference: Carson and Wheeler, 1968 (B,D).

MISCELLANEOUS DIPTERA

A number of shore inhabiting flies have been taken, usually as adults
only, from the mouths of crabholes. Others (species largely unidentified) are
known in the immature stages from water in the burrow.

Family CHIRONOMIDAE
Species undetermined.

Localities: Lagos, Nigeria; Banana, Congo.

Crab Host: Cardisoma armatum.

Type: ?

References: Bruce-Chwatt and Fitz-John, 1951:118; Wanson, 1935:

578.

46

Contributions in Science

No. 220

Family CHLOROPIDAE
Genus Lasiopleura sp.

Locality: Not specified.

Crab Host: Not recorded.

Type: ?

Reference: Wheeler, 1960:210.

Family DOLICHOPODIDAE
Asydetus carcinophilus Parent, 1937

Distribution: Hawaii; coastal.

Crab Host: Ocypode ceratophthalma.

Type: Transient. Adults rest and hide in the burrow mouth. Immatures
undoubtedly develop elsewhere.

References: Wheeler, 1960:210; Williams, 1938:126-129.

Family EMPIDIDAE
Chersodromia hawaiiensis Melander, 1938

Distribution: Hawaii; coastal.

Crab Host: Not recorded, probably Ocypode.

Type: Transient. Adults found on the beach in the near vicinity of
burrows.

Reference: Melander, 1938:57.

Family EPHYDRIDAE
Hecamede sp.

Locality: Not specified.

Crab Host: Not recorded.

Type: ?

Reference: Wheeler, 1960:210.

Undetermined “larva A”

Locality: Admiralty Islands.

Crab Host: Cardisoma hirtipes.

Type: Symbiosis? Larvae found in branchial chambers of preserved
crabs.

References: Baylis, 1915; Keilin, 1921.

Undetermined “larva B”

Locality: Christmas Island (Indian Ocean).

Crab Host: Gecarcoidea lalandii (humei).

Type: Symbiosis? Larvae found in branchial chambers of preserved
crabs.

References: Baylis, 1915; Keilin, 1921.

1972

Burrowing Land Crabs of the World

47

Family TETHINIDAE

Rhicnoessa sp.

Locality: Not specified.

Crab Host: Not recorded.

Type: ?

Reference: Wheeler, 1960:210.

Order COLEOPTERA
Family DYTISCIDAE
Bidessus rogersi Young, 1941
Distribution: Florida; interior.

Crab Host: Procambarus rogersi rogersi.

Type: Transient. Primarily a flatwoods species occurring in slow streams,
ditches, cypress swamps, ponds, and other lenitic situations.

Reference: Young, 1954:17, 64-65.

Family HELODIDAE

Helodes ? sp.

Locality: Costa Rica.

Crab Host: Cardisoma crassum.

Type: ? Bionomics insufficiently known. Larva only.

Reference: Hogue and Wirth, 1968:6.

MISCELLANEOUS INSECTS
Order HEMIPTERA
Family GELASTOCORIDAE
Mononyx grandicollis Germar, 1840
Distribution: West Africa; interior and coastal?

Crab Host: Sudanonautes africanus.

Type: Transient or accidental. A streamside mud flat inhabitant.
Reference: Hanney, 1960:100.

Family VELIIDAE
Microvelia or aria Drake, 1952
Distribution: Costa Rica.

Crab Host: Not recorded.

Type: ?

Reference: Drake, 1952:14-15.

Family CYDNIDAE
Sehirus tibialis Stal, 1853
Distribution: West Africa; interior and coastal?

Crab Host: Sudanonautes africanus.

Type: Accidental. Ground burrowing species.

Reference: Hanney, 1960:100 (as Legnotus tibialis).

48

Contributions in Science

No. 220

Order ACARINA

Two mites of different families are known as symbiotic associates of
certain land crabs:

Family TYROGLYPHIDAE
Rhizoglyphus sp.

Distribution: Dry Tortugas Islands; small islands.

Crab Host: Gecarcinus lateralis.

Type: Commensal or parasitic. Single nymphal specimen known.
Reference: Pearse, 1929:230.

Family LAELAPTIDAE
Laelaps cancer Pearse, 1929
Distribution: Dry Tortugas Islands; small islands.

Crab Host: Gecarcinus lateralis.

Type: Commensal or parasitic. All stages found in branchial chambers
and on gills.

References: Pearse, 1929:229-230; 1932:112.

Family UNDETERMINED
Species Undetermined
Locality: West Indies.

Crab Host: Gecarcinus ruricola.

Type: ?

Reference: Carson, 1967:342.

Class CRUSTACEA

Apart from the crab host itself, certain other aquatic Crustacea appear
to be members of the crabhole community.

Order COPEPODA
Cancrincola jamaicensis Wilson, 1913
Distribution: Jamaica, Key West; coastal.

Crab Host: Cardisoma guanhumi.

Type: Parasitic? Cling to gill filaments of crab with their second antennae
and maxillipeds. Probably feed on host’s blood or secretions.

References: Pearse, 1932: 1 12; Wilson, 1913:264-268.

Order EUCOPEPODA
Cyclops sp.

Localities: Lagos, Nigeria; Banana, Congo.

Crab Host: Cardisoma armatum.

Type: ?

References: Bruce-Chwatt and Fitz-John, 1951:118; Wanson, 1935:

578,

1972

Burrowing Land Crabs of the World

49

ACKNOWLEDGMENTS

For direct assistance and contributions to the preparation of this paper
the authors wish especially to thank the following individuals: Drs. John N.
Belkin, University of California, Los Angeles; Jocelyn Crane, New York
Zoological Society; John S. Garth, Allan Hancock Foundation, University
of Southern California; D. J. D. Griffin, Australian Museum; William Macnae,
University of Witwatersrand; E. L. Peyton and Botha de Meillon, Southeast
Asia Mosquito Project; Shivaji Ramalingam, University of Malaya; and
William Stephenson, University of Queensland. We are also indebted to
many institutions and persons who have offered assistance and guidance in
connection with our field trips through Central and South America, East
Africa and Australia, from which were derived the original portions of this
paper. Most significantly these are: the American Philosophical Society;
California State College, Fullerton; National Geographic Society; Los Angeles
County Museum of Natural History Foundation; and Mr. Richard Dwyer,
Dr. David Grajeda, and Dr. Purvis L. Martin.

Literature Cited

Adames, A. J. 1971. Mosquito studies (Diptera, Culicidae) XXIV. A revision of the
crabhole mosquitoes of the genus Deinocerites. Contrib. Amer. Entomol. Inst.
7:1-154.

Aders, W. M. 1917. Insects injurious to man and stock in Zanzibar. Bull. Entomol.
Res. 7:391-401.

Alcock, A. 1900. Materials for a carcinological fauna of India. No. 6. The Brach-
yura Catometopa or Grapsoidea. J. Asiatic Soc. Bengal 69:279-456.
Altevogt, R. 1959. Okologische und ethologische studien an Europas einziger
Winkerkrabbe Uca tangeri Eydoux. Zeit. Morph. Okol. Tiere 48:123-146.

1962. Akustische epiphanomene im sozialverhalten von Uca tangeri in

Sudspanien. Sond. Verh. Deutschen Zool. Ges. Wien 22:309-315.

Amos, D. W. 1944. Mosq. Control. Training Manual. Suva, Fiji. 43 p.

Andrews, C. W. 1900. Land Crustacea, in A monograph of Christmas Island
(Indian Ocean). British Museum Natur. Hist., London XIII, 337 p., 22 pi.
Barnard, K. H. 1947. Descriptions of new species of South African Decapod Crus-
tacea, with notes on synonomy and new records. Ann. Mag. Natur. Hist.,
Ser. 11, 13:361-392.

1950. Descriptive catalogue of South African Decapod Crustacea. Ann.

S. Afr. Mus. 38:1-837.

Barraud, P. J. 1934. Family Culicidae. Tribes Megarhinini and Culicini. The fauna
of British India, including Ceylon and Burma. Diptera 5:1-463.

Baylis, H. A. 1915. A parasitic oligochaete and other inhabitants of the gill-
chambers of land crabs. Ann. Mag. Natur. Hist., Ser. 8, 15:378-381.

Behre, E. H. 1949. Notes on the occurrence of Cardisoma guanhumi Latreille at
Grand Isle, Louisiana. Proc. Louisiana Acad. Sci. 12:19-22.

Belkin, J. N. 1962. The mosquitoes of the South Pacific. Univ. of Calif. Press,
Berkeley. 608 p. -1- 412 pi.

1969. Culex ( Melanoconion ) annulipes invalid. Mosq. Syst. Newsletter

1:68.

, S. J. Heinemann, and W. A. Page. 1970. The Culicidae of Jamaica

(Insecta, Diptera). Contrib. Amer. Entomol. Inst. 6(1): 1-458 (also Inst.
Jamaica, Bull. Sci., Ser. 20).

50

Contributions in Science

No. 220

, and C. L. Hogue. 1959. A review of the crabhole mosquitoes of the genus

Deinocerites (Diptera, Culicidae). Univ. Calif. Pub. Entomol. 14:411:458.

, C. L. Hogue, P. Galindo, T. H. G. Aitken, R. X. Schick, and W. A.

Powder. 1965. Mosquito studies (Diptera, Culicidae) II. Methods for the col-
lection, rearing and preservation of mosquitoes. Contrib. Amer. Entomol. Inst.
1(2): 19-78.

Berlin, O. G. W. 1969. Mosquito studies (Diptera, Culicidae) XII. A revision of
the neotropical subgenus Howardina of Aedes.. Contrib. Amer. Entomol. Inst.
4(2): 1-190.

Bliss, D. D. 1964. Environmental regulation of growth in the Decapod Crustacean
Gecarcinus lateralis. Gen. Comp. Endocrinol. 4:15-41.

Bonne, C., and J. Bonne-Webster. 1925. Mosquitoes of Surinam. A study of
Neotropical mosquitoes. Mededeeling Koloniaal Instituut te Amsterdam 21,
Afdeeling Tropische Hygiene, No. 13. 558 p. + 84 figs.

Borradaile, C. A. 1902. Land Crustaceans. Gardiner’s fauna and geography of
the Maidive and Laccadive Archipelagoes. 1:9-100.

Bott, R. 1954. Dekapoden (Crustacea) aus El Salvador. 1. Winkerkrabben ( Uca ).
Senck. Biol. 35:155-180.

1955. Dekapoden (Crustacea) aus El Salvador. 2. Litorale Dekapoden,

ausser Uca. Senck. Biol. 36:45-72.

Boyce, R. 1909. Mosquito or man? Murray, London xvi, 267 p. + 44 fig.

Bram. R. A. 1967. Contributions to the mosquito fauna of southeast Asia. II. The
genus Culex in Thailand (Diptera: Culicidae). Contr. Amer. Entomol. Inst.
2(1): 1-296.

, and M. Rattanarithikul. 1967. Six new species of the Culex ( Lopho -

ceraomyia ) mammilifer group from Thailand (Diptera: Culicidae). Proc.
Entomol. Soc. Wash. 69:1-17.

Branch, N., and E. L. Seabrook. 1959. Culex (Culex) scimitar, a new species of
mosquito from the Bahama Islands. (Diptera, Culicidae ). Proc. Entomol.
Soc. Wash. 61:216-218.

Bright, D. B. 1966. The land crabs of Costa Rica. Rev. Biol. Trop. 15:183-203.

Brooke Worth, C., J. De Sousa, and M. P. Weinbren. 1961. Studies on the life-
history of Aedes (Skusea) pembaensis (Theobald) (Diptera, Culicidae). Bull.
Entomol. Res. 52:257-261.

Bruce-Chwatt, L. J., and R. A. Fitz-John. 1951. Mosquitoes in crab-burrows on
the coast of West Africa and their control. J. Trop. Med. Hyg. 54:116-121.

Burkenroad, M. 1947. Production of sound by the fiddler crab, Uca pugilator
Bose, with remarks on its nocturnal and mating behavior. Ecology 28:458-462.

Burnett, G. F. 1960. Filariasis research in Fiji, 1957-1959. J. Trop. Med. Hyg.
63:153-162.

Cabrera, J. A. 1965. Contribuciones carcinologicas. I. El primer estadio zoea en
Gecarcinus lateralis (Freminville) (Brachyura Gecarcinidae) procedente de
Veracruz, Mexico. Ann. Inst. Biol. Mex. 36:173-187.

Calman, W. T. 1909. On Decapod Crustacea from Christmas Island, collected by
Dr. C. W. Andrews. Proc. Zool. Soc. London 1909:703-713, 1 pi.

1911. The life of Crustacea. London, Methuen and Co. XVI, 289 p.

Campbell, B. M. 1967. The Australian Sesarminae (Crustacea: Brachyura): five
species of Sesarma ( Chiromantes ). Mem. Queen. Mus. 15:1-19.

Carson, H. L. 1967. The association between Drosophila carcinophila Wheeler and
its host, the land crab Gecarcinus ruricola (L.). Amer. Mid. Natur. 78:324-343.

, and M. R. Wheeler. 1968. Drosophila endobranchia, a new droso-

philid associated with land crabs in the West Indies. Ann. Entomol. Soc. Amer.
61:675-678.

Carter, H. F., and D. P. Wijesundara. 1948. Notes on some Ceylon culicine
mosquitoes. Ceylon J. Sci. 23:135-151.

1972

Burrowing Land Crabs of the World

51

Castro, G. M. de Oliveira. 1932. Estudo sobre uma especie de Culex, que se cria
em buracos de guayamu. (Dipt., Culicidae). Rev. Entomol. 2:97-105.
Cerqueira, N. 1938. Nota sobre o genero Deinocerites : A sua presenca no Estado
do Maranhao, Brasil (Dipt., Culicidae). Rev. Entomol. 8:289-291.

Chace, F. A. 1953. Zoological results of a fifth expedition to East Africa. VI.

Decapod Crustacea. Bull. Mus. Comp. Zool. 110:427-443.

, and H. H. Hobbs. 1969. The freshwater and terrestrial Decapod Crus-
taceans of the West Indies with special reference to Dominica. Bull. U. S. Nat.
Mus. 292:1-258.

, and L. B. Holthuis. 1948. Land and freshwater Decapod Crustacea

from the Leeward Group and Northern South America, in Hummelinck,
Studies on the fauna of Curasao, Aruba, Bonaire and the Venezuelan Islands
3:21-28.

Cheesman, L. E. 1922. Observations on the land crab, Car disoma armatum, with
especial regard to the sense organs. Proc. Zool. Soc. London 1922:361-363.

1923. Notes on the pairing of the land crab, Cardisoma armatum. Proc.

Zool. Soc. London 1923:173.

Cheneveau, R. 1934. Note sur les trous de crabes a Anecho (Togo). Bull. Soc. Pat.
Exot. 27:590-593.

Colless, D. H. 1957. Notes on the culicine mosquitoes of Singapore III.— Larval
breeding-places. Amer. Trop. Med. Parasitol. 51:102-116.

1965. The genus Culex, subgenus Lophoceraomyia, in Malaya (Diptera:

Culicidae). J. Med. Entomol. 2:261-307.

Coluzzi, M. 1964. Morphological divergences in the Anopheles gambiae complex.
Riv. Malariol. 43:197-232.

Cott, H. B. 1930. The zoological society’s expedition to the Zambesi, 1927: No. 2.
Observations on the natural history of the land crab Sesarma meinerti, from
Beira, with special reference to the theory of warning colours. Proc. Zool. Soc.
London 1929:679-692.

Crane, J. 1941a. Eastern Pacific expeditions of the New York Zoological Society
XXVI. Crabs of the genus Uca from the West Coast of Central America.
Zoologica 26:145-207.

1941b. Eastern Pacific expeditions of the New York Zoological Society

XXIX. On the growth and ecology of Brachyuran crabs of the genus Ocypode.
Zoologica 26:297-310.

1943a. Crabs of the genus Uca from Venezuela. Zoologica 28:33-44.

1943b. Display, breeding and relationships of fiddler crabs (Brachyura,

genus Uca ) in the Northeastern United States. Zoologica 28:217-223.

1947. Eastern Pacific expeditions of the New York Zoological Society

XXXVIII. Intertidal Brachygnathous crabs from the west coast of tropical
America with special reference to ecology. Zoologica 32:69-95.

1957. Basic patterns of display in fiddler crabs (Ocypodidae, genus Uca).

Zoologica 42:68-82.

Crosnier, A. 1965. Crustaces Decapodes Grapsidae et Ocypodidae. Faune de
Madagascar 18:1-143.

Dakin, W. J., I. Bennett, and E. C. Pope. 1952. Australian Seashores. Sydney,
Angus and Robertson, xii, 372 p.

Dalziel, J. M. 1920. Crabholes, trees, and other mosquito sources in Lagos. Bull.
Entomol. Res. 11:247-270.

Day, J. H., N. Millard, and G. J. Broekhuysen. 1954. The ecology of South
African estuaries. Part IV. The St. Lucia System. Trans. Roy. Soc S. Afr.
34:129-156.

Delfinado, M D. 1966. The culicine mosquitoes of the Philippines, tribe Culicini
(Diptera, Culicidae). Mem. Amer. Entomol. Inst. 7:1-252.

52

Contributions in Science

No. 220

1967. Contributions to the mosquito fauna of southeast Asia.— I The

genus Aedes, subgenus Neomacleaya Theobald in Thailand. Contrib. Amer.
Entomol. Inst. 1(8): 1-56.

1968. Contributions to the mosquito fauna of southeast Asia.— Ill The

genus Aedes subgenus Neomacleaya Theobald in southeast Asia. Contrib.
Amer. Entomol. Inst. 2(4): 1-74.

Dembowski, J. 1925. On the “speech” of the fiddler crab, Uca pugilator. Trav.
Inst. Nenchi 3:48.

Downes, J. A. 1966. Observations on the mating behavior of the crabhole mos-
quito Deinocerites cancer (Diptera, Culicidae). Can. Entomol. 98:1169-1177.

Drake, D. J. 1952. Two new Microvelia Westwood (Hemiptera; Veliidae). Bull.
Brook. Entomol. Soc. 47:13-15.

Dunn, L. H. 1928. Further observations on mosquito breeding in tree-holes and
crabholes. Bull. Entomol. Res. 18:247-250.

Duret. J. P. 1967. Dos especies nuevas de Culex ( Melanoconion ) (Diptera-Culici-
dae). Neotropica 13:77-84.

Dyar, H. G. 1928. The mosquitoes of the Americas. Carnegie Inst., Washington.

616 p.

, and F. Knab. 1906. The larvae of Culicidae classified as independent

organisms. J. New York Entomol. Soc. 14:169-230.

Edwards, F. W. 1928. The early stages of some Singapore mosquitoes. Bull.
Entomol. Res. 18:337-357.

| _. 1929. Mosquito notes.— VIII. Bull. Entomol. Res. 20:321-343.

Esaki, T. 1940. A preliminary report on the entomological survey of the Micro-
nesian Islands under the Japanese Mandate, with special reference to the insects
of economic importance. Proc. Sixth Pacific Sci. Congr. 4:407-415.

Evan, B. R. 1962. Survey for possible mosquito breeding in crawfish holes in New
Orleans. Mosq. News 22:255-257.

Feliciano, C. 1962. Notes on the biology and importance of the land crab,
Cardisoma guanhumi Latreille of Puerto Rico. Spec. Contrib. Inst. Marine
Biol., Univ. Puerto Rico, iii, 29 p.

Finnegan, S. 1931. Report on the Brachyura collected in Central America, the
Gorgona and Galapagos Islands, by Dr. Crossland on the “St. George” expe-
dition to the Pacific, 1924-25. J. Linn. Soc. London, Zool. 37:607-673.

Fisk, F. W. 1941. Deinocerites spanius at Brownsville, Texas, with notes on its
biology and a description of the larva. Ann. Entomol. Soc. Amer. 34:543-550.

Floch, H., and E. Abonnenc. 1945. Les moustiques de la Guadeloupe (II), les
genres Megarhinus, Aedes, Culex, Deinocerites, Mansonia et Wyeomyia. Pub.
Inst. Pasteur. Guy. Franc. Inini. 1 10:1-48.

Forattini, O. P. 1958. “Culicidae”’ que se criam em buracos de carangueijos
(Diptera). Rev. Brasil Biol. 18:175-179.

Forattini, O. P., E. X. Rabello, and R. L. Heredia. 1956. Notas sobre Culicidae
(Diptera). Arq. Fac. Hig. Saud. Pub. Univ. Sao Paulo 10:85-88.

Forattini, O. P., E. X. Rabello, and D. Pattoli. 1956. Nota sobre a larva e pupa
de Culicoides insignis Lutz, 1913 (Diptera, Ceratopogonidae). Rev. Brasil
Entomol. 4:195-198.

1957. A brief note on breeding places of Culicoides in Sao Vicente,

Brazil. Mosq. News 17:312-313.

1958. Culicoides da regiao neotropical (Diptera, Ceratopogonidae). II—

Observacoes sobre biologia em condicSes naturais. Arq. Fac. Hig. Saud. Pub.
Univ. Sao Paulo 12:1-52.

Forest, J., and D. Guinot. 1961. “Crustaces decapodes brachyoures de Tahiti et
des taumotu.” Expedition francaise sur les recifs coralliens de la Nouvelle-
Caledonie, Paris, xi, 195 p. + 18 pi.

1972

Burrowing Land Crabs of the World

53

Fox, I. and W. A. Hoffman. 1944. New neotropical biting sandflies of the genus
Culicoides (Diptera: Ceratopogonidae). Puerto Rico J. Pub. Health Med.
20:108-111.

Galindo, P. 1967. Preliminary observations on the colonization and bionomics of
the crabhole breeding mosquito Deinocerites pseudes Dyar and Knab, 1909.
Mosq. News 27:187-190.

Garth, J. S. 1948. The brachyura of the “Askoy” expedition with remarks on
carcinological collecting in the Panama Bight. Bull. Amer. Mus. Natur. Hist.
92:1-66.

1960. Distribution and affinities of the brachyuran Crustacea. (Sympo-
sium: The biogeography of Baja California and adjacent seas. Part II. Marine
biotas). Syst. Zool. 9:105-123.

George, R. W., and W. Knott. 1963. The ocypode ghost crabs of Western Aus-
tralia (Crustacea, Brachypura). J. Roy. Soc. W. Aust. 46:15-21.

Gibson-Hill, C. A. 1947. Field notes on the terrestrial crabs of Christmas Island.
Bull. Raffles Mus. 18:43-52.

Gifford, C. A. 1963. Some observations on the general biology of the land crab,
Cardisoma guanhumi (Latreille) in South Florida. Biol. Bull. 123:207-223.

Goiny, E. C. C. et al. 1957. The eggs of Aedes ( Skusea ) pembaensis Theobald dis-
covered on crabs. East Afr. Med. J. 34:1-2.

Gordon, I. 1934. Resultats Scientifiques du Voyage aux Indes Orientales Neer-
landaises, Crustacea brachyura. Mem. Musee Roy. Hist. Natur. Belg. Hors.
Ser. 3, 15:1-78.

1937. Notes on several Indo-Pacific species of Sesarma (Crustacea

brachyura). Proc. Linn. Soc. London 149:150-156.

Gorgas Memorial Laboratory. 1970. Entomology, Blood meals of Deinocerites
mosquitoes, p. 19-20. in Forty-first annual report of the work and operations
of the Gorgas Memorial Laboratory, fiscal year 1969. Gorgas Memorial
Laboratory, Panama City, Panama. 35 p.

Grabham, M. 1905. Notes on some Jamaican Culcidae. Can. Entomol. 37:401-411.

Grayson, M. A., S. Srihongse, and P. Galindo. 1967. Isolation of St. Louis
encephalitis virus from Deinocerites pseudes in Panama. Mosq. News 27:204.

Green, J. P. 1964. Morphological color change in the Hawaiian ghost crab, Ocy-
pode ceratophthalma (Pallas). Biol. Bull. 126:407-413.

Gressitt, J. L. 1954. Introduction. Insects of Micronesia. 1:1-257.

Haeger, J. S., and J. Phinizee. 1959. The biology of the crabhole mosquito Deino-
cerites cancer. (Typed report of Florida Antimosquito Assoc.).

Hanney, P. W. 1960. Some notes on the insect fauna of crabholes in northern
Nigeria. Entomologist 93:99-100.

Hediger, H. 1933. Beobachtungen an der marokkanischen winkerkrabbe, Uca
tangeri (Eydoux). Verh. Schweiz. Naturforsch. Ges. 114:388-389.

u 1934. Zur biologie und psychologie der flucht bei tieren. Biol. Zentralbl.

Leipzig. 54:21-40.

Heisch, R. B., H. H. Goiny, and M. Ikata. 1956. A new vector of filariasis in East
Africa. Trans. Roy. Soc. Trop. Hyg. 50:421-422.

Herreid, C. F. 1963. Observations on the feeding behavior of Cardisoma guanhumi
(Latreille) in Southern Florida. Crustaceana 5:176-180.

1967. Skeletal measurements and growth of the land crab Cardisoma

guanhumi Latreille. Crustaceana 13:39-44.

, and C. A. Gifford. 1963. The burrow habitat of the land crab, Cardi-
soma guanhumi (Latreille). Ecology 44:773-775.

Hill, R. B., and C. D. 1948. The mosquitoes of Jamaica. Bull. Inst. Jamaica, Sci.
Ser., No. 4. 60 p.

Hogue, C. L., and D. B. Bright. 1969. Study of the biologies of land crabs and
their burrow associates. Assoc. Trop. Biol. Newsletter 18:9-11.

54

Contributions in Science

No. 220

1971. Observations on the biology of land crabs and their burrow asso-
ciates on the Kenya coast. Los Angeles Co. Mus., Contrib. Sci. 210.

Hogue, C. L. and W. W. Wirth. 1968. A new Central American sand fly breeding
in crabholes (Diptera, Ceratopogonidae). Los Angeles Co. Mus., Contrib. Sci.
152.

Holthuis, L. B. 1953. Enumeration of the decapod and stomatopod Crustacea
from Pacific Coral Islands. Atoll Res. Bull. 24:1-66.

1959. The Crustacea decapoda of Suriname (Dutch Guiana). Zool. Verh.

Leiden 44: 1-296.

Hopkins, G. H. E. 1952. Mosquitoes of the Ethiopian region I. — Larval bionomics
of mosquitoes and taxonomy of culicine larvae, 2nd. ed. Trustees Brit. Mus.
(Natur. Hist.), London. 355 p.

Howard, L. O., H. G. Dyar and F. Knab. 1915. The mosquitoes of North and
Central America and the West Indies. 3:1-523.

Hubbard, W. G. 1894. Notes and exhibitions of specimens. Proc. Entomol. Soc.
Wash. 3:184.

Ingram, A., and J. W. S. Macfie. 1917. The early stages of certain west African
mosquitoes. Bull. Entomol. Res. 8:135-154.

Ingram, A., and B. de Meillon. 1927. A mosquito survey of certain parts of
South Africa, with special reference to the carriers of malaria and their con-
trol. (part 1). Pub. So. Africa Inst. Med. Res. 4:1-81, 15 pi.

Keilin, D. 1921. On some dipterous larvae infesting the brachial chambers of land
crabs. Ann. Mag. Natur. Hist., Ser. 9, 8:601-608.

King, W. V., and H. Hoogstraal. 1946a. Three new species of Aedes from Nether-
lands New Guinea. Proc. Entomol. Soc. Wash. 48:95-106.

1946b. The New Guinea species of Culex ( Culiciomyia ), with descrip-
tions of two new species. Proc. Biol. Soc. Wash. 59:143-154.

1947. New Guinea species of mosquitoes of the genus Aedes, subgenus

Aedes. J. Wash Acad. Sci. 37:113-134.

Knab, F. 1910. Mosquito habits and mosquito control. Science 31:868-869.

Knight, K. L., R. M. Bohart, and G. E. Bohart. 1944. Keys to the mosquitoes
of the Australasian region, including a synopsis of their distribution and
breeding habits. Nat. Res. Conn., Div. Med. Sci., Wash. 71 p.

Knight, K. L., and W. B. Hull. 1953. The Aedes mosquitoes of the Philippine
Islands. III. Subgenera Aedimorphus, Banksinella, Aedes, and Cancraedes
(Diptera, Culicidae). Pacif. Sci. 7:453-481.

Komp, W. H. W. 1956. Copulation in crabhole mosquitoes (Diptera, Culicidae).
Proc. Entomol. Soc. Wash. 58:349-351.

Kumm, H. W. 1931. Studies on Aedes larvae in southwestern Nigeria and in the
vicinity of Kano. Bull. Entomol. Res. 22:65-74.

Lane, J. 1942. Dixinae e Chaoborinae. Revisao das especies neotropicas (Diptera,
Culicidae). Rev. Entomol. 13:81-148, 4 pi.

Leicester, G. F. 1908. The Culicidae of Malaya. Stud. Inst. Med. Res. F. M. S.
3(3): 18-261.

Lien, J. C. 1968. New species of mosquitoes from Taiwan (Diptera: Culicidae).
Part III. Five new species of Aedes. Trop. Med. 10:95-115.

Lumsden, N. H. R. 1955. Entomological studies, relating to yellow fever epi-
demiology, at Gede and Taveta, Kenya. Bull. Entomol. Res. 446:149-183.

Lutz, A. 1912. Contribuicao para o estudo ceratopogoninas hematofagas do Brasil.
Parte geral primeira memoria. Mem. Inst. Oswaldo Cruz 4:1-33.

1913. Contribuicao para o estudo ceratopogoninas hematofagas do

Brasil. Parte sistematica segunda memoria. Mem. Inst. Oswaldo Cruz 5:45-73,
3 pi.

1972

Burrowing Land Crabs of the World

55

MacDonald, W. W. 1957. Malaysian parasites — XVI An interim review of the
nonanopheline mosquitoes of Malaya. Stud. Inst. Med. Res. Malaya No. 28.
34 p.

Macfie, J. W. S. 1946. A new species of Culicoides (Diptera, Ceratopogonidae)
from Fiji. Proc. Roy. Entomol. Soc. London, Ser. B, 15:15-16.

, and A. Ingram. 1916. New culicine larvae from the Gold Coast. Bull.

Entomol. Res. 7:1-18.

Macnae, W. 1963. Mangrove swamps in South Africa. J. Ecol. 51:1-25.

1966. Mangroves in Eastern and Southern Australia. Aust. J. Bot.

14:67-104.

, and M. Kalk (Eds.). 1969. A natural history of Inhaca Island,

Mozambique. Witwatersrand Univ. Press, Johannesburg, vi, 163 p.

Man, J. G. de. 1879. On some new or imperfectly known Podophthalmous Crus-
tacea of the Leyden Museum. Notes Leyden Mus. 1:1-61.

Manning, R. B. and A. J. Provenzano. 1961. The occurrence of U tides cordatus
(Linnaeus, 1763) (Decapoda) in the United States. Crustaceana 2:158-159.
Martini, E. 1914. Some new American mosquitoes. Insecutor Inscitiae Men-
struus 2:65-76.

Mattingly, P. F. 1958. The culicine mosquitoes of the Indomalayan area.
Part III. Genus Aedes Meigen, subgenera Paraedes Edwards, Rhinoskusea
Edwards and Cancraedes Edwards. Trustees Brit. Mus. (Natur. Hist.), Lon-
don. 61 p.

1959. The culicine mosquitoes of the Indomalayan area. Part IV.

Genus Aedes Meigen subgenera Skusea Theobald, Diceromyia Theobald,
Geoskusea Edwards, and Christophersiomyia Barraud. Trustees Brit. Mus.
(Natur. Hist.), London. 61 p.

Melander, A. L. 1938. A new empidid fly in Hawaii. Proc. Hawaiian Entomol.
Soc. 10:57.

Miers, E. J. 1880. On a collection of Crustacea from the Malaysian region. Ann.

Mag. Natur. Hist., Ser. 5, 5:226-239, 304-317.

Millard, N. A., and A. D. Harrison. 1954. The ecology of South African estuaries.

Part V. Richard’s Bay. Trans. Roy. Soc. S. Afr. 34:157-179.

Milne Edwards, A. 1879. Memoire sur les Crustaces decapodes du genre Dyno-
mene. Ann. Sci. Nat. (VI), 8:1-11.

Milne Edwards, H. 1834. Histoire naturelle des Crustaces, Paris, 1:1-360.

1837. Historic naturelle des Crustaces, Paris, 2:1-531.

Miyake, S. 1939. Notes on Crustacea brachyura collected by Professor Teiso
Esaki’s Micronesia expeditions 1937-1938, together with a checklist of Mic-
ronesian brachyura. Rec. Oceanogr. Works, Japan, 10:168-247.
Montchadsky, A. S., and I. Garcia A. 1966. Las larvas de los mosquitos (Diptera:
Culicidae) de Cuba. Su biologia y determinacion. Poeyana (A) 28:1-93.
Murphy, R. C. 1944. Mountain and sea in the Choco (the sixth installment in the
story of the “Askoy” expedition). Natur. Hist. 53:474-481.
de Oliveira, L. P. H. 1946. Estudos ecologicos dos Crustaceos comestiveis Uga
e Guaiamu, Cardisoma guanhumi Latreille e Ucides cordatus (L.). Mem.
Inst. Oswaldo Cruz 44:295-322.

Ortmann, A. E. 1894. Die Decapoden-Krebse des Strassburger Museum, 8. Zool.
Jahrb., Syst. 7:683-772.

Pearse, A. S. 1914a. On the habits of Uca pugnax (Smith) and U. pugilator (Bose).
Trans. Wisconsin Acad. Sci. 17:791-802.

1914b. Habits of fiddler crabs. 415-427. in Ann. Rep. Smith. Inst.

for 1913.

1916. An account of the Crustacea collected by the Walker expedi-
tion to Santa Marta, Colombia. Proc. U.S. Nat. Mus. 49:531-556.

56

Contributions in Science

No. 220

1929. Two new mites from the gills of land crabs. Pap. Tortugas

Lab. Carnegie Inst. Wash. 26:227-230.

1932. Observations on the parasites and commensals found associ-
ated with crustaceans and fishes at Dry Tortugas, Florida. Pap. Tortugas Lab.
Carnegie Inst. Wash. 28:107-115.

Pesta, M. 1931. Ergebnisse der Osterreichischen Biologischen Costa-Rica Expedi-
tion 1930. I. Teil. Crustacea decapoda aus Costa-Rica. Ann. Naturhist. Mus.
Wien 45:173-181.

Peters, W. 1963. Mosquitoes of New Guinea (Diptera: Culicidae) IV. — Two new
species, genera Aedes Meigen and Armigeres Theobald. Proc. Roy. Ent. Soc.
Lond., Ser. B, 32:1-11.

Peyton, E. L. 1970. Studies on Uranotaenia at SEAMP A plea for further material.
Mosq. Syst. Newsletter 2:2-5.

Peyton, E. L., J. F. Reinert, and N. E. Peterson. 1964. The occurrence of
Deinocerites pseudes Dyar and Knab in the United States, with additional
notes on the biology of Deinocerites species of Texas. Mosq. News 24:449-458.

Peyton, E. L., and T. M. Klein. 1970. Five new species of Uranotaenia from
southeast Asia. Proc. Ent. Soc. Wash. 72:243-251.

Phillip, C. B. 1931. Two new species of Uranotaenia (Culicidae) from Nigeria,
with notes on the genus in the Ethiopian Region. Bull. Entomol. Res.
22:183-193.

Pocock, R. I. 1888. On the Arachnida, Myriopoda and land-Crustacea of Christ-
mas Island. Proc. Zool. Soc. London 1888:556-560.

Pratt, H. D., and E. L. Seabrook. 1952. The occurrence of Culex iolambdis Dyar
in Florida and Puerto Rico, with a description of the larva (Diptera, Culicidae).
Proc. Entomol. Soc. Wash. 54:27-32.

Pratt, H. D., W. W. Wirth, and D. G. Denning. 1945. The occurrence of Culex
opisthopus Komp in Puerto Rico and Florida, with a description of the larva
(Diptera, Culicidae). Proc. Entomol. Soc. Wash. 47:245-251.

Provost, M. W., and J. S. Haeger. 1967. Mating and pupal attendance in Deino-
ce.ites cancer and comparisons with Opifex fuscus (Diptera: Culicidae). Ann.
Entomol. Soc. Amer. 60:565-574.

Qutubuddin, M. 1946(1947). A new species of Uranotaenia, Culicidae-Diptera
from Hyderabad (Deccan). Indian J. Entomol. 8:117-118.

1951. Uranotaenia mattinglyi sp. n., a new species of mosquito (Diptera,

Culicidae) from Hyderabad (Deccan), India. Proc. Roy. Entomol. Soc. Lon-
don, Ser. B, 20:107-109.

Rathbun, J. J. 1914. New species of crabs of the families Grapsidae and Ocypo-
didae. Proc. U. S. Nat. Mus. 47:69-85.

1918. The Grapsoid crabs of America. Spec. Bull. U.S. Nat. Mus.

97:1-461.

1921. The brachyuran crabs collected by the American Museum

Congo expedition, 1909-1915. Bull. Amer. Mus. Natur. Hist. 43:379-474.

1933. Brachyuran crabs of Puerto Rico and the Virgin Islands. In

Scientific survey of Puerto Rico and the Virgin Islands, (New York Academy
of Sciences). 15:1-121.

Ray, C. 1967. Gecarcinus lateralis Freminville in Texas. Texas J. Sci. 19:109.

Riqueau. 1929. Les trous de crabes, gites a larves. Bull. Soc. Pathol. Exot. 22:175-
179.

Sakai, T. 1940. Bio-geographic review on the distribution of crabs in Japanese
waters. Rec. Oceanogr. Works, Japan 11:27-63.

Salmon M. 1965. Waving display and sound production in the courtship behavior
o' Uca pugilator, with comparisons to U. minax and U. pugnax. Zoologica
' 0:123-149.

1972

Burrowing Land Crabs of the World

57

, and J. F. Stout. 1962. Sexual discrimination and sound production

in Uca pugilator Bose. Zoologica 47:15-20.

Scanlon, J. E., and E. L. Peyton. 1967. Anopheles ( Anopheles ) tigertti, a new
species of the Aitkenii group from Thailand. Proc. Entomol. Soc. Wash.
69:18-23.

Scharff, J. W., and M. W. F. Tweedie. 1942. Malaria and the mud lobster. Trans.
Roy. Soc. Trop. Med. Hyg. 36:41-44.

Schmitt, W. L. 1921. The marine decapod Crustacea of California. Univ. Calif.
Publ. Zool. 23:1-470.

Shropshire, J. B., and J. Zetek. 1927. Unusual Anopheles habitats in the Canal
Zone. Amer. J. Trop. Med. 7:331-338.

Silas, E. G., and C. Sankarankutty. 1960. On the castle building habit of the
crab Car disoma carnifex (Herbst) (Family Gecarcinidae) of the Andaman
Islands. J. Mar. Biol. Assoc. India, 1960. 2:237-240.

Sloof, R., and E. N. Marks. 1965. Mosquitoes (Culicidae) biting a fish (Perio-
phthalmidae). J. Med. Entomol. 2:16.

Stebbing, T. R. 1910. South African Crustacea (Part IV). Ann. S. Afr. Mus. 6: 1-96.

Steffan, W. A. 1966. A checklist and review of the mosquitoes of the Papuan
subregion (Diptera: Culicidae). J. Med. Entomol. 3:179-237.

Stone, A. 1950. The larva of Culex foliaceus Lane (Diptera, Culicidae). Rev.
Entomol. 21:237-239.

, and J. A. Hair. 1968. A new Culex ( Melanoconion ) from Florida

(Diptera, Culicidae). Mosq. News 28:39-41.

, and P. Barreto. 1969. A new genus and species of mosquito from

Colombia, Galindomyia leei (Diptera, Culicidae, Culicini). J. Med. Entomol.
6:143-146.

Surtees, G. 1958. Notes on the breeding habits of some culicine mosquitoes
(Diptera: Culicidae) in southern Ghana. Proc. Roy. Entomol. Soc. London,
Ser. A, 33:88-92.

Symes, C. L. 1960. Observations on the epidemiology of filariasis in Fiji. J. Trop.
Med. Hyg. 63:1-44.

Tamashiro, M. 1964. Observations on the larval ecology of Aedes ( Stegomyia )
polynesiensis Marks on Aitutaki, southern Cook Islands. WHO/EBL/22: 1-29,
7 fig. (Mimeographed).

Teal, J. M. 1958. Distribution of fiddler crabs in Georgia salt marshes. Ecology
39:185-193.

Templis, C. H., and P. Galindo. 1970. Feeding habits of five species of Deino-
cerites mosquitoes collected in Panama. J. Med. Entomol. 7:175-179.

Tesch, J. J. 1917. Synopsis of the genera Sesarma, Metasesarma, Sarmatium and
Clistocoeloma, with a key to the determination of the Indo-Pacific species.
Zool. Meded. Rijks. Mus. Natur. Hist. Leyden 3:127-260.

1918. The decapoda brachyura of the Siboga expedition. I. Hymeno-

somidae, Retroplumidae, Ocypodidae, Grapsidae and Gecarcinidae. Siboga
Exped. Monogr. 39c, (livre 82): 1-148.

Tweedie, M. W. F. 1940. New and interesting Malaysian species of Sesarma and
Utica (Crustacea, Brachyura). Bui.. Raffles Mus. 16:88-113.

1947. On the brachyura of Christmas Island. Bull. Raffles Mus.

18:27-42.

1950. Papers on the fauna of the Cocos-Keeling Islands. Bull. Raffles

Mus. 22:105-148.

1954. Notes on Grapsoid crabs from the Raffles Museum, Nos. 3,

4 and 5. Bull. Raffles Mus. 25:118-123.

van den Assem, J. 1961. Mosquitoes collected in the Hollandia area, Netherlands
New Guinea, with notes on the ecology of larvae. Tijd. Entomol. 104:17-30.

58

Contributions in Science

No. 220

van Someren, E. C. C., C. Teesdale, and M. Furlong. 1955. The mosquitoes
of the Kenya Coast; Records of occurrence, behavior and habitat. Bull.
Entomol. Res. 46:463-493.

Villalobos, A., and J. Cabrera. 1964. Los cangrejos del genero Gecarcinus como
un peligro potencial para la agricultura de las zonas aledanas a nuestros
litorales. Folia Entomol. Mexicana 7-8:27-29.
von Hagen, H. O. 1961. Nachtliche aktivitat von Uca tangeri in Sudspanien.
Naturwissenschaften 48:1-140.

1962. Frielandstudien zur Sexual-und fortplanzungsbiologie von Uca

tangeri in Andulusien. Zeit. Morph. Okol. Tiere 51:611-725.

1968. Studien an peruanischen winkerkrabben {Uca). Zool. Jaarb.

Syst. 95:395-468.

Wanson, M. 1935. Note sur les trous de crabes, gites larvaires. Ann. Soc. Belg.
Med. Trop. 15:575-585.

Ward, M. 1934. Notes on a collection of crabs from Christmas Island, Indian
Ocean. Bull. Raffles Mus. 9:5-26.

Webb, G. E. 1922. Note on some young stages of Gecarcoidec lalandii Milne-
Edwards. Ann. Mag. Natur. Hist. 10:530-534.

Wheeler, M. R. 1960. A new genus and two new species of neotropical flies
(Diptera: Drosophilidae). Entomol. News 71:207-213.

Williams, F. X. 1938. Asyndetus carcinophilus Parent (Diptera, Dolichopodidae).

Proc. Hawaiian Entomol. Soc. 10: 126-129.

Wilson, C. B. 1913. Crustacean parasites of West Indian fishes and land crabs,
with descriptions of new species. Proc. U. S. Nat. Mus. 44:189-277.

Wirth. W. W., and P. H. Arnaud. 1969. Polynesian biting midges of the genus
Culicoides (Diptera: Ceratopogonidae). Pacific Insects 11:507-520.
Wood-Mason, J. 1873. On a new genus and species of land crabs from the Nico-
bar Islands. J. Asiat. Soc. Bengal 42:258-262.

Wright, H. O. 1966. Comparative studies of social behavior in Grapsoid crabs.

Ph.D. Dissertation, Univ. California, Berkeley. 226 p.

Young, F. N. 1954. The water beetles of Florida. Univ. Florida Stud., Biol. Sci.
Ser. 5(1): 1-238.

Accepted for publication September 10, 1970

Printed in Los Angeles, California by Continental Graphics

CzLgtf

NUMBER 221
FEBRUARY 16, 1972

FOSSIL MICROTINES FROM LATE
CENOZOIC DEPOSITS IN THE
ANZA-BORREGO DESERT, CALIFORNIA,
WITH THE DESCRIPTION OF A NEW
SUBGENUS OF SYNAPTOMYS

By Richard J. Zakrzewski

CONTRIBUTIONS IN SCICNCC

NATURAL HISTORY MUSEUM

LOS ANGELES COUNTY

CONTRIBUTIONS IN SCIENCE is a series of miscellaneous technical papers
in the fields of Biology, Geology and Anthropology, published at irregular intervals
by the Natural History Museum of Los Angeles County. Issues are numbered sep-
arately, and numbers run consecutively regardless of subject matter. Number 1 was
issued January 23, 1957. The series is available to scientific institutions and scien-
tists on an exchange basis. Copies may also be purchased at a nominal price. Inquiries
should be directed to Virginia D. Miller, Natural History Museum of Los Angeles
County, 900 Exposition Boulevard, Los Angeles, California 90007.

INSTRUCTIONS FOR AUTHORS

Manuscripts for CONTRIBUTIONS IN SCIENCE may be in any field of
Life or Earth Sciences. Acceptance of papers will be determined by the amount
and character of new information. Although priority will be given to manuscripts
by staff members, or to papers dealing largely with specimens in the collections of
the Museum, other technical papers will be considered. All manuscripts must be
recommended for consideration by the curator in charge of the proper section or
by the editorial board. Manuscripts must conform to those specifications listed
below and will be examined for suitability by an Editorial Committee including
review by competent specialists outside the Museum.

Authors proposing new taxa in a CONTRIBUTIONS IN SCIENCE must
indicate that the primary type has become the property of a scientific institution of
their choice and cited by name.

MANUSCRIPT FORM.— ( 1 ) The 1964 AIBS Style Manual for Biological
Journals is to be followed in preparation of copy. (2) Double space entire manu-
script. (3) Footnotes should be avoided if possible. Acknowledgments as footnotes
will not be accepted. (4) Place all tables on separate pages. (5) Figure legends and
unavoidable footnotes must be typed on separate sheets. Several of one kind may
be placed on a sheet. (6) An abstract must be included for all papers. This will be
published at the head of each paper. (7) A Spanish summary is required for all
manuscripts dealing with Latin American subjects. Summaries in other languages
are not required but are strongly recommended. Summaries will be published at
the end of the paper. (8) A diagnosis must accompany any newly proposed taxon.
(9) Submit two copies of manuscript.

ILLUSTRATIONS. — All illustrations, including maps and photographs, will
be referred to as figures. All illustrations should be of sufficient clarity and in the
proper proportions for reduction to CONTRIBUTIONS page size. Consult the 1964
AIBS Style Manual for Biological Journals in preparing illustration and legend copy
for style. Submit only illustrations made with permanent ink and glossy photo-
graphic prints of good contrast. Original illustrations and art work will be returned
after the manuscript has been published.

PROOF.— Authors will be sent galley proof which should be corrected and
returned promptly. Changes in the manuscript after galley proof will be billed to
the author. Unless otherwise requested, page proof also will be sent to the author.
One hundred copies of each paper will be given free to each author or divided
equally among multiple authors. Orders for additional copies must be sent to the
Editor at the time corrected galley proof is returned. Appropriate order forms will
be included with the galley proof.

Virginia D. Miller
Editor

FOSSIL MICROTINES FROM LATE CENOZOIC DEPOSITS
IN THE ANZA-BORREGO DESERT, CALIFORNIA,

WITH THE DESCRIPTION OF A NEW SUBGENUS
OF SYNAPTOMYS

By Richard J. Zakrzewski1

Abstract: Remains of fossil microtines from the Late
Cenozoic deposits of the Anza-Borrego desert are extremely
rare. Microtines are found at only four out of 450 sites. The
specimens found represent three genera: Synaptomys, Microtus,
and an indeterminate one. The remains of Synaptomys when
considered with similar forms described previously from Idaho
and Kansas permit the naming of a new subgenus and species.

The presence of the microtines adds credence to the suggestion of
a more equable climate than at present in the area. The stage-of-
evolution of the microtines confirms a Late Blancan to Irving-
tonian time span for the deposits.

Introduction

Fossil-bearing deposits in the Anza-Borrego Desert of southern California
have been recently described by Downs and White (1968). These deposits
range in age from ?mid-Pliocene (Hemphillian) to mid-Pleistocene (Irving-
tonian) and the fossils are arbitrarily placed into three local faunas. Over
ninety vertebrate taxa have been reported, among which are three genera of
microtine rodents.

The microtines are conspicuous in the deposits because of their extreme
rarity. From 450 sites through 12,000 feet of beds, microtines are found at
only four localities. This paucity may be ascribed to one or both of the follow-
ing causes: 1. the southern location of the area; microtines are primarily
boreal and temperate forms, and 2. the presence of the cricetine genera
Neotoma and Sigmodon, which conservatively outnumber microtines in the
deposits of this area some 100 to one and parallel them in dietary preferences.
For example, Baker (1969) points out that S. hispidus is replacing Pedomys
ochrogaster in the Recent fauna of the south-central states as the primary
grass-eating rodent.

Unfortunately, not all the localities from which the microtines were ob-
tained are in the type section, but rather from areas which have been subjected
to an unknown amount of faulting. Consequently, exact stratigraphic relation-
ships are difficult to determine. A detailed study of the geology of this region
is presently underway. An attempt at correlation is based on the stage-of-
evolution of the microtines. The nomenclature of the teeth follows Zakrzewski
(1967).

department of Earth Sciences, and Sternberg Memorial Museum, Fort Hays
Kansas State College, Hays, Kansas 67601.

1

2

Contributions in Science

No. 221

Systematic Paleontology
Synaptomys Baird 1857
Metaxyomys, subgen. nov.

Holotype — Synaptomys vetus Wilson, 1933, Carnegie Inst. Wash. Publ.
440: 124-126, fig. 3. Specimen 1364 C.I.T. now in LACM.

Referred species:

Synaptomys landesi Hibbard, 1954, Jour. Mammal. 35, (n. 2): 249-252,
Fig. 1C.

Synaptomys anzaensis, n. sp.

Subgeneric diagnosis. —The subgenus Metaxyomys is distinguished by its
Mx (Fig. 1A, B, I), in which the first and second alternating triangles are con-
fluent as in the subgenus Mictomys (Fig. 1C), while the external re-entrant
angles are well developed as in the subgenus Synaptomys (Fig. ID).

Discussion and additional description.— The, decision to place the species
listed above into a new subgenus was based on the fact that these forms
possess a number of characters which are intermediate between those found
in the two extant subgenera, Synaptomys and Mictomys. Likewise, the sub-
genus, Metaxyomys combines characters which are now found independently
in the two extant subgenera, as shown by the morphology of the Mx.

Cement may be present or absent in the posterior-external re-entrant
angle of the Mx in Metaxyomys. In Mictomys cement is absent in all the
external re-entrants of the Mx. In Synaptomys it is present. The anterior loop
of the Mj is more vole-like in Metaxyomys than it is in either of the other two
subgenera.

The third and fourth alternating triangles in the M2 and M3 of Metaxy-
omys resemble Synaptomys while the first and second alternating triangles
are confluent as in Mictomys.

The upper molar teeth of Metaxyomys tend to more closely resemble
those of Mictomys because of the slight development of a third internal re-
entrant which helps to better isolate the most posterior triangle or loop (Fig.
1 E-H ) . However, even in this character, some Metaxyomys specimens show
an intermediate development between that of the two extant subgenera.

The capsular process for the reception of the lower incisor is located at
the very anterior end of the M3 in Metaxyomys. In Synaptomys, although a
partial north-south cline has been demonstrated (Hibbard, 1963), the incisor
terminates more posteriorly in every case and usually behind the M3. In Recent
specimens of Mictomys the incisor ends at the posterior portion of the M2.
One fossil species of Mictomys, M. kansasensis, has the lower incisor termi-
nate behind the M3 (Hibbard, 1952). The grooves on the upper incisors are
apparently also intermediate in position, but approach more closely the posi-
tion observed in Synaptomys (Wilson, 1933).

Etymology.— The subgenus Metaxyomys takes its name from the Greek
metaxy = between, in allusion to its intermediate nature and mys for mouse.

Synaptomys ( Metaxyomys ) vetus Wilson
(Figures 1A, E)

Synaptomys vetus Wilson, 1933, Carnegie Inst. Wash. Publ. 440: 124-126,
Figs. 2-3.

1972

Fossil Microtines From Anza-Borrego Desert

3

Holotype — Same as type of subgenus.

Horizon and type locality — Upper part of the Glenns Ferry Formation,
early Pleistocene, Grand View local fauna, Owyhee County, Idaho.

Emended diagnosis— Synaptomys vetus is distinguished by its Mx, on

Figure 1. Occlusal views of microtine teeth. A-I, Synaptomys, A-D, RMiS, E-H,
RM3s, A, S. ( Metaxomys ) vetus, UMMP V59973; B, S. (M.) landesi, UMMP
V59982; C, S. ( Mictomys ) borealis, KU 43256; D, 5. (S.) cooped, KU 5041; E,
S. ( Metaxomys ) vetus, UMMP V56322; F, S. (M.) landesi, UMMP V59972; G,
S. ( Mictomys ) borealis, KU 43256; H, S. (S.) cooped, KU 5041; I, S. ( Metaxomys )
anzaensis, holotype, LACM 19684; J, Microtine, gen. et sp. indet., LACM 24647,
LM3; K, Microtus calif ornicus? , LACM 24540, RMi.

4

Contributions in Science

No. 221

which the anterior loop generally possesses a deep, angular, internal re-entrant
and a shallow, external re-entrant. The second alternating triangle of the Mx
is generally rounded in appearance. The M3 has a posterior loop which is
ellipsoidal in shape and lacks enamel along the entire posterior face.

Additional description .—Synaptomys vetus was chosen as the type of the
subgenus because it is the best known member of the group. The Mx is char-
acterized by the first and second alternating triangles being confluent and
having well-developed external re-entrant angles (Fig. 1A). The anterior loop
appears to be generally smaller than those of other members of the subgenus.
There is also a tendency for the internal re-entrant on the anterior loop of
S. vetus to be more angular than in the other species and on occasion (5/22)
this re-entrant is filled with cement. S. vetus is the only member of the sub-
genus in which this character has been found thus far. Cement is present in
this re-entrant in the subgenus Mictomys but is lacking in the subgenus
Synaptomys. The external re-entrant on the anterior loop of S. vetus is fairly
shallow.

Generally the second alternating triangle appears rounded in outline
(Fig. 1A) because of the flatness of the external surface due to the absence of
enamel. On occasion (6/22), generally in ontogenetically younger specimens,
the second alternating triangle is more triangular in nature. The triangularity
of the second alternating triangle of Mx is a character which predominates in
Synaptomys landesi (Fig. IB). Characters of the other lower molars will be
considered under the discussion of S. landesi.

In the lower jaw of Synaptomys vetus and S. (S.) coo peri there is a ridge
present between the tooth row and the ascending ramus. In S. vetus, a very
shallow temporal fossa is found, exterior to this ridge on the ascending ramus,
in a position approximate with the alveolar boundary between M2 and M3. An
even less pronounced fossa occurs in S. cooperi in approximately the same
position, while in S. ( Mictomys ) borealis the area between the tooth row and
ascending ramus descends gradually to a deep point behind M3. There is no
obvious fossa or possibly the strengthening of the musculature, if any, occurs
in the area behind the M3. The condition which exists in the other species of
the subgenus Metaxyomys is unknown. The ascending ramus of S. vetus
thickens in the post-fossa area. The diastemal region appears to be more ro-
bust in the subgenus Metaxyomys than in either of the extant subgenera. The
lower incisor of S. vetus terminates at the anterior edge of the M3.

As mentioned above the posterior loop of the M3 is very diagnostic in
Synaptomys vetus. The loop is ellipsoidal in shape and enamel is absent from
the entire posterior face (Fig. IE). The first alternating and third alternating
triangles open broadly into the second which is poorly developed to give a
C-shaped pattern to the alternating triangles of the M3. Characteristics of the
other upper molars will be considered under the discussion of S. landesi.

Synaptomys ( Metaxyomys ) landesi Hibbard
(Figures IB, F)

Synaptomys ( Mictomys ) cf. S. (M.) vetus Wilson. Hibbard, 1941, Kansas
Geol. Surv Bull. 38, pt. 7: 213-214, pi. 2.

Synaptomys vetus Wilson. Hibbard, 1949, Bull. Geol. Soc. Amer. 60: 1424.

1972

Fossil Microtines From Anza-Borrego Desert

5

Synaptomys ( Synaptomys ) landesi Hibbard, 1954, Jour. Mammal. 35, (n. 2) :
249-252, Fig. 1C.

Holotype.- UMMP V29961, fragmentary left ramus with MrM3.

Horizon and type locality.— Type section of the Crooked Creek Forma-
tion, early Pleistocene, Borchers local fauna, Meade County, Kansas.

Emended diagnosis— Synaptomys landesi is distinguished by its Mx, on
which the anterior loop possesses a shallow internal and external re-entrant
angle. The outline of the second alternating triangle of the M. is generally
triangular in shape. The posterior loop of the M3 is triangular in shape and the
enamel is present along most of the posterior face. S. landesi appears to be
slightly larger than the other members of the subgenus.

Additional description— Synaptomys landesi has the first and second alter-
nating triangles of the Mx confluent, as do the other members of the subgenus.
The second alternating triangle is closed off from the third. This closure was
a character used by Hibbard (1954) to separate S. landesi from S. vetus. In
the latter species the second triangle opened into the third. While this is true
of some specimens, the majority of S. vetus specimens (19/22) also have
these triangles closed. This variation in the degree of closure may be due to
differences in the ontogenetic age of the individuals. The younger individuals
show less closure. There is a tendency for the second alternating triangle of the
M. in S. landesi to be more triangular in shape (Fig. IB) than in the other
species assigned to Metaxyomys, though there is some rounding exhibited in
a few specimens ( 2/ 8 ) .

Both of the re-entrants on the anterior loop are shallow. There is no
evidence of cement in either of these re-entrants. From the sample on hand
it appears that Synaptomys landesi is the largest of the three species (Fig. 2).
This size difference is especially apparent in the height of the tooth.

The M2 has the first and second alternating triangles confluent as in the
subgenus Mictomys. The second triangle of S. landesi seems to be better
developed, however. The third and fourth alternating triangles open broadly
into each other, but the shape approaches more closely that of the subgenus
Synaptomys for these triangles. The first external re-entrant of the M2 in S.

1 1

\ B | C

S. (M.)

anzaensis (1)

1 A

Bl c

S. (M)

Sandesi (7)

S. (M)

vetus 08)

Figure 2. Bar diagrams showing variations in size parameters of Synaptomys
( Metaxomys ) Mis. Vertical line represents the mean of the sample, the darkened
area two standard errors of the mean, the white areas a standard deviation on either
side of the mean, and the horizontal line the observed range. A, measurements
(mm) of the width of the occlusal surface; B, length; and C, height of enamel
crown. Size of sample is found in parentheses beside specific name.

6

Contributions in Science

No. 221

landesi is short and broad as in Mictomys. The second external re-entrant is
larger and more anteriorly directed than in either of the two Recent subgenera.

The M3 of Synaptomys landesi resembles its M2, in that the first and
second alternating triangles resemble Mictomys, but are not as confluent, while
the third and fourth alternating triangles more closely resemble Synaptomys.
These M2 and M3 patterns do not differ significantly from those observed in
S. vetus.

The characters of the lower jaw of Synaptomys landesi are largely un-
known. The diastemal region is as robust and the incisor ends in the same
position as in S. vetus.

The M1 and M2 of Synaptomys landesi resemble more closely those of
the subgenus Mictomys, in that there is a slight development of an internal
re-entrant angle near the posterior end of the teeth. The development of this
re-entrant varies and in some cases is represented only by a shallow groove.
These upper molars, likewise, do not differ significantly from those of the
other two species of Metaxyomys except possibly in size.

Three M3s of Synaptomys landesi are known. All three M3s have the
posterior loop shaped similarly to that of Mictomys but are not quite as
angular (Fig. IF). Enamel is absent from the posterior internal surface. The
above two features readily distinguish S. landesi from S. vetus. In two of the
M3s, alternating triangles one and three open into two to give a C-shaped pat-
tern. In the third tooth the middle external re-entrant nearly extends across
the entire width of the tooth isolating the third alternating triangle. The first
triangle still opens into the second.

Synaptomys ( Metaxyomys ) anzaensis, n. sp.

(Figure II)

Holotype— LACM 19684, isolated RMX.

Horizon and type locality.— Palm Springs Formation, probable middle
Pleistocene, Vallecito Creek local fauna, San Diego County, California, loc.
6683.

Paratypes.- LACM 24539 RM1, LACM 19071 2RM2s.

Diagnosis.— Synaptomys anzaensis is distinguished from the other mem-
bers of the genus by the shape of the anterior loop on the M1? which is very
vole-like and has deep internal and external re-entrant angles. The anterior
loop is larger than that of other members of the genus examined.

Description of holotype.— The Mt of Synaptomys anzaensis has the first
and second alternating triangles confluent. The remaining connections be-
tween the loops and triangles are closed. The anterior loop is vole-like and
relatively large when compared to other members of the genus. The re-
entrant angles on the anterior loop are relatively deep but are rounded and lack
cement. Cement is present in the remaining re-entrants of the tooth. The sec-
ond and third internal re-entrants curve more forward than in S. cooperi.
This feature appears to give the triangles a greater individuality than in the
latter species. The external triangles of S. anzaensis are more rounded than
those of S. vetus (Fig. II).

Etymology.— Anza from the Anza-Borrego Desert, ensis a suffix denoting
a geographic location.

1972

Fossil Microtines From Anza-Borrego Desert

7

Description of paratypes— These teeth have been characterized above
in the discussions of the upper molars. They are not diagnostic enough, except
possibly in size, to be separated from other members of the subgenus.

Discussion.— The lower jaw and M3 of Synaptomys anzaensis are un-
known at this time. I believe that if found their characteristics will be rela-
tively close to those of the other members of the subgenus. To name a new
species on the basis of so little material often presents problems, but the
characteristics of the tooth, geographic separation of the localities, and the
relatively large samples of the other members in the subgenus are sufficient to
warrant this approach.

Relationships— Because of the intermediate nature of the species which
make up the Metaxyomys group it may appear that one or more could be
ancestral to the extant subgenera of Synaptomys. The chief argument against
this relationship is that good representatives of one extant subgenus are known
from the fossil record prior to the first known appearance of Metaxyomys.
This is Synaptomys ( Synaptomys ) rinkeri from the Dixon local fauna of Ne-
braskan age (Hibbard, 1956). The first known occurrence of the subgenus
Mictomys is Synaptomys (M.) meltoni from the Cudahy fauna of Kansan
age (Paulson, 1961). Faunas in which the subgenus Metaxomys is found are
presently considered to be of Aftonian age (Hibbard, personal communica-
tion). If Metaxyomys types served as a common ancestor the divergence must
have occurred some time in Pliocene time and the forms discussed above have
carried on as a generalized type before becoming extinct.

Microtine, gen. and sp. indet.

(Figure 1J)

Material —L ACM 24647 RM3; LACM 24648 3M1s, 2M2s, 2 M0s; loc.
1357.

Description.— The above specimens represent a small, primitive micro-
tine with rooted teeth that lack cement in the re-entrant angles and have
essentially no or only very slightly-developed dentine tracts. The M2s possess
the typical microtine pattern which consists of a posterior loop and four
alternating triangles. The posterior loop is closed off from the alternating
triangles. In one of the M2s, considered to be a young adult individual, the first
and second triangles open into each other, but are closed off from the third and
fourth triangles which open into each other. These features impart a trilophed
appearance to the tooth. The other M2, in an adult stage of wear, shows a
closure between the first and second alternating triangles. The enamel on the
occlusal surface is very thick. There is no cement in the re-entrants and den-
tine tracts are essentially absent.

The M*s of this species also possess the typical microtine pattern which
consists of an anterior loop and four alternating triangles. The pattern most
closely resembles that of Pliophenacomys from the Fox Canyon local fauna
of Kansas (Hibbard, 1950). This similarity results from the development of a
posterior internal re-entrant which isolates the fourth triangle from the third.
An incipient “re-entrant pit” (Zakrzewski, 1969) is also present at the base of
this re-entrant. There is also an incipient development of a re-eritrant, or at
least a slight depression, on the posterior external surface. There is a groove

8

Contributions in Science

No. 221

on the anterior face of the ontogenetically youngest specimen. The enamel is
of equal thickness around the entire crown. Dentine tracts are incipient or
very slight on the internal triangles and on the posterior face. Cement is ab-
sent from the re-entrants. The teeth are three rooted, with the medial-internal
root being very reduced. The M1s of this microtine approach Ophiomys taylori
and the Fox Canyon Pliophenacomys in size. Microtine, gen. and sp. indet.,
differs from the former form because of the development of the posterior re-
entrants, being slightly more hypsodont, having slightly better-developed
dentine tracts, and a more reduced medial-internal root. It differs from the
latter by the following characteristics: slightly more hypsodont; less-developed
roots, especially the medial-internal; less-developed dentine tracts; and shal-
lower internal re-entrants, with the exception of the most posterior one which
is deeper.

The M2s of Microtine, gen. and sp. indet., also resemble those of Plio-
phenacomys from the Fox Canyon local fauna. This similarity, again, results
from the development of a posterior-internal re-entrant which tends to isolate
the third alternating triangle. Pliophenacomys differs from the form under
consideration in the following ways: Pliophenacomys is more robust for the
size of the tooth; the enamel is thinner; the posterior-internal re-entrant is
better developed which makes the third triangle more equilateral; it possesses
a higher percentage of re-entrant pits; and the dentine tracts are slightly better
developed. The M2s of Microtine, gen. and sp. indet., also resemble that of a
yet undescribed microtine in the collection of the University of Michigan
Museum of Paleontology from the Sand Draw local fauna of Nebraska. The
chief difference between the forms is that in the California specimens the
posterior-internal re-entrant angle is deeper and the teeth have three roots,
while the Nebraska specimen has only two roots.

In addition to the M1? which is unfortunately missing in this case, the
most diagnostic of the microtine teeth is the M3. The M3 of the microtine in
question is characterized by an anterior loop, two alternating triangles and a
posterior loop. The anterior loop opens slightly into the first triangle; the first
triangle is confluent with the second; and the second triangle opens slightly
into the posterior loop (Fig. 1J). The re-entrant angles are perpendicular
except for the posterior-internal one which curves slightly back. The two
anterior re-entrants are almost opposed. The enamel is very thick and equal
around the entire tooth. Dentine tracts are absent. The tooth is two rooted.
With the exception of the last character, the above description removes the
California specimens from the genus Pliophenacomys which they resemble in
the other upper molars. The M3 of Pliophenacomys consists primarily of two
lophs. The alternation of the triangles is not readily apparent. These features
of the M3 in Pliophenacomys were the chief reason why some microtines
earlier placed in the genus were removed and the genus Ophiomys was erected
for them (Hibbard and Zakrzewski, 1967). Though at about the same stage-of-
evolution as some species of Ophiomys, the California specimens probably do
not belong to that genus. The degree of confluency of the alternating triangles
on the M3 of the California specimen, along with the other characters of that
tooth, suggests that another line of microtine is present. This line of reasoning
is supported by the other molar teeth which morphologically appear closer

1972

Fossil Microtines From Anza-Borrego Desert

9

to the Pliophenacomys line. The M3 under discussion most closely resembles
one belonging to a microtine from the Sand Draw local fauna. The Sand Draw
M3 was found at the locality mentioned above in the discussion of the M2s.
The two M3s differ from each other only very slightly. Unfortunately for this
study several genera and species of microtine are represented in the Sand
Draw local fauna and not all of the specimens can be assigned properly be-
cause their remains are not associated. The two teeth from the Sand Draw
discussed above are thought to be related to each other solely because of their
morphological similarity to the Vallecito specimens, which by the laws of
probability should represent one species. It might be mentioned here that the
triangles on some young M3 specimens of Ophiomys taylori approach the
above teeth in degree of openness but are not as confluent, nor is the posterior
internal re-entrant well developed. Possibly these microtines could have been
derived from Ophiomys but they certainly represent a different line than the
other species so far assigned to that genus. Only additional material will show
the true relationship of this form.

Microtus californicus?

(Figure IK)

Referred material- LACM 24540 RM1? loc. 6814; LACM 8252 RM1?
loc. 1942; LACM 24649 left edentulous lower jaw, loc. 6683.

Description of material— LACM 24540 is an isolated RMj characterized
by a posterior loop, five alternating triangles, and an anterior loop. The ante-
rior loop can be considered to be composed of two additional alternating tri-
angles and a small anterior loph (Fig. IK). The posterior loop and the alter-
nating triangles open very slightly into each other except on the anterior loop
where the two triangles open broadly into the anterior loph. The enamel is
differentiated into thin tracts on the posterior sides of the loops and triangles
and thick tracts on the anterior edges, except on the anterior loop where the
enamel is absent from the entire anterior face. Cement is present in all the
re-entrants. The tooth is evergrowing. LACM 24540 measures 3.39 mm in
length, 1.28 mm in width, and 4.72 mm in height.

The other Mx (LACM 8252) is broken and still emplaced in a block of
matrix. Only a portion of the anterior loop and 3 alternating triangles are
visible. The tooth is assigned to M. californicus? for the following reasons: it is
evergrowing, it has the enamel differentiated into thin and thick tracts, it has
an anterior loop similar to LACM 24540, and it has cement in the re-entrants.

The lower jaw (LACM 24649) is assigned to M. californicus? because
the alveolus for the M1 shows that the tooth was evergrowing and of a type
similar to LACM 24540 in terms of loops and alternating triangles. A rela-
tively well-developed temporal fossa is also present. A similarly developed
fossa is characteristic of the genus Microtus but is not found in Synaptomys.
The latter genus is the only other microtine with evergrowing teeth present
at this locality. However, it may be that the lower jaw represents another
genus and/or species.

LACM 24540 in its stage of development, most closely resembles
Microtus californicus among the Recent microtines. The fossil differs from
the Recent forms chiefly in that the anterior loop is more rounded and the 6th

10

Contributions in Science

No. 221

alternating triangle is better developed. The fossil is larger than the specimens
assigned to M. californicus in the collection of the Department of Biological
Sciences, Fort Hays Kansas State College. The latter remains are thought to
represent at least two and possibly three different subspecies on the basis of
locality data with the specimens. The fossil is slightly smaller than the speci-
men of M. c. aequivocatus (UMMZ 79598) examined at the University of
Michigan. It may be that the fossil represents a subspecies of M. californicus.
It may also be that of a new species ancestral to extant M. californicus.
But as the M1 occlusal pattern of Microtus is highly variable it seems best to
follow the above approach till more material becomes available.

Paleoecology and Age of Localities

Microtines are generally found in boreal and/or relatively humid areas.
This suggests that during the time the fossils lived the climate in the area of
the Anza-Borrego Desert was more humid. Microtines are not found in this
area at present. As the exact stratigraphic position of all the localities which
contain microtines is not known (see discussion below), it cannot be stated
with certainty whether the presence of the microtines coincides with an alter-
nation of climate because of glaciation or whether other factors are involved.

All of the Synaptomys specimens and the edentulous Microtus lower jaw
were found at loc. 6683. This might suggest that this locality was near a bog
or marsh.

Downs and White (1968) compared the forms present in these deposits
with the distribution in time of all known mammalian genera which have a
fossil record in the late Pliocene and/or Pleistocene of North America. On the
nature of the genera and their associations they arbitrarily divided the fossils
in the superposed type section into three local faunas. The lowermost local
fauna is the Layer Cake of early Blancan age, the middle local fauna is the
Arroyo Seco of late Blancan age, and the uppermost local fauna is the Valle-
cito Creek of Irvingtonian age.

Loc. 1942, from which one of the RM^ (LACM 8252) of Microtus
californicus? is known, is the only locality of the four where microtines are
found in the type section. Its position, 12,100 feet from the base of the section,
places loc. 1942 in the Vallecito Creek local fauna. Savage (1951) assigned
2 RMjS from the type Irvingtonian local fauna to M. californicus. This fact
adds support for considering the Vallecito Creek l.f. to be of Irvingtonian age.

Loc. 6814 is not in the type section. A RMX of Microtus californicus? is
found here. From the above discussion this locality is tentatively placed in the
Vallecito Creek l.f.

Loc. 1357, where the remains of Microtine gen. and sp. indet. are found,
is also not in the type section. The microtine from this locality appears to be
most closely related to a species as yet undescribed from the Sand Draw local
fauna of Nebraska. The Sand Draw l.f. is thought to be late Blancan in age.
Hibbard (1970, footnote, p. 414) considers the Sand Draw l.f. to be Pleisto-
cene, but to represent a time prior to continental glaciation. If a direct correla-
tion can be assumed loc. 1357 would then be a part of the Arroyo Seco l.f.
At present there is not enough evidence to substantiate whether or not this is
indeed the case.

1972

Fossil Microtines From Anza-Borrego Desert

11

Loc. 6683, where the remains of Synaptomys anzaensis are found, is like-
wise not a part of the type section. The other species of Synaptomys assigned
to the subgenus Metaxyomys are found in faunas considered Aftonian in age.
The Borchers If. of Kansas is considered to represent an interglacial stage
because of the nature of the fauna and was thought to be Yarmouthian because
of its position above a volcanic ash (Hibbard, 1941). It has recently been sug-
gested to Hibbard that the ash below the Borchers l.f. is not the Pearlette-like
ash above the Cudahy l.f. He now considers the Borchers l.f. to be Aftonian
in age (Hibbard, personal communication). Support for this thesis is the pres-
ence of some relict Pliocene species in the Borchers l.f. (Sorex taylori, Perog-
nathus pearlettensis, Etadonomys tiheni, and Hypolagus sp.), and the better
fit which would be obtained for the size chronocline in Ondatra demonstrated
by Semken (1966). Hibbard is also working on a warm fauna from Ellsworth
County, Kansas, which has an entirely different faunal complement from the
Borchers, and is thought to be Yarmouthian on stratigraphic evidence. The
Grand View l.f. of Idaho is correlated with the Borchers l.f. primarily on the
stage-of-evolution of the Synaptomys and Ondatra which are known from
both local faunas (Hibbard, 1959). Direct correlation on the basis of Meta-
xyomys would suggest an Aftonian age for loc. 6683. However, if the edentul-
ous lower jaw of Microtus californicus? from loc. 6683 is correctly assigned,
it may mean that Metaxyomys is a relict in California and this locality is of a
younger age. Only more detailed work will enable us to know the true rela-
tionships of the deposits and contained faunas. The above discussion merely
points out how scanty our information is and some of the possibilities one
should be aware of.

Acknowledgments

I am especially indebted to Theodore Downs, Chief Curator of Earth Sci-
ences, Los Angeles County Museum of Natural History (LACM), Claude W.
Hibbard, Museum of Paleontology, University of Michigan (UMMP), and
John A. White, Idaho State University Museum, for the" opportunity to study
and report on the specimens under their care and for their critical reading of
the manuscript.

I would also like to thank the following for the permission to examine
specimens under their care: Eugene D. Fleharty, Department of Biological
Sciences, Fort Hays Kansas State College; Emmet T. Hooper, Museum of
Zoology, University of Michigan (UMMZ); and J. Knox Jones, Museum
of Natural History, University of Kansas (KU).

Graduate Faculty Research Grants (5326 and 5409) from Fort Hays
Kansas State College provided for the line drawings by Jerry L. Maxfield.
This study began at Idaho State University while I was on a postdoctoral
fellowship sponsored jointly by Idaho State University and the Los Angeles
County Museum of Natural History (NSF-GB 5116).

12

Contributions in Science

No. 221

Literature Cited

Baker, R. H. 1969. Cotton rats of the Sigmodon fulviventer group. Univ. Kans.

Mus. Nat. Hist., Misc. Publ. 51: 177-232.

Downs, T., and J. A. White. 1968. A vertebrate faunal succession in superposed
sediments from late Pliocene to middle Pleistocene in California. Inter. Geol.
Cong. Report XXIII Session, Proc. Sec. 10: 41-47.

Hibbard, C. W. 1941. The Borchers fauna, a new Pleistocene interglacial fauna
from Meade County, Kansas. Kans. Geol. Surv. Bull. 38: 197-220.

1950. Mammals of the Rexroad formation from Fox Canyon, Kansas.

Contrib. Mus. Paleontol., Univ. Mich. 8: 113-192.

1952. Vertebrate fossils from late Cenozoic deposits of central Kansas. Univ.

Kans. Paleontol. Contrib. Vert. Art. 2: 1-14.

1954. A new Synaptomys, an addition to the Borchers interglacial (Yar-
mouth?) fauna. Jour. Mammal. 35: 249-252.

1956. Vertebrate fossils from the Meade formation of southwestern Kansas.

Papers Mich. Acad. Sci. 41: 145-200.

1959. Late Cenozoic microtine rodents from Wyoming and Idaho. Papers

Mich. Acad. Sci. 44: 3-40.

1963. A late Illinoian fauna from Kansas and its climatic significance.

Papers Mich. Acad. Sci. 48: 187-221.

1970. Pleistocene mammalian local faunas from the Great Plains and Cen-
tral Lowland Provinces of the United States, p. 395-433. in Pleistocene and
Recent environments of the Central Great Plains. Dept. Geology, Univ. Kans.
Spec. Publ. 3. 433 p.

Hibbard, C. W., and R. J. Zakrzewski. 1967. Phyletic trends in the late Cenozoic
microtine Ophiomys gen. nov., from Idaho. Contrib. Mus. Paleontol., Univ.
Mich. 21: 255-271.

Paulson, G. R. 1961. The mammals of the Cudahy fauna. Papers Mich. Acad. Sci.
46: 127-153.

Savage, D. E. 1951. Late Cenozoic vertebrates of the San Francisco Bay Region.

Univ. Calif. Publ., Bull. Dept. Geol. Sci. 24: 339-410.

Semken, H. A. Jr., 1966. Stratigraphy and paleontology of the McPherson Equus
beds (Sandahl local fauna), McPherson County, Kansas. Contrib. Mus.
Paleontol., Univ. Mich. 20: 121-178.

Wilson, R. W. 1933. A rodent fauna from later Cenozoic beds of southwestern
Idaho. Carnegie Inst. Wash. Publ. 440: 117-135.

Zakrzewski, R. J. 1967. The primitive vole, Ogmodontomys, from the late Cenozoic
of Kansas and Nebraska. Papers Mich. Acad. Sci. 52: 133-150.

1969. The rodents from the Hagerman local fauna, Upper Pliocene of Idaho.

Contrib. Mus. Paleontol., Univ. Mich. 23: 1-36.

Accepted for publication September 13, 1971

Printed in Los Angeles, California by Continental Graphic

NUMBER 222
FEBRUARY 22, 1972

A NEW NIGHT SNAKE FROM MEXICO
(SERPENTES: COLUBRIDAE)

By James R. Dixon and Carl S. Lieb

CONTRIBUTIONS IN SCICNCC

NATURAL HISTORY MUSEUM

H

• LOS ANGELES COUNTY

CONTRIBUTIONS IN SCIENCE is a series of miscellaneous technical papers
in the fields of Biology, Geology and Anthropology, published at irregular intervals
by the Natural History Museum of Los Angeles County. Issues are numbered sep-
arately, and numbers run consecutively regardless of subject matter. Number 1 was
issued January 23, 1957. The series is available to scientific institutions and scien-
tists on an exchange basis. Copies may also be purchased at a nominal price. Inquiries
should be directed to Virginia D. Miller, Natural History Museum of Los Angeles
County, 900 Exposition Boulevard, Los Angeles, California 90007.

INSTRUCTIONS FOR AUTHORS

Manuscripts for CONTRIBUTIONS IN SCIENCE may be in any field of
Life or Earth Sciences. Acceptance of papers will be determined by the amount
and character of new information. Although priority will be given to manuscripts
by staff members, or to papers dealing largely with specimens in the collections of
the Museum, other technical papers will be considered. All manuscripts must be
recommended for consideration by the curator in charge of the proper section or
by the editorial board. Manuscripts must conform to those specifications listed
below and will be examined for suitability by an Editorial Committee including
review by competent specialists outside the Museum.

Authors proposing new taxa in a CONTRIBUTIONS IN SCIENCE must
indicate that the primary type has become the property of a scientific institution of
their choice and cited by name.

MANUSCRIPT FORM.— (1 ) The 1964 AIBS Style Manual for Biological
Journals is to be followed in preparation of copy. (2) Double space entire manu-
script. (3) Footnotes should be avoided if possible. Acknowledgments as footnotes
will not be accepted. (4) Place all tables on separate pages. (5) Figure legends and
unavoidable footnotes must be typed on separate sheets. Several of one kind may
be placed on a sheet. (6) An abstract must be included for all papers. This will be
published at the head of each paper. (7) A Spanish summary is required for all
manuscripts dealing with Latin American subjects. Summaries in other languages
are not required but are strongly recommended. Summaries will be published at
the end of the paper. (8) A diagnosis must accompany any newly proposed taxon.
(9) Submit two copies of manuscript.

ILLUSTRATIONS.— All illustrations, including maps and photographs, will
be referred to as figures. All illustrations should be of sufficient clarity and in the
proper proportions for reduction to CONTRIBUTIONS page size. Consult the 1964
AIBS Style Manual for Biological Journals in preparing illustration and legend copy
for style. Submit only illustrations made with permanent ink and glossy photo-
graphic prints of good contrast. Original illustrations and art work will be returned
after the manuscript has been published.

PROOF.— Authors will be sent galley proof which should be corrected and
returned promptly. Changes in the manuscript after galley proof will be billed to
the author. Unless otherwise requested, page proof also will be sent to the author.
One hundred copies of each paper will be given free to each author or divided
equally among multiple authors. Orders for additional copies must be sent to the
Editor at the time corrected galley proof is returned. Appropriate order forms will
be included with the galley proof.

Virginia D. Miller

Editor

A NEW NIGHT SNAKE FROM MEXICO
(SERPENTES: COLUBRIDAE)

By James R. Dixon1 and Carl S. Lieb1

Abstract: A new species of Hypsiglena is described from
eastern Queretaro, Mexico. Taken from the arid tropical scrub
forest of the Jalpan Valley, the new species is distinguished from
other Hypsiglena by having a longer tail and broader dorsal
bands. The new species is morphologically distinct from the
nearest population of H. torquata jani.

During the past 110 years, or since the original description of Hypsiglena
torquata (Gunther), 1860 and H. ochrorhyncha Cope, 1860, the status of the
two taxa has been a subject of much debate.

Cope (1860) proposed the genus Hypsiglena (type species: ochrorhyn-
cha) from specimens collected at Cape San Lucas, Baja California. Ten
months earlier, Gunther (1860) described a snake ( torquata ) from Nicaragua
that he assigned to the genus Leptodeira. Later Gunther (1895) reassigned
torquata to Hypsiglena and included ochrorhyncha as a synonym. He recog-
nized the fact that two nape patterns existed: torquata with a cream-colored
nuchal band followed by a single dark nuchal band, and ochrorhyncha which
lacked the light band but possessed a dark nuchal band or spots.

Dunn (1936) placed the genus Hypsiglena in the synonymy of Lepto-
deira, asserting that the presence or absence of grooves in the posterior maxil-
lary teeth was of no significant value. Taylor (1938) suggested retaining
Hypsiglena and presented a redefinition of Hypsiglena. Tanner’s (1944)
systematic review of the genus Hypsiglena supported Taylor’s views.

Three additional taxa of importance to this problem were named sub-
sequent to Cope’s description of the genus. Duges (1866) described Liophis
jani from Guanajuato, Mexico; Stejneger (1893) described Hypsiglena texana
from near Laredo, Texas; and Taylor ( 1938) named a new race, H. torquata
dunklei, from near Forlon, Tamaulipas, Mexico.

Boulenger (1894) placed H. texana Stejneger in the synonymy of H.
ochrorhyncha Cope. Cope (1900) agreed with this, as did Taylor (1938).
No mention was made by Boulenger or Cope of the status of Liophis jani.
Even though texana was considered a race of ochrorhyncha by Stejneger
and Barbour (1917), it was not generally recognized until Tanner’s (1944)
systematic review of the genus was published.

Taylor (1938) placed the species Liophis jani in the synonymy of tor-
quata. However, Smith (1943) resurrected jani and gave it subspecific status

department of Wildlife and Fisheries Sciences, Texas A&M University, College
Station, Texas 77843.

1

2

Contributions in Science

No. 222

with H. ochrorhyncha. This arrangement was followed by Tanner (1944)
and Smith and Taylor (1945). Dixon (1965) reviewed the status of jani,
texana, and ochrorhyncha, and proposed relegating H. o. texana to synonymy
of H. o. jani. Taylor’s (1938) race, Hypsiglena t. dunklei, was elevated to a
specific level by Tanner ( 1944). Dixon (1962) utilizing additional specimens,
returned dunklei to a race of torquata based on the similarity of nuchal
patterns.

Though H. torquata and H. ochrorhyncha were considered distinct
species by earlier authors, Dunn (1936) suggested they were conspecific.
Taylor (1938) and Tanner (1944) considered the possibility of combining
the two forms, but favored retaining the two as separate species. Bogert and
Oliver (1945) favored the arrangement of Dunn, based on similarities of
squamation and habitat of snakes with both nuchal types from southern
Sonora, Mexico. As a result, the races of ochrorhyncha ( texana , jani) became
races of torquata. This arrangement was maintained for the next 20 years
until Dixon (1965) proposed the separation of ochrorhyncha from torquata
at the specific level, based on the presence or absence of a light cream colored
nuchal collar. Tanner (1966) and Hardy and McDiarmid (1969) suggested
returning to Dunn’s 1936 classification.

An attempt was made to solve the species problem by Mr. Ernest Tanzer,
through the use of karyotyping, examination of all preserved material, and
examination of hatchlings from clutches of eggs from areas of overlap between
the two forms. Unfortunately, before the completion of his monograph of the
genus, Ernest Tanzer died in January, 1971.

For the past three years we (Dixon, Ketchersid and Lieb) have under-
taken a study of the herpetofauna of the Mexican state of Queretaro and
recently discovered what we consider a new species of Hypsiglena in the iso-
lated valley of the Rio Jalpan. In memory of Ernest Tanzer, a close friend and
ardent field companion, we propose the new species be known as

Hypsiglena tanzeri, new species
Figure 1A, B

Holotype — Texas Cooperative Wildlife Collection 34079, male, taken
5 km E. Jalpan, Queretaro, Mexico, 762 m, 99° 27' W. 21° 13' N., by Frank
Guyer, 13 April 1971.

Par atype. —Los Angeles County Museum of Natural History (72068)
male, taken 0.8 km W. Landa de Matamoros, 1067 m; Queretaro, Mexico,
by Carl Lieb and Douglas Albaugh, 20 May 1971.

Diagnosis.— Hypsiglena tanzeri is readily distinguished from populations
of snakes that have been referred to H. ochrorhyncha by possessing a broad
nuchal collar. It most closely resembles H. torquata from which it is readily
distinguished by the presence of wider (usually reaching the second dorsal
scale row) and longer (usually 3 to 4 scales long) brown to black bands on
the anterior two-thirds of the body. H. torquata has dorsal blotches reaching

1972

A New Night Snake From Mexico

3

the sixth or seventh scale row and 2 to 3 scales long. Additionally H. tanzeri
has the following characters: tail length, 22.5 per cent of the total length;
postocular stripe not continuous with nape blotch, and 7 supralabials; whereas,
H. torquata has a tail length 19.5 per cent or less of the total length; a postocu-
lar stripe that may or may not join the nuchal blotch; usually (99 per cent of
specimens examined) 8 or more supralabials.

Description of the Holotype— Head and body proporations normal for
Hypsiglena; total length 328 mm, tail 73 mm, ratio of tail to total length .225;
dorsal scale formula 21-21-17; 2 pair of chin shields of about equal length;
ventrals 175; caudals 69, including tip; supralabials 7-7; infralabials 10-10,
preoculars 2-2, postoculars 2-2, loreal single; temporals 1 + 2 + 3; third and
fourth supralabials entering eye; fifth and sixth supralabials and sixth infra-
labials largest of series; first five infralabials contact first pair chin shields,
primary temporal contacts parietal, both secondary temporals, fifth and sixth
supralabials; maxillary teeth 9-9, followed by a distema and two large un-
grooved teeth.

Color pattern consists of 37 black bands and spots on body; anterior 26
bands 16 to 17 scale rows wide, reducing the intercalary spots to narrow
brownish black lines on scale rows one and two on most of the body; dorsal
color pattern above the last 26 ventrals consists of three rows of spots, the
outer rows alternating with the middle row; interspaces between major bands
one to two scales in length, grayish white with dark pigment in the center of
each scale; nape band ten scales in length, 15 scales if the median anterior
projection of the nape band (Fig. 1A) is included in the count; dorsal surface
of head heavily pigmented with brownish black spots on a ground color of
dark gray; nasal-prefrontal area heavily pigmented, less so from frontal to
posterior edge of parietal; postocular stripe reaching one scale row beyond
last supralabial on left side, terminating on last supralabial on right side, not
connected to nape blotch; two coalesced black spots on medial edge of center
of parietals; labials dirty white with scattered marks of brownish black; mental
and anterior two infralabials heavily pigmented with brownish black; center
of each scale between parietals and anterior edge of nuchal blotch densely
pigmented with brownish black; venter cream white.

Variation— The single paratype (Fig. IB) differs from the type in having
the band color dark brown rather than black; 21 large anterior bands; posterior
bands small, forming an alternating series of smaller spots above the last 58
ventrals. The paratype lacks an anterior median projection of the nuchal
blotch. Otherwise, the general color pattern is similar to the type. Unfortu-
nately, the paratype is missing about one-third of its tail, and tail/ total length
ratio could not be determined. The only scale difference between the paratype
and the type is 178 ventrals in the former.

Remarks— Data from 315 specimens of Hypsiglena were made available
to us by Mrs. Glenda Tanzer and 200 additional specimens were examined

4

Contributions in Science

No. 222

Figure 1. Dorsal view of the holotype (A) and paratype (B) of Hypsiglena tanzeri,
and two male specimens (C and D) oi H. torquata from Queretaro.

1972

A New Night Snake From Mexico

5

by Dixon. The 515 specimens represent at least three distinct geographic
populations; (A) northern Sinaloa, Sonora, and Arizona; (B) the remainder
of Sinaloa, Nayarit, Jalisco, Michoacan, Guerrero, and Morelos; (C) all of
the Mexican Plateau, the lowlands of northeastern Mexico, Texas, New
Mexico, Oklahoma, and Kansas. These will be discussed in greater detail else-
where. Population “(C)” is of special interest since it is geographically closest
to the locality of H. tanzeri. The general variation of the former population is
as follows: Counts of ventrals plus caudals of 206 male specimens range from
200 to 232 (x = 215.0), in 177 females, 200 to 231 (x = 214.4); midbody
scale rows of both sexes generally 21 (in 373 specimens), with 23 rows occur-
ring three times, 20 (4), and 19 (5); scale rows just anterior to the vent are
highly variable, with 19 occurring in one specimen; 18 in 4, 17 in 275, 16 in
76, and 15 in 27. Preoculars usually two with a 1-1 combination occurring
in ten specimens, 1-2 in 8, 2-2 in 358, 2-3 in 5, 3-3 in 2; postoculars usually
two, with 1-1 occurring in eight specimens, 1-2 in 11, 2-2 in 358, 2-3 in 5,
and 3-3 in 1; supralabials usually eight, with 7-7 occurring in five specimens,
7-8 in 8, 8-8 in 359, and 8-9 in 11; infralabials usually ten, with 9-9 in 20
specimens, 9-10 in 31, 10-10 in 296, 10-1 1 in 20, 1 1-1 1 in 13, 1 1-12 in 2, and
12-12 in 1; loreals usually one, with 1-2 in 3 specimens, and 2-2 in 1. Body
blotches vary from 41 to 72 (x = 51,8) in 186 females, and 36 to 69 (x = 49.3)
in 215 males.

From the large series available, only 10 male specimens of H. torquata

TCWC 34082 TCWC 34079

Figure 2. Midbody color patterns of H. torquata (left) and H. tanzeri (right) from
Queretaro, Mexico.

6

Contributions in Science

No. 222

jani were found that represented the nearest geographic localities to the area
where H. tanzeri has been taken. The comparative material is represented by
one specimen from the Mexican state of Hidalgo, four from San Luis Potosi,
two from Queretaro (Fig. 1, C and D), and three from Guanajuato. Their
characteristics are: ventrals, 153-166 (x=159.3); caudals, 48-55 (x = 50.2);
ventrals + caudals, 201 to 219 (x = 210.4); supralabials, 8 in eight specimens,
7 in two; infralabials, 9 in two, and 10 in eight; preoculars, 1 in two, 2 in eight;
postoculars, 2 in all, third and fourth supralabials entering orbit in two, fourth
and fifth in eight; tail length ranges from 15.8 to 18.8 (x= 17.5) per cent of
total length; body blotches vary from 39 to 53 (x = 44.9); width of fifteenth
dorsal blotch ranges from 7 to 8 scale rows wide; all specimens with two to
three rows of intercalary spots.

Three specimens of H. torquata jani were taken 18 and 19 May, 1971,
in the Rio Extoraz Valley in Queretaro, some 58 airline kilometers from the
type locality of H. tanzeri. Their color pattern is typical of nearly all races of
H. torquata and show no approach to the condition found in H. tanzeri
(Fig. 2).

The differences presently used to separate the currently recognized races
of H. torquata are comparatively small. Most are based on the number of
ventrals and caudals, subtle differences in color pattern, and reduction in the
number of midbody scale rows. When these “racial” criteria are considered
with respect to the characters of H. tanzeri and adjacent populations of H.
torquata, additional support is gained for recognition of H. tanzeri at the
specific level.

Specimens of H. torquata jani have been taken from the desert in the
states of Queretaro and San Luis Potosi, Guanajuato, and Hidalgo, from
tropical deciduous forest in southern Tamaulipas and eastern San Luis Potosi,
and from arid scrub in southern San Luis Potosi. Although none have yet
come from the arid tropical scrub in the Jalpan Valley, they may someday be
taken there.

The soils in the vicinity of the type locality consist of Karst Limestone on
the slopes, with alluvial top soil on the valley floor. Dense thickets of cacti
and non-thorny shrubs cover the limestone hillsides and cultivated crops are
grown on the valley floor. The other nocturnal snakes which have been taken
from the same general area are Elaphe guttata, Lampropeltis triangulum,
Leptoderia septentrionalis, and Trimorphodon tau.

Acknowledgments

We wish to thank Juan Cifentes L., Biologist in the Department de Estu-
dios Biologicos Pesqueros de la Direccion General de Pesca e. Indus. Conexas.
for granting us a permit for our studies in Mexico. We would also like to thank
Douglas Albaugh, Frank Guyer, and James C. Kroll for their assistance in the
field.

1972

A New Night Snake From Mexico

7

Resumen

Se describe una nueva especie de Hypsiglena de la parte occidental de
Queretaro, Mexico. Colectada en la zona de mezquital (arido tropical) del
valle de Jalpan. La nueva especie se distingue de las otras Hypsiglena por tener
la cola mas larga y bandas dorsales mas anchas. La nueva especie es mor-
fologicamente diferente de las poblaciones mas cercanas de H. torquata jani.

Literature Cited

Bogert, C. and J. A. Oliver. 1945. A preliminary analysis of the herpetofauna of
Sonora. Amer. Mus. Nat. Hist. Bull. 83:301-425.

Boulenger, G. A. 1894. Catalogue of the Snakes in the British Museum (Natural
History). London. Vol. 2, 382 p.

Cope, E. D. 1860. Catalogue of the Colubridae in the Museum of the Academy of
Natural Sciences of Philadelphia, with notes and descriptions of new species.
Part 2, Proc.. Acad. Nat. Sci. Phila. 12:241-266.

1900. The Crocodilians, Lizards, and Snakes of North America. U.S. Natl.

Mus. Rep. 1898:153-1270.

Dixon, J. R. 1962. Three additional specimens of the night snake, Hypsiglena
dunklei. Herpetologica 8(2) : 134-135.

1965. A taxonomic reevaluation of the night snake Hypsiglena ochrorhyn-

cha and relatives. Southwest Nat. 10(2) : 125-131.

Dunn, E. R. 1936. Notes on North American Leptodeira. Proc. Nat. Acad. Sci.
22:106-119.

Duges, A. 1865. Du Liophis janii. Mem. Acad. Sci. Lett. Montpellier, 6:32-33.
Gunther, A. 1860. Description of Leptoderia torquata, a new snake from Central
America. Ann. Mag. Nat. Hist. Ser. 3, 5:169-171.

1895. Biologia Centrali-Americana. Reptilia and Batrachia. 1885-1902.

London. 326 p.

Hardy, L. M., and R. W. Me Diarmid. 1969. The amphibians and reptiles of Sina-
loa, Mexico. Univ. Kans. Publ., Mus. Nat. Hist. 18:39-252.

Stejneger, L. 1893. The Death Valley Expedition. A biological survey of parts of
California, Nevada, Arizona, and Utah. Part II: 2. Reptiles and Batrachians.
North American Fauna 7:159-234.

Stejneger, L., and T. Barbour. 1917. A checklist of North American amphibians
and reptiles. Harvard Univ. Press, Cambridge. 125 p.

Smith, H. M. 1943. Summary of the collection of snakes and crocodilians made in
Mexico under the Walther Rathbone Bacon Traveling Scholarship. Proc. U.S.
Nat. Mus. 93:393-504.

Smith, H. M., and E. H. Taylor. 1945. An annotated checklist and key to the snakes
of Mexico. U.S. Nat. Mus. Bull. 187:1-239.

Tanner, W. W. 1944. A taxonomic study of the genus Hypsiglena. Great Basin Nat.,
5(3-4) :25-92.

1966. The night snakes of Baja California. Trans. San Diego Soc. Nat. Hist.

14(15): 189-196.

Taylor, E. H. 1938. On Mexican snakes of the genera Trimorphoden and Hyp-
siglena. Univ. Kans. Sci. Bull. 25(16) :357-383.

Accepted for publication October 11, 1971

5 73

Ct-Im*

NUMBER 223
FEBRUARY 23, 1972

MORE VERTEBRATES, INCLUDING A
NEW MICROSAUR, FROM THE UPPER
PENNSYLVANIAN OF CENTRAL COLORADO

By Peter Paul Vaughn

CONTRIBUTIONS IN SCICNCC

8

NATURAL HISTORY MUSEUM • LOS ANGELES COUNTY

CONTRIBUTIONS IN SCIENCE is a series of miscellaneous technical papers
in the fields of Biology, Geology and Anthropology, published at irregular intervals
by the Natural History Museum of Los Angeles County. Issues are numbered sep-
arately, and numbers run consecutively regardless of subject matter. Number 1 was
issued January 23, 1957. The series is available to scientific institutions and scien-
tists on an exchange basis. Copies may also be purchased at a nominal price. Inquiries
should be directed to Virginia D. Miller, Natural History Museum of Los Angeles
County, 900 Exposition Boulevard, Los Angeles, California 90007.

INSTRUCTIONS FOR AUTHORS

Manuscripts for CONTRIBUTIONS IN SCIENCE may be in any field of
Life or Earth Sciences. Acceptance of papers will be determined by the amount
and character of new information. Although priority will be given to manuscripts
by staff members, or to papers dealing largely with specimens in the collections of
the Museum, other technical papers will be considered. All manuscripts must be
recommended for consideration by the curator in charge of the proper section or
by the editorial board. Manuscripts must conform to those specifications listed
below and will be examined for suitability by an Editorial Committee including
review by competent specialists outside the Museum.

Authors proposing new taxa in a CONTRIBUTIONS IN SCIENCE must
indicate that the primary type has become the property of a scientific institution of
their choice and cited by name.

MANUSCRIPT FORM.— ( 1 ) The 1964 AIBS Style Manual for Biological
Journals is to be followed in preparation of copy. (2) Double space entire manu-
script. (3) Footnotes should be avoided if possible. Acknowledgments as footnotes
will not be accepted. (4) Place all tables on separate pages. (5) Figure legends and
unavoidable footnotes must be typed on separate sheets. Several of one kind may
be placed on a sheet. (6) An abstract must be included for all papers. This will be
published at the head of each paper. (7) A Spanish summary is required for all
manuscripts dealing with Latin American subjects. Summaries in other languages
are not required but are strongly recommended. Summaries will be published at
the end of the paper. (8) A diagnosis must accompany any newly proposed taxon.
(9) Submit two copies of manuscript.

ILLUSTRATIONS. — -All illustrations, including maps and photographs, will
be referred to as figures. All illustrations should be of sufficient clarity and in the
proper proportions for reduction to CONTRIBUTIONS page size. Consult the 1964
AIBS Style Manual for Biological Journals in preparing illustration and legend copy
for style. Submit only illustrations made with permanent ink and glossy photo-
graphic prints of good contrast. Original illustrations and art work will be returned
after the manuscript has been published.

PROOF.— Authors will be sent galley proof which should be corrected and
returned promptly. Changes in the manuscript after galley proof will be billed to
the author. Unless otherwise requested, page proof also will be sent to the author.
One hundred copies of each paper will be given free to each author or divided
equally among multiple authors. Orders for additional copies must be sent to the
Editor at the time corrected galley proof is returned. Appropriate order forms will
be included with the galley proof.

Virginia D. Miller
Editor

MORE VERTEBRATES, INCLUDING A NEW MICROSAUR, FROM
THE UPPER PENNSYLVANIAN OF CENTRAL COLORADO

By Peter Paul Vaughn1

Abstract: More fossils, mostly vertebrate, have been col-
lected from a black shale low in the Sangre de Cristo Formation
in the Arkansas River valley of central Colorado. Some of the
fossils probably represent “residents” in the pond, but most of
the bones seem to have been washed in from nearby. None of
the recently found specimens conflicts with the estimate of Late
Pennsylvanian, probably Missourian, age based on the previously
reported fossils (Vaughn, 1969); and recognition of the pely-
cosaur Edaphosaurus cf. E. ecordi corroborates this estimate. The
presence of bisaccate gymnosperm pollen is attributed to prox-
imity of the area to elements of the Ancestral Rocky Mountains.

An almost complete, articulated skeleton of a new kind of micro-
saur requires the naming of a new genus and species, Trihecaton
howardinus. This form is clearly a microsaur, as shown by the
structure of the first vertebra and the character of the scales;
but it is remarkable in the infolding of the enamel of the
marginal teeth, and in the possession of large presacral inter-
centra with capitular facets. There are also large intercentral
haemal arches. A new family, Trihecatontidae, is based on the
genus, but the position of this family within the Microsauria
is obscure. T. howardinus seems primitive in a number of
respects, but it occurs too high in the stratigraphic column to
be regarded as an actual “urmicrosaur.” More materials of the
diadectid Desmatodon hesperis are now on hand. A braincase
and connected parts of the dermal roof of an apparently im-
mature specimen show essential similarity to the Early Permian
Diadectes, but the basipterygoid joint was mobile, and, as rarely
seen in Diadectes, there are narrow fenestrae between the post-
parietal and tabular regions. The teeth of a juvenile maxilla con-
trast with those of the holotypic maxilla in several ways: smaller
number, much greater relative length and more incisiform aspect
of the first two, separation by longer spaces, and lack of wear
facets. An analysis of the replacement cycle of the teeth in the
holotypic maxilla of Desmatodon hesperis shows a much longer
replacement wave than in Diadectes, and also reveals the exist-
ence of a “gap” between the second and third teeth. In its longer
replacement wave, Desmatodon hesperis may also differ from the
type species, Desmatodon hollandi, and it may become necessary
to set up a new genus. The problems of the origin and affinities
of the diadectids remain unsolved.

Research Associate in Vertebrate Paleontology, Natural History Museum of Los
Angeles County; and Department of Zoology, University of California, Los Angeles,
California 90024.

1

2

Contributions in Science

No. 223

Introduction

I have recently published on vertebrate fossils found low in the Sangre
de Cristo Formation of central Colorado (Vaughn, 1969). These fossils,
which seem to be of Late Pennsylvanian, more specifically Missourian, age
represent the only sizable North American tetrapod fauna of that age known
from west of the Garnett quarry of eastern Kansas (see Peabody, 1952),
although tetrapod trackways are known from elsewhere in the Pennsylvanian
of Colorado. They are of special interest in that they occur in the region of
the Ancestral Rocky Mountains of late Paleozoic time. The present paper is a
report on additional materials recently recovered from the same site; these
include remains of new faunal elements, an almost complete skeleton of a
new kind of microsaur, and materials that contribute significantly to our
understanding of the only known Pennsylvanian diadectid, Desmatodon.

The quarry is near the town of Howard in the valley of the Arkansas
River, in NW V* NEV4 SW14 sec. 22, T 49 N, R 10 E, Fremont County,
Colorado. It is about 1450 feet above the base of the Sangre de Cristo Forma-
tion as defined by Brill (1952) in a two- to three-foot thick black shale that
he designated as part of “Interval 300”; the total thickness of the formation
in this region is about 8800 feet. The steeply dipping attitude of the beds has
made quarrying difficult, and the shale is not yet completely exposed, but it
may be said in general that it is a lens-shaped deposit that probably represents
a pond, perhaps an oxbow within the general system of stream channels
indicated in this part of the formation. The presently exposed face is about
70 feet in length, from where it is cut by a small stream that flows close to a
soil-covered bank, to where it pinches out to the northwest. A study of the
layering of the shale and the pattern of distribution of the fossils within the
shale is in progress and will be reported at a later date.

I have been fortunate to have had access to Mr. Walter Pierce’s manu-
script on the stratigraphy of the Howard area (unpublished master’s thesis,
Colorado School of Mines, 1 969) , which has assisted greatly in my field work.
The Interval 300 quarry is in the upper part of Pierce’s “Unit V”; within the
lower part of this unit, which has a maximum thickness of about 3700 feet,
lies the limestone that Brill (1952) correlated with the Whiskey Creek Pass
Limestone to the south and with the Jacque Mountain Limestone to the north.
The Sangre de Cristo Formation as defined by Brill begins directly above this
limestone. Pierce demonstrates an angular unconformity between Unit V and
the overlying Unit VI, but it is not yet known how much time is represented
in this break. Pierce also shows that the dominant sedimentary pattern of
Unit V is the point-bar type, with streams that flowed generally toward the
northwest.

The Sangre de Cristo Formation was deposited in the southern half of
the trough that lay between the Late Pennsylvanian and Early Permian Front
Range to the east and the Uncompahgre Highland to the west. Mallory
(1958, 1960) has presented paleogeographic reconstructions of the trough

1972

New Vertebrates

3

and surrounding highlands. There is some debate as to the proper name of
the formation; Brill (1952) extended the name Sangre de Cristo northward
from New Mexico and southern Colorado, but Chronic (1958) has recom-
mended that this name not be used in central Colorado and that instead the
older term, Maroon, be retained to emphasize the essential continuity of the
beds with the Maroon Formation deposited in the northern half of the trough.
Geologists at the Colorado School of Mines are currently engaged in studies
of the general region around the Interval 300 quarry, and more detailed
stratigraphic analyses will soon appear, but for the present we may rely on
Brill’s terminology. It must also be noted that the lower parts of the Sangre de
Cristo Formation, in Brill’s usage of the term, are not everywhere of the same
age; in central Colorado the lower part, which includes Interval 300, is
apparently Late Pennsylvanian, but the equivalent strata in northern New
Mexico are marine and are assigned to the Madera Formation whereas the
terrestrial Sangre de Cristo Formation as mapped there is probably entirely
of Early Permian age (see Brill, 1952; Vaughn, 1969).

Additions to the Flora and Fauna

Most plant remains from the Interval 300 quarry are very poorly pre-
served, but there are carbonized bits of wood, and Calamites impressions are
recognizable. Stratigraphically somewhat lower, but still within Pierce’s Unit
V and above the limestone correlated with the Whiskey Creek Pass Limestone
by Brill, better preserved plants are found in SW XA SW XA NWV4 sec. 22.
These include Calamites pith casts, carbonized impressions of Walchia fronds,
and impressions of fernlike plants. Below this, in Pierce’s Unit IV— roughly
equivalent to the upper part of the Minturn Formation of Brill’s usage— are
found more Calamites remains and also bisaccate gymnosperm pollen, which
latter Scott (1967) takes as an indication of Permian rather than Pennsyl-
vanian age. Microscopic examination of the black shale of Interval 300 also
reveals bisaccate gymnosperm pollen. This may seem to conflict with the
evidence presented earlier (Vaughn, 1969) and below for Late Pennsylvanian
age of Interval 300, but not necessarily. Clapham (1970), in a study of Per-
mian pollen from Oklahoma, came to the conclusion that gymnosperms pro-
ducing bisaccate pollen were upland forms. Langenheim (1959:569) has
pointed out that our picture of late Paleozoic plant successions is based largely
on the record from farther east, where extensive uplands did not exist in the
Pennsylvanian, and that, therefore, “lowland plants became known as Penn-
sylvanian indices and upland plants as Permian indices in the mid-western
and eastern United States. . . however, the upland environments of the An-
cestral Rockies were close to the marginal swamp and marine environments
early in the Pennsylvanian. The result is the presence of ‘Permian’ plants in
intimate association with ‘Pennsylvanian’ plants and marine invertebrates.”
Elias (1970:696) accounts in much the same way for recent discoveries of
Walchia, previously not known to occur below the Upper Pennsylvanian, in

4

Contributions in Science

No. 223

the Middle Pennsylvanian of Oklahoma; he feels that “ Walchia undoubtedly
existed ... in the ancient highlands which rose in conjunction with the very
early Pennslyvanian orogenies in the areas of the Arbuckle and Ouachita up-
lifts.” The principle is similar to that expressed by Axelrod (1952) when he
suggested that angiosperms may have been present in the Permian and Triassic,
but “hidden” in uplands remote from the lowland areas of deposition. It seems
probable that the presence of bisaccate gymnosperm pollen in and even below
Interval 300 is only a matter of ecological difference, that is, a reflection of
the upland conditions of the nearby Ancestral Rocky Mountains and not
really inconsistent with the more compelling evidence of Late Pennsylvanian
age presented by the vertebrates.

Shells of small pelecypods were noted in my earlier paper; the only other
invertebrate remains that have been recognized in Interval 300 are parts of
carapaces of the branch iopod Cyzicus.

The vertebrates known to date from Interval 300 are:

Elasmobranch fishes
A xenacanth
Palaeoniscoid fishes
One or more kinds
Labyrinthodont amphibians
?Anthracosaurs

A large ?embolomere
Temnospondyls

Several small rhachitomes including the dissorophid lAmphibamus
Lepospondyl amphibians
Aistopods

Coloraderpeton brilli Vaughn, 1969
Microsaurs

Trihecaton howardinus, new genus and species
?Cotylosaurs

Desmatodon hesperis Vaughn, 1969
Pelycosaurian reptiles

A “medium-sized” ophiacodont
A small sphenacodont

Edaphosaurus aff. E. raymondi (Case, 1908)

Edaphosaurus cf. E. ecordi Peabody, 1957

The additions to my earlier list are: the xenacanth, the dissorophid
labyrinthodont, the new microsaur, and Edaphosaurus cf. E. ecordi’, it is also
now clear that I am dealing with more kinds of small rhachitomes than I
previously recognized. The great difficulty is in the disarticulated nature of
almost all the materials, which occur mostly as isolated bones. It is obvious
that more faunal elements are represented than can be identified.

The xenacanth “shark” is represented by only one spine (UCLA VP

1972

New Vertebrates

5

17361), but this, with its two rows of recurved denticles, is sufficient for iden-
tification. The presence of this freshwater form at this site is no surprise.

The remains of palaeoniscoid fishes are frustrating. Mostly, the scales
are isolated, but they also occur in articulated patches, and sometimes small
parts of skulls and fins are found in association with these patches. The scales
are almost all small, smooth and shiny, with diamond-shaped outlines and
with the peg-and-socket articulation common among palaeoniscoids, but rare
signs of different scales indicate the presence of more than one kind of palaeo-
niscoid. That these thin scales are found in patches as often as they are would
seem to indicate that palaeoniscoids, at least, were “resident” in the pond.
Perhaps also to be counted as a resident is the aistopod, Coloraderpeton
brilli, vertebrae and osteoderms of which are frequently found in articulation
or very close association. It may even be that trails of tiny pellets found in a
number of places in the quarry are the result of burrowing activity of this
aistopod. The presence of coprolites, some containing palaeoniscoid scales,
also points to a resident fauna. The scattered remains of almost all the other
faunal elements seem to indicate that they were washed in from elsewhere, but
parts of what was apparently a single individual of Desmatodon hesperis were
found not far removed from one another on the same plane, and the specimen
of the microsaur described below is almost complete and in nearly perfect
articulation. There are no signs of wear on the bones that would indicate
transportation from a distant source. It is possible, of course, that some of
the scattering of the bones may be due to predatory activity of the xenacanths,
but the lack of any discernible bite marks makes it seem more likely that the
specimens were washed in from the immediately surrounding area, perhaps
during periods of flooding of the adjacent streams.

In my earlier paper, I illustrated premaxillary and palatine bones (UCLA
VP 1700 and 1699) of what I took to be a large rhachitomous labyrinthodont.
More cranial parts of this form (UCLA VP 1737) are now at hand, including
another premaxilla and an associated large part of a dentary bone with small
teeth that are slightly recurved at their tips. The premaxilla has a formidable
tusk of about 25 mm length, with an oval base 10 mm long but only 6 mm
wide. All the teeth show deep infolding of the enamel and dentine in their
basal portions. The general shape of the preserved part of the dentary and
the traces of the meckelian fenestrae cause me to think that, contrary to my
earlier identification, this form may be an anthracosaur, perhaps an embolo-
mere similar to the large N eopteroplax conemaughensis described by Romer
(1963) from the Conemaugh Group of Ohio. Other elements, including a
squamosal bone (UCLA VP 1701), seem to support this. Better materials
must be found before a positive identification can be made, but it is clear
that a labyrinthodont amphibian of crocodilelike proportions was present in
the area.

*The designation UCLA VP refers to the vertebrate paleontological collection of the
University of California, Los Angeles.

6

Contributions in Science

No. 223

Many parts of small to “medium-sized” rhachitomes are represented in
the collection from Interval 300. Some of these, as a fairly well-preserved
dentary bone with slim, conical teeth (UCLA VP 1738), are of trimerorha-
choid aspect; but most are indeterminate. One small, slender dentary with
numerous, tiny teeth (UCLA VP 1739) seems clearly to be of a dissorophid.
It falls within the size range of dentaries of Amphibamus lyelli from the upper-
most part of the Allegheny Group of Ohio (Carroll, 1964), which it closely
resembles, and it may represent a species of that genus.

Descriptions of remains of a number of pelycosaurian reptiles including
a “medium-sized” ophiacodont and a small sphenacodont were given in my
earlier paper. More materials of these forms continue to appear. These
include: the basi-parasphenoid part of a braincase of the ophiacodont (UCLA
VP 1740), with a distance of 26 mm between the lateralmost points of the
basipterygoid processes; and a vertebra of the small sphenacodont (UCLA
VP 1741), with a centrum about 11 mm long and a neural arch that shows
the characteristic excavations on its lateral surfaces. These finds help to dem-
onstrate the taxonomic diversity of the fauna and also reinforce the develop-
ing picture of a variety of pelycosaurs in the Late Pennsylvanian (see DeMar,
1970).

My earlier paper also recorded the presence of the little edaphosaurian
pelycosaur Edaphosaurus aff. E. raymondi. It is now evident that there was
also another small edaphosaur. UCLA VP 1742 is a partial neural spine that
is strikingly similar in both shape and size to the spine of Edaphosaurus ecordi
Peabody, 1957. As in that species, the spine is laterally flattened but flares
distally within the anteroposterior plane and has only incipiently developed
tubercles along the sides. The holotype of E. ecordi was found near Garnett
in eastern Kansas, in a shale that forms part of a lagoonal deposit within the
Stanton Formation, Missourian Stage, Upper Pennsylvanian (see Peabody,
1952). The general nature and fauna of the Garnett quarry, from which come
the only known specimens of the reptile Petrolacosaurus kansensis, are quite
different from the Interval 300 quarry, but Peabody (1957:949) points out
that the holotype of E. ecordi was found near the bottom of the lagoonal
deposit, “in heavily carbonaceous shale containing a more typical, coalswamp
flora than higher and off-shore deposits in the lagoon.”

The new specimens of Desmatodon hesperis and the holotype of the
new microsaur are described in detailed fashion below. There remain, of
course, many skeletal elements that defy taxonomic identification at this time,
but they do indicate the presence of a diverse fauna and hold out the promise
of further interesting finds.

I have already shown that the vertebrates previously reported indicate
Late Pennsylvanian, probably Missourian, age (Vaughn, 1969). This is based
largely on the essential similarity of Desmatodon hesperis to D. hollandi
known from the middle part of the Conemaugh Group west of Pittsburgh,
Pennsylvania. The association of remains of Edaphosaurus raymondi with

1972

New Vertebrates

7

the holotype of D. hoUandi supports this estimate of approximate time-
equivalence. The primitive nature of the aistopod from Interval 300, Colora-
derpeton brilli, may perhaps also be cited as evidence for Pennsylvanian,
rather than Permian, age.

There is nothing among the recently found vertebrate materials from
Interval 300 that would contradict Missourian age; rather, Edaphosaurus cf.
E. ecordi provides corroboration. Sturgeon and Hoare (1968) place the Stan-
ton Formation, from which comes the holotype of E. ecordi , at about the same
level as the Ames Limestone of the Conemaugh Group. The Round Knob
Formation, from which comes the holotype of Desmatodon holiandi , lies
shortly below the Ames. The Stanton Formation is near the top of the Mis-
sourian Stage and thus a closer, but not necessarily more accurate, estimate
of the age of Interval 300 might be late Missourian, Nor does an estimate of
Missourian age conflict with any other data— excepting perhaps the bisaccate
gymnosperm pollen which, as I have said, may merely reflect environmental
difference; according to Brill (1952), the Sangre de Cristo Formation in the
area of Interval 300 rests conformably on rocks of Desmoinesian (Middle
Pennsylvanian) age.

A New Microsaur

A well-preserved, articulated skeleton of a hitherto unknown kind of
microsaurian amphibian is remarkable in a number of features that make it
important in consideration of the origin and relationships of the Microsauria.
Its distinctness from all previously known microsaurs requires the naming of
a new family.

Order MICROSAURIA Dawson, 1863
TRIHECATONTIDAE, new family

This family is based on the new genus Trihecaton. described below.
Because this is the only known genus, definition of the family is tentative:
microsaurs with infolded enamel on the marginal teeth, and with large pre-
sacra i intercentra with capitular facets for the ribs.

Trihecaton , new genus

Type species: Trihecaton howardinus , new species.

Diagnosis: Marginal teeth simple cones with shallow infolding of enamel.
Prominent coronoid process on lower jaw. Thirty-six presacral vertebrae— as
nearly as can be determined. First vertebra of characteristic microsaurian
structure, with forward-facing articular facet on either side of short “odon-
toid” process on anterior face of centrum, and deep notochordal pit on
posterior face; two costal facets on either side of first centrum; first neural
arch incomplete dorsally. Pleurocentra the dominant central elements, but
large intercentra, with capitular facets, in presacral vertebral column. Inter-
central haemal arches in tail. Almost regular alternation in shape of presacral
neural spines. Presacral ribs articulate with transverse processes low on

8

Contributions in Science

No. 223

anterior parts of neural arches and also with intercentra. Stem of interclavicle
very short, with abrupt termination. Well-developed limbs. Entepicondylar
foramen in humerus. Scales of characteristic microsaurian pattern, with
radiate striae.

Etymology: Named for the Interval 300 quarry. From the Greek treis,
three, and hekaton, hundred.

Trihecaton howardinus, new species

Holotype: UCLA VP 1743 (Figs. 1; 2; 3, A-C), an articulated skeleton
that lacks most of the skull although a maxilla and a mandible are present.
The entire presacral vertebral column and most of the ribs are preserved.
There are large parts of the left pectoral girdle and the interclavicle, and most
of the left front limb, as well as parts of the right front limb. The pelvic girdle
is poorly preserved, but the left femur is complete, and the distal half of the
right femur is also preserved. Scales are present throughout the region of the
vertebral column.

Referred specimen: UCLA VP 1744 (Fig. 3, D), a partially articulated
series of twelve caudal vertebrae with haemal arches. Same kind of scales
present as in holotype. Found immediately adjacent to the plastered-out block
of matrix that contained the holotype, and probably part of the same
individual.

Horizon and locality: Collected by a field party from the University of
California, Los Angeles, in the summer of 1970, from Interval 300 of the
section of the Sangre de Cristo Formation measured by Brill (1952:870),
about 1450 feet above the base of the formation, in NW!4 NE14 SWV4 sec.
22, T 49 N, R 10 E, Fremont County, Colorado. The age is Late Pennsyl-
vanian, probably Missourian. In European terms, the horizon would be
within the lower part of the Stephanian Series. The species is named after
the nearby town of Howard.

Diagnosis: The same as for the genus as this is the only known species,
but a note on size may be appropriate. The length of the skull, as known
from the lower jaw, is about 2.5 cm, and the length of the presacral vertebral
column is about 16 cm.

Description: Most of the features of the holotype and the referred speci-
men can be readily seen in the illustrations (Figs. 1, 2, 3). The anterior quar-
ter of the presacral vertebral column is bent to the right and backward. The
girdles and limbs are in their proper places relative to the column, but the
maxilla, mandible and the few other preserved cranial fragments are dis-
placed and lie adjacent to the interclavicle.

The left maxilla lies directly behind the interclavicle. It seems to be nearly

Figure 1. Trihecaton howardinus, new genus and species: photograph of the
holotypic specimen, UCLA VP 1743. Dusted with white powder to bring out details.
White material surrounding the matrix is plaster. Smallest divisions on the scale are
millimeters.

1972

New Vertebrates

9

, x

10

Contributions in Science

No. 223

complete except for the lack of a small posteriormost part. As preserved, it is
14 mm long; its greatest depth, about 2 mm, occurs about a third way back
in its length, and posterior to this it tapers gradually to a depth of less than
0.5 mm. Nineteen slender, conical teeth are present, and there are spaces for
three more. The anteriormost two are very slightly recurved, but the rest are
straight. The longest, about 1.8 mm, are in the region of greatest depth of the
maxilla; anterior to this they become slightly shorter, and posteriorly they

Figure 2. Trihecaton howardinus, new genus and species: details of holotypic speci-
men, UCLA VP 1743. A, anterior part of holotype, with mandible in lateral view,
elements of pectoral girdle in internal view, and vertebrae in various views — first
vertebra seen from behind; B, mandibular teeth at greater magnification. Abbre-
viations: a, articular region of mandible; c, coracoid plate; h, left humerus; ic, inter-
centrum following vertebra 6; icl, interclavicle; mx, maxilla; r, left radius; tp,
transverse process of vertebra 7; u, right ulna; v. 1, first vertebra. Size indicated by
the 1 cm scales.

1972

New Vertebrates

11

decrease gradually to very small size— the exact length of the posteriormost
teeth cannot be determined, due to breakage. A number of the teeth clearly
show infolding of the enamel, expressed externally by fine grooves that are
especially obvious in the basal portions; that the infolding is shallow is evident
from the fact that those teeth that were worn, either prior to or during prep-
aration, lack the grooves.

Almost all of the left mandible is preserved, exposed in lateral view.
Although the posterior portion of the surangular, to which is attached an
articular bone that projects mediad as a long spur, is broken and slightly
displaced from the rest of this element, the clean fracture premits confident
measurement of the total length of the mandible, 26 mm; this provides, of
course, an excellent index to length of skull. The lateral surface of the mandi-
ble, especially the dentary, is irregularly pockmarked by small, subcircular
pits; this is also true of the maxilla. The mandible is slender in most of its
length, with a depth of only 3 mm midway along the tooth row, but posteriorly
it rises in a prominent coronoid process, 6 mm deep, in which region the
dentary overlaps the anterior part of the surangular and is capped by a con-
spicuous coronoid bone. The dentary meets a long splenial along the ventral
border of the mandible. The angular bone is badly fractured, but it seems
to have extended about as far posteriorly as did the surangular. Nine teeth
are preserved in their entirety, stumps of three others are present, and there
are spaces for about seven more. As in the maxilla, the teeth are slender cones,
with the more anterior ones slightly recurved. Five of the teeth clearly show
the grooves related to the infolding of the enamel; these grooves are more

Figure 3. Trihecaton howardinus, new genus and species: details of holotypic
specimen, UCLA VP 1743, and referred specimen, UCLA VP 1744. A, vertebrae
17-21 of holotype in dorsal view, and associated ribs; B, first vertebra of holotype
in anterior view, some matrix left adherent to permit reunion with rest of specimen;
C, fragmentary scales taken from near vertebra 17 of holotype; D, caudals from
posterior portion of vertebral string of referred specimen. Size of A, B and D
indicated by the 1 cm scale; size of C indicated by the 2 mm scale.

12

Contributions in Science

No. 223

prominent at the bases but extend about three-quarters of the way up the teeth.
In the rest of the teeth, which are worn, the grooves are not apparent. As in
the maxilla, the longest teeth, about 1.8 mm, lie a short distance back in the
jaw; anteriorly and posteriorly the teeth become shorter, with the shortest
lying just in front of the root of the coronoid process. The anteriormost teeth
are preserved only as stumps, and their exact lengths cannot be ascertained.

A few scraps of other cranial elements lie near the maxilla and mandible,
but they cannot be identified with confidence.

As nearly as can be determined, there are 36 presacral vertebrae, in a
column that was about 16 cm long. What I count as the last presacral has
been pushed upward to partially override the poorly preserved, but obviously
more massive, remains of what I judge to be a sacral, closely appressed to the
pelvic girdle. The first vertebra is displaced from the second by several milli-
meters, but there is no evidence of an intervening vertebra. It is conceivable
that there may have been 37 presacrals, but I am fairly confident that 36 is the
correct number. There seems to have been only one sacral, but this cannot be
stated positively. Scraps of caudals are present in the holotype, but the nature
of the caudals is evident only in the referred specimen, and their number is
unknown. Slight rotation and displacement of various vertebrae in the anterior
part of the column display some in lateral, some in ventral, and some in dorsal
view. Farther posteriorly, most of the vertebrae are seen in dorsal view, but
rotation of vertebra 14 and loss of the neural arches of 15 and 16 allow inter-
centra to be seen in that region— as well as in the anteriormost parts of the
column. These circumstances are fortunate, because the highly fissured nature
of the matrix would make it hazardous to attempt to display the column from
below.

Wherever in the column a series of well-preserved vertebrae is exposed
in dorsal view, there can be seen an almost regular alternation in the shape
of the neural spines. This is especially clear in vertebrae 17-21 and 26-35. The
alternation is not quite as marked as in Pantylus (Carroll, 1968) but is pro-
nounced nevertheless. In the “low” type (e.g., vertebrae 27 and 29), the spine
runs the entire length of the neural arch as a low, rounded ridge. In the “high”
type (e.g., vertebrae 28 and 30), the low ridge occupies only the anterior half
of the arch; posteriorly, the spine broadens abruptly, rises slightly higher, and
sends a short process forward. In the more anterior portions of the column
this process is generally single, but posteriorly it is divided into a side-by-side
pair of processes. That the alternation is not completely regular is shown by
vertebrae 17-21, where the sequence is low-high-low-high-high. The signif-
icance of this alternation, which occurs also in other late Paleozoic tetrapods,
is obscure. I have attempted an analysis of a similar pattern in the Early Per-
mian reptile Captorhinus in terms of the system of interspinous ligaments, with
particular regard to an “extended nuchal ligament” (Vaughn, 1970), but the
presence of this pattern in the long-trunked Trihecaton casts doubt on such
an explanation.

1972

New Vertebrates

13

The first vertebra fits so well into the general microsaurian pattern (see
Carroll and Baird, 1968) that only its salient features need description. Its
length, from the tip of the “odontoid” process to the posterior edge of the
centrum, is 4 mm. The “odontoid” process, on the anterior face of the cen-
trum, is short, is rounded on its anteroventral surface, and is channeled by
the neural canal on its dorsal surface. On either side of this process is a large,
subcircular, slightly concave articular facet for the occiput; these facets face
directly anteriad. The distance between the tips of the winglike buttresses for
these facets is 7.5 mm. Each of these wings has on its posterolateral surface
two small, slightly raised and centrally dimpled facets, one above the other;
these are obviously for the first rib. The posterior face of the centrum, with a
deep notochordal pit, has a width of only 3.5 mm. The greatest height of the
centrum is 4 mm; the highest points of the neural arch are 8 mm above the
bottom of the centrum. The lateral halves of the neural arch almost meet at a
point anteriorly, but posteriorly they diverge widely. A short, blunt spine rises
above each posterior zygapophysis. The lack of junction of the halves of the
neural arch is not unique; Miss Eleanor Daly (personal communication, 1971 )
has shown me a similar condition in certain microsaurs from the Lower Per-
mian of Oklahoma.

The shape of the centra (pleurocentra) can be easily made out in several
of the anterior vertebrae and also in the region of vertebrae 14-16. In general,
the centra are like those of the microsaur Tuditanus (see Peabody, 1959; Car-
roll and Baird, 1968). They are notochordal, are excavated on their lateral
surfaces in such a way that they appear pinched, and a rounded ridge forms
the concave ventral border. They are different from those of Tuditanus in
that the ventral portions of both the anterior and posterior lips of the centra
are bevelled for the reception of the large intercentra. The centra that can be
measured are each somewhat more than 4 mm long, and are about 5 mm wide
and 3.5 mm high at the posterior end. Vertebrae 15 and 16 show that the floor
of the neural canal paralleled the hourglass-shaped notochordal canal.

The neural arches are not swollen. They are firmly joined to the centra,
but the lines of the neurocentral sutures are evident as curved ridges. The
neural spines have already been described. The zygapophyses are oval with
long anteroposterior axes, with their articular surfaces in the horizontal plane.
Stout transverse processes about 1.5 mm long occur near the junctions of the
neural arches and centra on all the presacral vertebrae excepting the first, and
possibly the second, which is poorly preserved. They are kidney-shaped in
cross-section, concave side to the rear, with the upper end lying approximately
midway in the length of the vertebra and the lower end near the front edge
of the centrum, apparently in contact with the apex of the intercentrum. The
costal facets of the transverse processes face forward, laterally and ventrally.

An intercentral scrap is visible in front of the second vertebra, large parts
of the intercentra are exposed in front of vertebrae 3 and 6, and a complete
one lies in apparently its life position against the posterior rim of the centrum

14

Contributions in Science

No. 223

of vertebra 6. Parts of intercentra may also be seen in front of vertebrae 14,
15 and 17. The intercentrum in front of vertebra 14 seems to have been as
large as the one behind vertebra 6; this, plus the presence of large haemal
arches in the caudal series, makes it seem likely that intercentra were present
throughout the entire presacral column.

As shown by the one behind vertebra 6, the intercentrum is a large
crescent that extends halfway up along the bevelled lip of the centrum. The
ventral portion is much swollen toward the sides, almost bulbous, and has
an anteroposterior length of 2 mm, almost half as long as the centrum. On
either side, the wedge-shaped ascending part of the intercentrum bears a large,
concave articular facet that faces posteriorly as much as it does laterally.
This ascending part is abruptly truncated where it meets— actually lies against
—the ventral end of the transverse process. This proximity of capitular and
tubercular facets is consonant with the structure of the ribs described below.
The intercentrum in front of vertebra 14 also shows the capitular facet.

The first several ribs are not preserved, but the costal facets on the first
vertebra show that ribs must have been present in this region. Indeed, the
well-developed transverse processes back to and on the last presacral, plus
ribs either articulated or in close association with almost all the vertebrae,
show that there were ribs throughout the presacral column. The ribs articu-
lated on the left sides of vertebrae 18 and 19 have gradually tapered shafts
about 14 mm long. Ribs farther forward in the column have expanded distal
ends. The structure of the head is best seen in partial ribs lying to the right of
vertebrae 17 and 19. The dorsoventral length of the head is about 4 mm.
The strongly convex capitulum is separated by only 0.5 mm from the slightly
concave tuberculum; the connecting web is much thinner than either the
capitulum or tuberculum, but its margin is not incised. Distal to the head the
shaft narrows rapidly. Wherever else the heads of ribs can be seen, including
the anterior part of the column, the pattern is similar, and what can be seen
of the ribs immediately anterior to the pelvic girdle indicates that articulation
was with both transverse process and intercentrum throughout the presacral
column, excepting of course the first rib.

The referred specimen contains parts of twelve caudal vertebrae. Two
of these, in articulation but displaced from the rest, have transverse processes
that are fairly well developed but without the ventrad prolongations seen in
the presacral vertebrae. Intercentral elements are not attached to these two
vertebrae, but the bilaterally bevelled surfaces of the posteroventral parts of
the centra indicate that haemal arches were present. The other ten vertebrae,
of which the most anterior is represented by only a fragmentary haemal arch,
are in one string, but two vertebrae midway in this string have undergone
rotation in the vertical plane. This displacement of parts is in accord with the
jumbled nature of the pelvic bones of the holotype. A vertebra toward the
anterior end of the string has a small transverse process, but only small nub-
bins are present in the same positions on the more posterior vertebrae. The

1972

New Vertebrates

15

neural spines are thin, short, and inclined somewhat posteriad; there is no
sign of alternation in shape. Posteroventrally directed haemal arches are
present throughout the string; they are clearly separate from the centra, be-
tween which their proximal ends are wedged. A well-preserved haemal arch
toward the posterior end of the string is 7 mm long. Part way along the length
of each haemal arch, coinciding with the ventral limit of the haemal canal,
there is a bulbous swelling; distal to this the arch tapers, ending in a slight
dilatation in the more anterior ones, in a point in the more posterior ones.
The swollen portion comes to lie more proximally farther back in the string.
A bridge of bone completes the haemal canal dorsally. As previously men-
tioned, the referred specimen also includes scales identical to those of the
holotype. There is also a metatarsal or phalangeal element.

Although the pectoral girdle of the holotype is not entirely preserved,
it is clear that in most features it resembles that of Parity lus (see Carroll,
1968). A fragmentary element on the left side of the interclavicle may be
part of a cleithrum or clavicle, but this is not certain; an adjacent element
seems to be a displaced anterior rib with expanded distal end. The interclavicle
is seen in internal view. It is T-shaped, with an expanded bowl that must have
been about 15 mm in width but, quite unlike Pantylus, the stem is very narrow
and remarkably short, only about 1 mm. This shortness does not appear to be
due to fracture; at its termination the stem dilates slightly and ends abruptly
in a rugose surface. It would almost seem that it was continued in cartilage,
but this is highly unlikely in a dermal element. The coracoid plate of the left
scapulocoracoid, which lies in the region of vertebrae 4-6, is exposed in inter-
nal view. Most of the scapular blade is missing, but it is clear that, as in
Pantylus , it joined the coracoid plate anterior to the glenoid fossa. Also as in
Pantylus, the glenoid fossa is buttressed anteriorly by a ventrad projection
along which the articular surface is continued, and the part of the coracoid
plate posterior to the fossa is short and narrow. There is a coracoid foramen
medial to the junction of the scapular blade and coracoid plate. The greatest
anteroposterior length of the scapulocoracoid seems to have been about 13
mm.

Only the proximal portion of the right humerus is preserved, but the left
is complete and is nearly in articulation with the glenoid fossa. As in both
Tuditanus (Carroll and Baird, 1968) and Pantylus, there is an entepicondylar
foramen about three-fifths of the way from the head to the distal end of the
humerus, but the overall resemblance is closer to Tuditanus, in which the
entepicondyle is more widely flared than the ectepicondyle. In Pantylus
(Carroll, 1968, fig. 5), the capitellum lies on a prominently produced ectepi-
condyle, quite unlike the general tetrapod pattern. The distal and proximal
planes of the humerus are twisted at almost a right angle to one another as in
Pantylus (this feature is obscure in Tuditanus ), but the shaft is somewhat
more distinct. The articular ends are rugose, but there is no indication of
immaturity; all the usual processes of the head are present, and the capitellum

16

Contributions in Science

No. 223

is a well-developed hemisphere. The humerus is 15 mm long, 6 mm wide
across the head, slightly wider across the distal end, and has a minimum
thickness of about 2 mm in the shaft.

Only the left radius is present; this has a semicircular head, narrow shaft
—which is broken— and dilated distal end, much as in Pantylus. Only the right
ulna is complete. It is 10 mm long, has a prominent olecranon, and is much
narrower distally than proximally. There are closely associated but disarranged
carpal elements and two metacarpals near the epipodials on the left side, but
nothing can be confidently said of their original organization, nor of the num-
ber of digits.

The broken bones of the pelvic girdle overlap one another; while it is
possible to recognize the individual elements, distinctive features cannot be
discerned. The femur of the left side is complete, and the distal half of the
right one is also present. The femur is gently sigmoidal, about 17 mm long,
with a moderately developed adductor ridge and well-separated tibial condyles.
The shaft has a diameter of about 2 mm in its narrowest portion; the distance
across the distal condyles is 7.3 mm. A proximal fragment of an epipodial is
present, but there is nothing more distal. The nature of the remains and the
matrix demands “blind” removal of most of the blocks at the quarry, and in
this process the holotype was truncated just behind the pelvis, probably
separated from the referred specimen as mentioned earlier.

Imbricated scales may be seen alongside almost all the skeletal elements.
Their fragility makes them extremely difficult to remove and prepare satis-
factorily, but it has been possible to study the surface details of a number
of them. They resemble very closely those of Pennsylvanian microsaurs illus-
trated by Carroll and Baird (1968, fig. 20). They are oblong, have a ridge
along one border, and have radiate striae. The striae are especially well dis-
played on some of the scales of the referred specimen. A scale taken from
the region of vertebra 17 of the holotype is about 1 mm wide by 3 mm long.

It is possible that other parts already collected from Interval 300 will
eventually prove to be referable to Trihecaton howardinus. A specimen that
consists mostly of finely denticulate palatal elements but also includes a partial
maxilla with small, conical teeth (UCLA VP 1698) was identified in my
earlier paper as a labyrinthodont amphibian, because of the infolded enamel
(Vaughn, 1969). It may be of T. howardinus , but the specimen does not
include the characteristic scales that would make this clear. Further excava-
tion at Interval 300 is planned, and hopefully, better cranial materials may
soon be found.

Discussion: Trihecaton must be regarded as a microsaur. Among Paleo-
zoic tetrapods, the kind of first vertebra seen in T. howardinus is found only
among the lepospondyl amphibians; and the combination of pleurocentral-
intercentral construction of the vertebrae, well-developed limbs, and charac-
teristically microsaurian scales— plus the fact that the neural arches are
not, with the exception of the first vertebra, divided dorsally— rules out all

1972

New Vertebrates

17

lepospondyls but microsaurs. It is also clear that Trihecaton differs greatly
from all previously described microsaurs. No others have infolded enamel
(Carroll and Baird, 1968), although the bases of the teeth in Trachystegos
do show a coarse fluting (Carroll, 1966). Although haemal arches have been
found in several microsaurs including Pantylus (Carroll, 1968; Carroll and
Baird, 1968), and small presacral intercentra have been reported in Micro-
brachis (Brough and Brough, 1967), none of the known microsaurs ap-
proaches Trihecaton in its large presacral intercentra with capitular facets
for the ribs; and it must be remembered that Trihecaton also has well-
developed haemal arches. The recently described lepospondyl Acherontiscus
(Carroll, 1969) does have large intercentra, but the vertebral construction
is so similar to that of the embolomerous labyrinthodonts that Carroll
hesitated to assign Acherontiscus to any recognized order. The enamel in
Acherontiscus is not infolded. It is difficult to escape the impression that
Trihecaton is very primitive, despite its occurrence fairly high within the
stratigraphic range of known microsaurs.

Two groups usually included in the Microsauria, the Adelogyrinidae and
the Lysorophidae, are now considered by Carroll and Baird (1968) to be
somewhat separate stocks; this is also the view of Thomson and Bossy (1970),
who distinguish these two families from what they call “eumicrosaurs,” a
convenient term for present purposes. There does seem, however, to be
general agreement that the adelogyrinids, lysorophids and eumicrosaurs are
much more closely related to one another than any of these are to the re-
maining lepospondyls of traditional classification, the Nectridea and Aisto-
poda. Thomson and Bossy are of the opinion that the trend toward a
holospondylous vertebral condition seen among the lepospondyls is not a
reliable indicator of relationship but is, instead, a parallel tendency somehow
correlated with small size. They conclude that the “Lepospondyli” are not
a natural assemblage and suggest that the term be abandoned; with this I
heartily concur.

Trihecaton would seem to be allied with the eumicrosaurs. The lysoro-
phids are so different that they need hardly be considered, and no evidence
of pleurocentral-intercentral construction of adelogyrinid vertebrae has ever
been presented. Nevertheless, Trihecaton may represent something of a bridge
between adelogyrinids and eumicrosaurs. Thomson and Bossy point out that
the adelogyrinid jaw system was probably of the kinetic-inertial kind (see
Olson, 1961), whereas the small coronoid process and other cranial features
in the eumicrosaurs indicate a predominantly static-pressure system. Tri-
hecaton has a prominent coronoid process. It is unfortunate that more of the
skull is not known in Trihecaton , but in this seemingly annectent feature
we may have reason to regard this genus as primitive.

It is certainly unwise to assume that all amphibians were derived from
the Late Devonian ichthyostegalians — the variety of Mississippian forms is
too great— but perhaps we may assume, in conservative style and for the time

18

Contributions in Science

No. 223

being, that the microsaurs and the labyrinthodont amphibians did have a
common origin. If we do, we may perhaps regard the infolded enamel and
the large intercentra in Trihecaton as additional marks of primitiveness. This
suggestion invites dispute. Thomson and Bossy (1970) give reason to believe
that such characters of tooth and vertebral structure may not be the reliable
indices to relationship that they were once thought to be. However, their
argument that infolding of the enamel is merely a function of large tooth
size is considerably weakened by the occurrence of this character in such
a small form as Trihecaton, even granted that the infolding is shallow. With
regard to the vertebrae, Thomson and Bossy agree with Panchen (1967) that
the structure of the centrum is extremely plastic. Differential composition
of the centrum-multipartite versus unitary, differential emphasis on the parts
when multipartite — may merely reflect different responses to problems of
support and locomotion. Carroll (1969) regards the “embolomerous” con-
struction of the vertebrae in Acherontiscus as probably associated with
lengthening of the segments to assist in sinuous swimming movements. As
Thomson and Bossy say (1970:14), “the combined centrum (divided or
whole) is homologous in all amphibians,” that is, not too much stress should
be laid on the pleurocentrum and intercentrum in terms of strict homology.
Despite these cautions, it does still seem that a combination of features in
Trihecaton — infolding of the enamel, large presacral intercentra, prominent
coronoid process — give this genus an overall aspect of primitiveness.

Trihecaton, at a Stephanian horizon, is of course too late in time to be
considered as an actual “urmicrosaur.” Adelogyrinids are known from the
Lower Carboniferous (see Carroll, 1967); the problematic Acherontiscus,
with a skull similar to that of eumicrosaurs, probably occurs in the lowest
Upper Carboniferous (Carroll, 1969); and eumicrosaurs are known as low
as the Westphalian B level, in the Joggins Formation of Nova Scotia (Carroll,
1966). Perhaps we may regard Trihecaton as a relict.

Of the eumicrosaurs, Carroll and Baird (1968) tentatively recognize
the families Gymarthridae, Tuditanidae, Hyloplesionidae and Microbrachidae;
and Carroll (1968) suggests that we consider the pantylids as a separate
family. It is difficult to select from among these the family closest to the
Trihecatontidae. For example, Trihecaton resembles the microbrachids in
having a large number of presacral vertebrae (38-40 in Microbrachis) and
in the presence of presacral intercentra (see Brough and Brough, 1967); but
in its well-developed limbs and in the presence of an entepicondylar foramen
in the humerus, Trihecaton resembles the shorter-trunked tuditanids (29 pre-
sacrals in Tuditanus ) and pantylids (24 presacrals in Pantylus), with the
resemblance to Pantylus extending to details of the pectoral girdle (see Carroll
and Baird, 1968; Carroll, 1968). Aside, it may be noted that a long trunk
in microsaurs is usually thought to be a sign of aquatic habits, and the feeble
limbs in Microbrachis do seem to corroborate this, but Trihecaton has fairly
sturdy limbs. Because of the incomplete nature of most microsaurian remains,

1972

New Vertebrates

19

there is at present little ground for decision as to whether an elongate column
is primitive or advanced in this group. It may be of functional significance
that the two microsaurs known to have presacral intercentra, Trihecaton
and Microbrachis, also have long trunks. Perhaps the larger intercentra in
Trihecaton are correlated with the sturdier limbs; Parrington (1967) has
shown how the large intercentra in rhachitomous labyrinthodonts helped
strengthen the column in terrestrial locomotion while still allowing flexibility,
but rhachitomes are short-trunked. The alternation in shape of the neural
spines may at first glance seem to be a special resemblance to Pantylus (see
Carroll, 1968), but this phenomenon was apparently of widespread occur-
rence among late Paleozoic tetrapods and may be seen also in the Penn-
sylvanian lysorophid Molgophis (Dr. Donald Baird, personal communication,
1971) and even in the Early Permian reptile Captorhinus (Vaughn, 1970).
Trihecaton stands alone among known microsaurs in its infolded enamel and
its large presacral intercentra with capitular facets. It seems best, for now,
not to pursue the placement of the Trihecatontidae within the Microsauria,
but to leave this for later studies against, hopefully, a larger background of
known forms.

Present knowledge of Trihecaton does not help solve the question of
possible affinities of the microsaurs and the captorhinomorph reptiles. Romer
(1969), in a recent and thorough study of the cranial anatomy of the Early
Permian Pantylus , has demonstrated that the braincase and branchial arches of
this form have a definitely amphibian cast, and comes to the conclusion that
microsaurs and captorhinomorphs cannot have any antecedent-descendent
relationship. Nevertheless, distinctions between microsaurs and captorhino-
morphs continue to disappear; for examples, Carroll and Baird (1968) have
demonstrated that the microsaurian first vertebra is a compound of elements
probably homologous to the reptilian atlas and axis, and knowledge of Tri-
hecaton adds to the growing list of microsaurs that have pleurocentral-
intercentral vertebrae. Possibly this points to some remote community of
origin of captorhinomorphs and microsaurs, but this is far from clear.

New Information on Desmatodon hes peris

The genus Desmatodon was based by Case (1908) on Desmatodon
hollandi, the holotype of which is a fragment of maxilla with four complete
teeth and the root of a fifth (CM 1938, Carnegie Museum, Pittsburgh); the
holotype was collected from the Round Knob Formation, Conemaugh Group,
western Pennsylvania. I have named the species Desmatodon hesperis on the
basis of a complete left maxilla with twelve teeth (UCLA VP 1706) taken
from the Interval 300 quarry, and to this species I have already referred other
materials: more teeth including “incisors,” vertebrae, a humerus, and other
elements (Vaughn, 1969). I have also pointed out that, while D. hesperis
and D. hollandi are essentially similar as far as comparisons can be made,
the tendency toward more conical shape of the teeth in D. hesperis gives this

20

Contributions in Science

No. 223

species a somewhat more primitive aspect. More materials referable to
D. hesperis are now on hand from Interval 300, and these provide welcome
additional information on the nature of Desmatodon, the only known Penn-
sylvanian member of that odd group of primitive reptiles — or seymouriamorph
labyrinthodonts — called the diadectids.

UCLA VP 1745 (Fig. 4, A-C) consists of a braincase and firmly joined
fragments of the posterior elements of the dermal skull roof. It was found
on a plane that yielded various closely associated parts referable to Desma-
todon hesperis, including teeth and vertebrae, but even without this association
the diadectid nature of the specimen would be immediately obvious. It re-
sembles in almost all ways the corresponding portion of the skull in the Early
Permian Diadectes, and it may be safely assumed that the specimen is of
Desmatodon hesperis. It has been distorted in such a way that the angle
between the occipital surface and the dermal roof has been decreased, and
parts of the braincase on the left side have been moved forward to leave
a large gap between the ventral portions of the prootic and opisthotic. This
distortion makes it difficult to take accurate measurements, but the original
distance between the facets for the quadrate bones may be estimated at about
43 mm, and the distance between the lateralmost points of the paroccipital
processes was about 48 mm. The specimen is very probably of an immature
individual, to judge from the almost complete lack of ossification in the region
of the otic labyrinth and by its close association with the juvenile maxilla
described below. Because of the essential resemblance to Diadectes, for which
excellent illustrations are available (Olson, 1947, 1966; Watson, 1954),
description may be limited to salient features.

As in Diadectes, the opisthotic, supraoccipital, postparietal and tabular
bones are indistinguishably fused; but the exoccipitals are separate, and the
line of junction between the prootic and opisthotic can be easily discerned on
either side. It is impossible to trace the suture between the parietal and post-
parietal bones with any confidence although it does seem clear that, again
as in Diadectes, the postparietal has both occipital and roofing components.
A symmetrically curved line across the postparietal area separates the rugose
surface of the roof from the smoother occipital surface. A median ridge
running ventrad from the postparietal, and a transverse ridge across the
supraoccipital region mark out areas for attachment of occipital muscles.
The parietal foramen is enormous, with almost the same diameter as the
foramen magnum (about 12 mm), although distortion in both regions makes
this only a rough comparison; the proportions in Diadectes are similar. The
suture between the parietals can be followed for a short distance behind the
foramen. About 1.5 cm to the left of the parietal foramen there is a deep
longitudinal groove on the dorsal surface; this is seen also in certain Diadectes
specimens (Watson, 1954, fig. 22), in which the groove partially demarcates
the lateral parietal lappet from the main body of the parietal. A faint suture
can be traced along the floor of the posterior part of this groove; this supports

1972

New Vertebrates

21

Olson’s (1950) arguments that an intertemporal bone was present in dia-
dectids and that the parietal lappet is formed by this element. The nature
of the groove, showing incomplete ossification between two dermal elements,
is in accord with the pattern in Diadectes, of which Olson (1950:63) says,
“Fusion of adjacent elements in the occipital and temporal regions appears
to have been initiated at an early stage and to have progressed from the
inner to the outer surface of the skull”; and it is also additional evidence
of the immaturity of the specimen. Fractures and displacements in the region
of the supratemporal make it impossible to delimit this bone satisfactorily,
but an undulating longitudinal furrow probably marks its contact with the
tabular.

One of the most interesting features of the specimen is the presence,
on either side, of a fenestra between the postparietal and tabular regions.
The thin, finished borders leave no doubt that these openings actually existed.
The bone along the lateral border of the right fenestra is lacking, but the
left fenestra is completely bounded. It is about 14 mm long; toward its pos-
terior end it is about 8 mm wide as preserved but was probably narrower in
life — the fragment of the tabular in this region is displaced slightly to the
left, overriding the most posterior part of the facet for the supratemporal.
Desmatodon hesperis is not really different from Diadectes in this feature.
Although the fenestrae are rarely seen in Diadectes, they do occur in certain
specimens (Case, 1911; Huene, 1913). Olson (1947) has commented on
these openings in Diadectes and has pointed out that they do not lie in the
position of “normal” reptilian temporal fenestrae. In view of the sporadic
appearance of the fenestrae in Diadectes, and the indications that UCLA
VP 1745 is immature, it certainly would be unwise to consider the fenestrae
as diagnostic of Desmatodon hesperis, let alone the genus.

As in Diadectes, the supraoccipital sends a process under the postparietal,
terminating abruptly at a line that corresponds to the separation dorsally
between the roof and occipital surfaces. Despite the distortion that has driven
the left prootic forward, both prootics are well preserved and resemble the
same elements in Diadectes, extending upward and backward to their broad
junctions with the opisthotics. The foramina for the abducent nerves penetrate
the dorsum sellae, and a deep incisure that is more evident on the anterior
margin of the right prootic presumably marks the place of exit of the tri-
geminal nerve. Jutting forward from just above the dorsum sellae are bilater-
ally placed plates that must represent the pilae antoticae; it is not clear that
their lack of actual connection with the dorsum sefiae might not be due to
fracture and slight displacement, but they do seem to be separate elements,
and perhaps this is another sign of immaturity. Olson (1966) has illustrated
broad pleurosphenoids in this region in Diadectes, coosified with the dorsum
sellae. Above and partly behind the upper end of the prootic, there is on
the lateral surface of each opisthotic a deeply concave, elongate facet; this
is the socket for the quadrate bone. This unusual mode of articulation of

22

Contributions in Science

No. 223

quadrate and braincase is known also in Diadectes. This specimen helps show
that Olson (1966) has correctly interpreted the socket as occurring on the
opisthotic in Diadectes, not on the prootic as was stated by Watson (1954).

The parasphenoidal rostrum is incomplete, but enough is left to show
that it flared dorsad in front of the sella turcica. Only the left basipterygoid
process of the basisphenoid is preserved. This has smooth anterodorsal and

Figure 4. Desmatodon hesperis: A, dorsal, B, ventral, and C, left lateral views of
braincase and connected dermal roofing elements (UCLA VP 1745); D, upper part
of a right quadrate (UCLA VP 1746) in posterolateral view; E, lower part of
another, larger right quadrate (UCLA VP 1747) in posterior view. Abbreviations:
b, basipterygoid process; bs, basisphenoid; e, exoccipital; op, opisthotic; p, parietal;
pa, pila antotica; paf, parietal foramen; pi, parietal lappet ( - intertemporal? ) ;
pp, postparietal; pr, prootic; ps, parasphenoid; st, supratemporal; t, tabular. Because
of fusion in the temporal region, some of the abbreviations mark general areas
rather than distinctly demarcated elements. Line shading indicates matrix. Size
indicated by the 2 cm scale.

1972

New Vertebrates

23

anteroventral articular surfaces that are directed anteriorly and laterally.
It is obvious that the joint between braincase and palate was mobile, unlike
the condition in Diadectes where, despite an illustration by Watson (1954,
fig. 18) that may give the opposite impression, the basisphenoid and pterygoid
were firmly joined (see Olson, 1947). This feature could be interpreted as
a mark of primitiveness of Desmatodon hesperis, but it could also be argued
that it is merely another sign of immaturity of the specimen.

An upper part of a right quadrate bone (UCLA VP 1746, Fig. 4, D)
was found near the braincase, to which it is appropriate in size. It has a
well-defined condyle for articulation with the facet on the opisthotic bone.
A lower part of another right quadrate (UCLA VP 1747, Fig. 4, E) found
elsewhere in the quarry is obviously of a more mature individual, with a
width of 25 mm across the articular end. The articular surface is divided
into medial and lateral facets by a deep notch, and there is a large, rugose
tubercle on the posterior surface above the notch, indicating a stapedial
apparatus similar to that in Diadectes (see Olson, 1966). The resemblance
of these specimens to Diadectes is very close (see Watson, 1954, fig. 24),
and there can be little doubt that they are of Desmatodon hesperis.

A toothed right maxillary bone (UCLA VP 1748, Fig. 5, A, B) was
found only a few inches away from the above-described braincase. The
transversely widened, cusped teeth show that this specimen is of a diadectid,
surely Desmatodon hesperis. The thin anteriormost part, where it overlapped
the premaxilla, and the upper parts of the lateral wall have been broken away,
but it is otherwise complete. Except for size and for certain features of the
teeth, it is so similar to the holotypic maxilla that the description of the latter
(Vaughn, 1969) suffices for both. A small projection on the medial side
behind the last tooth is similar to one in the same position on the holotype,
making it clear that the specimen is complete posteriorly and that no teeth
have been lost. The new maxilla is 53 mm long as preserved, considerably
shorter than the 77 mm long holotypic maxilla, and it is apparent that it is
of a juvenile individual.

Whereas there are twelve teeth in the holotypic maxilla (Fig. 5, C), the
juvenile maxilla has only eight, which seem to correspond to the first eight
of the holotype. This correspondence is evident not only from the positions
of the teeth with respect to the anterior end of the maxilla, but also from the
fact that a narrow channel crossing the thickened medial surface of the maxilla
ends ventrally at the plane between the third and fourth teeth in both speci-
mens. Furthermore, the total length of the dental row in the juvenile is only
slightly less than the length occupied by the first eight teeth of the holotype,
42 mm and 47 mm respectively. The juvenile teeth are all much shorter
from front to back than the corresponding teeth of the holotype — for tooth
7 the respective dimensions are 2.6 mm and 5.2 mm — but they are separated
by much longer spaces. It is obvious that with growth, the teeth of D. hesperis
came to be replaced by more robust ones, and it is almost as obvious that

24

Contributions in Science

No. 223

the dental lamina must have become extended backward. These are common
phenomena that may be observed also in living lizards (Edmund, 1969), but
the elongation of the lamina in D. hesperis is striking when it is remembered
that the transition from the juvenile dentition to that represented in the holo-
type involved the addition of four teeth, one-third of what is presumably
the adult dentition. This is in sharp contrast to what is seen in the Early
Permian Diadectes. For example, an immature specimen of Diadectes san-

Figure 5. Desmatodon hesperis: A, lateral, and B, occlusal views of juvenile right
maxilla (UCLA VP 1748); C, lateral view of holotypic left maxilla (UCLA VP
1706). Line shading indicates matrix and, in A and B, epoxy resin used to repair
the specimen. Size indicated by the 2 cm scale.

1972

New Vertebrates

25

miguelensis has a total of eleven teeth and empty alveoli in a maxilla only
about 42 mm long (Lewis and Vaughn, 1965), and the number of maxillary
teeth is hardly greater in adults of the various species of Diadectes , ranging
from eleven to thirteen with eleven the usual number, as far as I have been
able to determine from specimens and the literature. A specimen of Diadectes
lentus at hand has a total of eleven teeth in a maxilla about 68 mm long.
It is interesting to note that Romer (1952) has described a battery of eight
small, diadectid-like cheek teeth contained in a length of only 1 1 mm, from
the Conemaugh Group, and that Langston (1963) has described six diadectid-
like teeth in a length of 10.5 mm, from the Lower Permian of Prince Edward
Island; these specimens are, however, very poorly preserved, and their sig-
nificance is difficult to assess.

Another notable contrast between the juvenile and holotypic dentitions
is the much greater relative length of the first two teeth in the juvenile. In
the holotype, the first two teeth are somewhat more acuminate than the
succeeding, and they jut out slightly beyond the general tooth row. The
difference is greatly accentuated in the juvenile, where the first two teeth
are markedly incisiform, bowed with concave lingual sides, and twice as
long as the succeeding teeth; the first and second teeth are each about 10 mm
long, but the third is only about 5 mm long. It must be noted that the base
of tooth 1 was lost during removal of the juvenile maxilla from the matrix,
but this part was cast in epoxy resin from the impression, and the restored
length is quite accurate. The teeth in the juvenile are, as might be expected,
narrower from side to side than the corresponding teeth of the holotype — for
tooth 7 the respective dimensions are 6 mm and 7.8 mm — but their lesser
anteroposterior length makes them more bladelike. Labial, central and lingual
cusps are present on teeth 5-8, but the lingual cusp is indistinct on teeth 3
and 4; all three cusps are distinct on teeth 3-8 of the holotype (actually, for
the labial “cusp” the word “shoulder” would be more appropriate, this
structure not being as well set off as in Diadectes) . There are no signs of attri-
tion on any of the teeth in the juvenile; this is remarkably different from the
holotype, in which there are distinct wear facets on most of the teeth.

A combination of ways in which the juvenile maxillary dentition differs
from that of the adult, much greater relative length and more incisiform
aspect of the first two teeth, smaller number of teeth, teeth more bladelike
and separated by longer spaces, lack of wear facets, fosters the suspicion
that the change from juvenile to adult may have included a shift in dietary
habit. Diadectes has been variously interpreted as herbivorous or mollusci-
vorous; small pelecypods frequently found in association would seem to
support the latter view, and it seems significant that small pelecypods are
also known alongside the remains of Desmatodon hesperis in the Interval
300 quarry. It is hard to guess as to the diet of young D. hesperis. It is, of
course, conceivable that the lack of wear facets on the juvenile teeth could
be a result of more rapid replacement— it is known in crocodilians, for

26

Contributions in Science

No. 223

example, that the replacement rhythm becomes slower with age (Edmund,
1969) —but this would not account for the other juvenile-adult differences.
Whatever further indications of a dietary shift may appear, it is at least
clear that the transition from juvenile to adult in Desmatodon hesperis in-
cluded much more pronounced dental changes than in Diadectes, in the im-
mature specimen of Diadectes sanmiguelensis previously mentioned there is
no great disparity between the first two maxillary teeth and those that follow.

Despite the above differences, the actual replacement of individual teeth
in Desmatodon hesperis seems to have occurred in essentially the same
manner as in Diadectes (see Edmund, 1960). A lingual pit was developed
alongside each tooth, the base of the tooth became eroded, and the replacing
tooth must have been ready for use soon after the overlying older tooth was
shed, as shown by the lack of empty alveoli in both the juvenile and holotypic
maxillae. Nevertheless, it will be shown that the cycle of tooth replacement
in Desmatodon hesperis was different from what is known in Diadectes.

A diagrammatic analysis of the tooth-replacement cycle in Desmatodon
hesperis, as inferred from the teeth of the holotypic maxilla, is presented in
Figure 6. On the basis of intactness or attrition of the crown, and the
degree of development of the lingual pit, five stages are arbitrarily de-
limited: newly erupted, not yet ankylosed (tooth 2); unworn but ankylosed
(teeth 10, 12); worn, small lingual pit (teeth 1, 3, 5, 7); worn, large
lingual pit (tooth 9); worn, base eroded (teeth 4, 6, 8, 11). If we assume
that a “gap” exists between teeth 2 and 3, it will be seen that the teeth

1 2 3 4 5 6 7 8 9 10 11 12

a, a - 4a

-•

O

• -•

Figure 6. Analysis of tooth-replacement cycle based on the holotypic maxilla of
Desmatodon hesperis (UCLA VP 1706). Filled-in and open symbols represent
members of the two alternating replacement series. The symbols also indicate
presence or absence of wear facets and the condition of the base, as: tooth 2, newly
erupted, not yet ankylosed; tooth 10, unworn but ankylosed; tooth 1, worn, small
lingual pit; tooth 9, worn, large lingual pit; tooth 4, worn, base eroded. The graph
represents the partial waves of replacement, with the lowest level corresponding to
newly erupted, not yet ankylosed, and the highest level corresponding to worn, base
eroded.

1972

New Vertebrates

27

can be arranged into two alternating replacement series, 1 -gap-4-6-8-1 0-1 2
and 2-3-5-7-9-11, and it will also be seen that the cephalad waves of replace-
ment overlap one another in an orderly manner, in accord with the usual
tetrapod pattern (see Edmund, 1960). In the graphic representation of the
waves, the five levels from lowest to highest correspond to the arbitrarily
delimited stages described above, in the same order. Only if we assume the
existence of the gap can the teeth be brought into such a pattern and, indeed,
there is evidence for the reality of this gap, as will be brought out below.
It is evident that the waves of replacement were long. Through graphic
analysis in “Edmundian” style, it can be shown that a complete replacement
wave included about seventeen teeth, from initial erosion of the old tooth
to destruction and loss of the new, from birth to death of a tooth so to
speak. This is considerably longer than replacement waves that have been
analyzed in Diadectes, where a complete wave apparently included only about
seven teeth (Edmund, 1960). The difference seems all the more profound
when it is appreciated that a complete replacement wave can be seen within
a jaw containing fourteen teeth in Diadectes, whereas it would take a jaw
of at least thirty-three teeth to show a complete wave in Desmatodon hesperis.

Because of damage suffered by the juvenile maxilla when it was dis-
covered through splitting of the enclosing matrix, most of the lingual pits
cannot be studied, but those alongside the first two teeth are plainly evident,
and these show that the second tooth has undergone some posteriorward
displacement; this is also indicated by fragmentation of the bone at the base
of the tooth. Thus it seems that there was originally an appreciable gap be-
tween the second and third teeth. In the holotypic maxilla there are short
diastemata between the first and second and the second and third teeth, in
marked contrast to the close packing in the rest of the row. However, a
remnant of the base of an older tooth at the second position shows that the
present tooth, newly erupted, and not yet ankylosed, must have moved
posteriorward as it came in, creating the diastema between itself and the
first tooth and reducing the gap between itself and the third. It is difficult
to say whether or not this gap, which “shows up” in the replacement pattern
as outlined above, was a constant feature in Desmatodon hesperis , but the
condition is reminiscent of that noted by Edmund (1969:134) in certain
of the teiid lizards where “Segments of the dentition may show regular
rhythms, but these are not continuous along the entire jaw. The explanation
may lie in the suppression of one or two adjacent tooth matrices to accommo-
date a single larger tooth.” In the holotypic maxilla of Desmatodon hesperis,
the disruption of regularity seems to be associated with development of the
two short diastemata.

The fragmentary holotypic maxilla of Desmatodon hollandi has only
four complete teeth and the root of a fifth; the represented stages are, from
anterior to posterior: unworn but ankylosed; worn, base eroded; worn, small
lingual pit; unworn but ankylosed; ?, large lingual pit. There are, of course,
too few teeth to permit a definite statement, but it would seem that the replace-

28

Contributions in Science

No. 223

ment wave in D. hollandi was much shorter, and thus nearer to the condition
in Diadectes , than in D. hes peris. As indicated in the initial description of
D. hesperis (Vaughn, 1969), this species seems on other bases to be some-
what more primitive than D. hollandi, and it is conceivable that the longer
replacement wave may present further grounds for differentiation, but nothing
is known of the ranges of variation within the two species.

Even if it may eventually become necessary to set up a new genus based
on D. hesperis, close relationship to D. hollandi would still be obvious; and
the cranial parts described above should allay any doubts as to the diadectid
affinities of Desmatodon. Although Desmatodon is more primitive in certain
features than Diadectes, the two genera are nevertheless essentially similar;
and it may be said definitely that the diadectid organization was well es-
tablished in the Late Pennsylvanian. The known Early Permian diadectids
form a closely knit group. Olson (1947) has summarized the various North
American species, recognizing two genera, Diadectes and Diasparactus.
Phanerosaurus and Stephanospondylus (probably synonyms) of the European
Lower Permian are basically like Diadectes but, like Desmatodon, appear
more primitive, having somewhat more acuminate teeth (Geinitz and Deich-
miiller, 1882, pi. 4) and, apparently, a mobile basipterygoid joint (Stappen-
beck, 1905, fig. 4).

The phylogenetic origin of the diadectids remains obscure although
Tseajaia from the Lower Permian of Utah does seem to provide a morpho-
logical link to seymouriamorph labyrinthodont amphibians (Vaughn, 1964).
The recently described Late Mississippian anthracosaur Mauchchunkia may,
as Hotton (1970) thinks, represent the ancestry of all “reptiliomorphs” includ-
ing diadectids, but intermediate forms are still unknown. Romer (1964) has
proposed that we consider Diadectes as a seymouriamorph rather than as a
cotylosaur. This is consonant with the recent suggestion by Carroll (1970)
that we exclude from definition as true reptiles all those forms that achieved
reptilian morphological characteristics independently of the line that passed
through romeriid captorhinomorphs; it would certainly be difficult to argue
for romeriid-diadectid affinities. Perhaps we should follow the advice of
Olson (1947) and think of diadectids as “parareptiles,” but in an informal
sense only, without intending the term as a taxonomic category.

Acknowledgments

I thank my hard-working field assistants, Messrs. Kerry Clegg, Marc
Gallup, Michael Novacek, and David Vaughn. Mr. Thurston Phetteplace,
of the Bureau of Land Management, helped smooth the way for quarrying
operations. Mr. Herbert Klug’s skill has been valuable in removal of the
difficult matrix. The drawings are the work of Mrs. Hermine Kavanau, and
the photograph of the microsaur was taken by Mr. Takeo Susuki. I am grate-
ful to the National Science Foundation for continued support of this study
through grant GB- 19971.

1972

New Vertebrates

29

Literature Cited

Axelrod, D. I. 1952. A theory of angiosperm evolution. Evolution 6:29-60.

Brill, K. G., Jr., 1952. Stratigraphy in the Permo-Pennsylvanian zeugogeosyncline
of Colorado and northern New Mexico. Geol. Soc. Amer., Bull. 63:809-880.

Brough, M. C., and J. Brough. 1967. Studies on early tetrapods. II. Microbrachis,
the type microsaur. Roy. Soc. London, Philos. Trans., ser. B, 252:131-146.

Carroll, R. L. 1964. Early evolution of the dissorophid amphibians. Harvard
Univ., Mus. Comp. Zool., Bull. 131:161-250.

1966. Microsaurs from the Westphalian B of Joggins, Nova Scotia. Linn.

Soc. London, Proc. 177:63-97.

1967. An adelogyrinid lepospondyl amphibian from the Upper Carbonif-
erous. Canadian Jour. Zool. 45: 1-16.

1968. The postcranial skeleton of the Permian microsaur Parity lus. Ca-
nadian Jour. Zool. 46:1175-1192.

1969. A new family of Carboniferous amphibians. Paleontology 12:537-

548.

1970. The ancestry of reptiles. Roy. Soc. London, Philos. Trans., ser. B,

257:267-308.

, and Donald Baird. 1968. The Carboniferous amphibian Tuditanus

[ Eosauravus ] and the distinction between microsaurs and reptiles. Amer. Mus.
Nat. Hist., Novitates 2337:1-50.

Case, E. C. 1908. Description of vertebrate fossils from the vicinity of Pittsburgh,
Pennsylvania. Carnegie Mus., Ann. 4:234-241.

1911. A revision of the Cotylosauria of North America. Carnegie Inst.

Washington, Publ. 145:1-122.

Chronic, John. 1958. Pennsylvanian rocks in central Colorado. Rocky Mountain
Assoc. Geol., Symposium on Pennsylvanian Rocks of Colorado and Adjacent
Areas, p. 59-63.

Clapham, W. B. 1970. Nature and paleogeography of Middle Permian floras of
Oklahoma as inferred from their pollen record. Jour. Geol. 78:153-171.

DeMar, Robert. 1970. A primitive pelycosaur from the Pennsylvanian of Illinois.
Jour. Paleont. 44:154-163.

Edmund, A. G. 1960. Tooth replacement phenomena in the lower vertebrates. Roy.
Ontario Mus., Life Sci. Div., Contrib. 52:1-190.

1969. Dentition. In Biology of the Reptilia, Carl Gans, A. d’A. Bellairs

and T. S. Parsons, eds., London and New York, Academic Press, p. 117-200.

Elias, M. K. 1970. Progress in correlation of Carboniferous rocks. Compte Rendu
6e Congres Internat. Strat. Geol. Carbonif., Sheffield, 1967, vol. 2:695-714.

Geinitz, H. B., and J. V. Deichmuller. 1882. Die Saurier der unteren Dyas von
Sachsen. Paleontographica 29:1-46.

Hotton, Nicholas, III. 1970. Mauchchunkia bassa, gen. et sp. nov., an anthraco-
saur (Amphibia, Labyrinthodontia) from the Upper Mississippian. Cleveland
Mus. Nat. Hist., Kirtlandia 12:1-38.

Huene, Friedrich Von. 1913. The skull elements of the Permian Tetrapoda in the
American Museum of Natural History, New York. Amer. Mus. Nat. Hist.,
Bull. 23:315-386.

Langenheim, R. L. 1952. Pennsylvanian and Permian stratigraphy of the Crested
Butte Quadrangle, Gunnison County, Colorado. Amer. Assoc. Petrol. Geol.,
Bull. 36:543-574.

Langston, Wann, Jr. 1963. Fossil vertebrates and the late Paleozoic red beds of
Prince Edward Island. Nat. Mus. Canada, Bull. 187:1-36.

30

Contributions in Science

No. 223

Lewis, G. E., and P. P. Vaughn. 1965. Early Permian vertebrates from the Cutler
Formation of the Placerville area, Colorado, with a section on Footprints from
the Cutler Formation by Donald Baird. U. S. Geol. Survey, Prof. Paper 503-
C: 1-50.

Mallory, W.W. 1958. Pennsylvanian coarse arkosic redbeds and associated moun-
tains in Colorado. Rocky Mountain Assoc. Geol., Symposium on Pennsylvanian
Rocks of Colorado and Adjacent Areas, p. 17-20.

1960. Outline of Pennsylvanian stratigraphy of Colorado. Geol. Soc. Amer.

and Rocky Mountain Assoc. Geol., Guide to the Geology of Colorado, p. 23-33.

Olson, E. C. 1947. The family Diadectidae and its bearing on the classification of
reptiles. Fieldiana: Geol. 11:1-53.

— 1950. The temporal region of the Permian reptile Diadectes. Fieldiana:

Geol. 10:63-77.

— 1961. Jaw mechanisms: rhipidistians, amphibians, reptiles. Amer. Zoologist

1:205-215.

... — 1966. Relationships of Diadectes. Fieldiana: Geol. 14:199-227.

Panchen, A. L. 1967. The homologies of the labyrinthodont centrum. Evolution
21:24-33.

Parrington, F. R. 1967. The vertebrae of early tetrapods. Colloq. Internal. Cent.
National Recherche Sci. 163:269-279.

Peabody, F. E. 1952. Petrolacosaurus kansensis Lane, a Pennsylvanian reptile from
Kansas. Univ. Kansas Paleont. Contrib., Vertebrata 1:1-41.

1957. Pennsylvanian reptiles of Garnett, Kansas: edaphosaurs. J. Pale-

ontol. 31:947-949.

1959. The oldest known reptile, Eosauravus copei Williston. Smithsonian

Misc. Collec. 139:1-14.

Romer, A. S. 1952. Late Pennsylvanian and Early Permian vertebrates of the
Pittsburgh- West Virginia region. Carnegie Mus., Ann. 33:47-113.

1963. The larger embolomerous amphibians of the American Carbonif-
erous. Harvard Univ., Mus. Comp. Zool., Bull. 128:415-454.

- 1964. Diadectes an amphibian? Copeia 1964:718-719.

1969. The cranial anatomy of the Permian amphibian Pantylus. Harvard

Univ., Mus. Comp. Zool., Breviora 314:1-37.

Scott, G.R. 1967. Upper part of Minturn Formation possibly of Permian age. U. S.
Geol. Survey, Prof. Paper 575-A:82.

Stappenbeck, Richard. 1905. fiber Stephanospondylus n. g. und Phanerosaurus
H. v. Meyer. Deutsch. geol. Ges., Zeitschr. 57:380-437.

Sturgeon, M. T., and R. D. Hoare. 1968. Pennsylvanian brachiopods of Ohio.
Ohio Dept. Nat. Resources, Div. Geol. Survey, Bull. 63:1-95.

Thomson, K. S., and K. H. Bossy. 1970. Adaptive trends and relationships in early
Amphibia. Forma et Functio 3:7-31.

Vaughn, P. P. 1964. Vertebrates from the Organ Rock Shale of the Cutler Group,
Permian of Monument Valley and vicinity, Utah and Arizona. J. Paleontol.
38:567-583

- 1969. Upper Pennsylvanian vertebrates from the Sangre de Cristo Forma-
tion of central Colorado. Los Angeles Co. Mus., Contrib. Sci. 164:1-28.

1970. Alternation of neural spine height in certain Early Permian tetra-
pods. So. Calif. Acad. Sci., Bull. 69:80-86.

Watson, D. M. S. 1954. On Bolosaurus and the origin and classification of reptiles.
Harvard Univ., Mus. Comp. Zool., Bull. 111:295-449.

Accepted for publication December 20, 1971

NUMBER 224
FEBRUARY 28, 1972

Ct~L ut

TWO NEW M1CROVELIA FROM
CRABHOLES IN COSTA RICA
(Hemiptera: Veliidae)

By John T. Polhemus and Charles L. Hogue

CONTRIBUTIONS IN SCICNCC

8

NATURAL HISTORY MUSEUM • LOS ANGELES COUNTY

CONTRIBUTIONS IN SCIENCE is a series of miscellaneous technical papers
in the fields of Biology, Geology and Anthropology, published at irregular intervals
by the Natural History Museum of Los Angeles County. Issues are numbered sep-
arately, and numbers run consecutively regardless of subject matter. Number 1 was
issued January 23, 1957. The series is available to scientific institutions and scien-
tists on an exchange basis. Copies may also be purchased at a nominal price. Inquiries
should be directed to Virginia D. Miller, Natural History Museum of Los Angeles
County, 900 Exposition Boulevard, Los Angeles, California 90007.

INSTRUCTIONS FOR AUTHORS

Manuscripts for CONTRIBUTIONS IN SCIENCE may be in any field of
Life or Earth Sciences. Acceptance of papers will be determined by the amount
and character of new information. Although priority will be given to manuscripts
by staff members, or to papers dealing largely with specimens in the collections of
the Museum, other technical papers will be considered. All manuscripts must be
recommended for consideration by the curator in charge of the proper section or
by the editorial board. Manuscripts must conform to those specifications listed
below and will be examined for suitability by an Editorial Committee including
review by competent specialists outside the Museum.

Authors proposing new taxa in a CONTRIBUTIONS IN SCIENCE must
indicate that the primary type has become the property of a scientific institution of
their choice and cited by name.

MANUSCRIPT FORM.-(l) The 1964 AIBS Style Manual for Biological
Journals is to be followed in preparation of copy. (2) Double space entire manu-
script. (3) Footnotes should be avoided if possible. Acknowledgments as footnotes
will not be accepted. (4) Place all tables on separate pages. (5) Figure legends and
unavoidable footnotes must be typed on separate sheets. Several of one kind may
be placed on a sheet. (6) An abstract must be included for all papers. This will be
published at the head of each paper. (7) A Spanish summary is required for all
manuscripts dealing with Latin American subjects. Summaries in other languages
are not required but are strongly recommended. Summaries will be published at
the end of the paper. (8) A diagnosis must accompany any newly proposed taxon.
(9) Submit two copies of manuscript.

ILLUSTRATIONS. — -All illustrations, including maps and photographs, will
be referred to as figures. All illustrations should be of sufficient clarity and in the
proper proportions for reduction to CONTRIBUTIONS page size. Consult the 1964
AIBS Style Manual for Biological Journals in preparing illustration and legend copy
for style. Submit only illustrations made with permanent ink and glossy photo-
graphic prints of good contrast. Original illustrations and art work will be returned
after the manuscript has been published.

PROOF.— Authors will be sent galley proof which should be corrected and
returned promptly. Changes in the manuscript after galley proof will be billed to
the author. Unless otherwise requested, page proof also will be sent to the author.
One hundred copies of each paper will be given free to each author or divided
equally among multiple authors. Orders for additional copies must be sent to the
Editor at the time corrected galley proof is returned. Appropriate order forms will
be included with the galley proof.

Virginia D. Miller
Editor

TWO NEW M ICR O V ELIA FROM CRABHOLES IN COSTA RICA
(Hemiptera: Veliidae)

By lOHN T. POLHEMUS1 AND CHARLES L. HOGUE2

Abstract: Microvelia inquilina, n. sp. and Microvelia
chanita, n. sp. are described from the Pacific Coast of Costa Rica.

These two species, along with another Costa Rican species,

M. or aria Drake (crab unrecorded), are found inhabiting land
crab burrows made by Cardisoma crassum and Ucides occi-
dentals respectively, but the nature of the association is unknown.

The two Microvelia described below were found inhabiting crabholes
on the Pacific Coast of Costa Rica. Previously, Microvelia oraria was the
only veliid known from crabholes and was described from an Atlantic Coast
locality in Costa Rica by Drake (1952).

Because veliids are poorly known in the Neotropical region, it is hazard-
ous to surmise that crabholes are the sole habitat of these new species. Yet
one of them, inquilina , has reduced eyes similar to the bromeliad-inhabiting
species laesslei Drake and Hussey. Compared to normal pond and stream
species, the ommatidia are larger but with only about half as many. For
example, Microvelia pulchella Westwood (a pond dweller) has an interocular
space to eye width ratio (I/W) of 2.43, whereas in inquilina and laesslei
I/W is 3.40 and 3.34 respectively. Small eyes may be an adaptation to special-
ized container habitats such as bromeliads and crabholes offer.

The material upon which these species are based was made available
by the junior author and Dr. Donald B. Bright, California State College,
Fullerton, from their collections in connection with a general study of the
biology of land crabs and their burrow associates (LCBA), a project con-
ducted with the support of grants from the American Philosophical Society.

All specimens were taken from samples of water extracted from deep
within land crab burrows with a simple bottle pump fixed with an intake
hose of one-half inch inside diameter. The bugs are hygrophobic and readily
come to the surface of water taken with the pump. The construction and
use of the device itself (small type mosquito pump) is described by Belkin
et al. (1965:70-71).

Both species were taken in the same locality and general habitat. How-
ever, their specific microhabitats are very different. Microvelia chanita was
found in only a single burrow, that of a full-grown crab, Cardisoma crassum.
The collectors observed that this crab typically constructs its burrows just
above the highest high tide line where they are never (or rarely) flooded but
receive ground water most of the year directly from the sea or from some

*3115 South York, Englewood, Colorado 80110

2Entomology Section, Natural History Museum of Los Angeles County, Los Angeles,
California 90007

1

2

Contributions in Science

No. 224

other proximate water body. In the present case the burrow was located only
a few feet from a fresh water seepage pond which was separated from the
sea by a slight rise and a distance of approximately 100 meters. At the time
of the collection the level of free water was depressed below the burrow
mouth, and the water was brackish (NaCl 2020 ppm). The bugs were ob-
served in and about the crabhole but most were collected with the mosquito
pump. None were seen on the surface of the nearby pond, but they might
have occurred there also. It is possible that the presence of these water bugs
at this site was simply accidental, the species being normally adapted to living
on open pools as is true of most of its relatives.

From the reduced eyes of inquilina, which may be a morphological trait
associated with life in a container habitat (see above), behavior, and that it
was collected repeatedly from crabholes (LCBA 526 represents a pooled
sample of 25 burrows of the crab, Ucides occidentals, all siphoned with the
bottle pump), this species is more likely than chanita to be a normal inhabi-
tant of this specialized microhabitat. Ucides occidentalis, unlike C. crassum ,
constructs its burrows at a low elevation where they are partially or completely
flooded by daily high tides or at least annual spring tides. This poses the
additional question: if its niche is destroyed for a portion of the day or year,
does this species become a littoral dweller, invade the burrows of other crabs,
or utilize a resistant stage (egg?) to pass this critical period? Extensive col-
lecting in the type locality of inquilina in the season of high tides has yielded
no specimens from C. crassum holes.

Microvelia inquilina Polhemus and Hogue, new species
DESCRIPTION

Apterous male, adult:

Size. — Very small, short, broad; length 1.15 mm; width 0.44 mm.

Coloration and vestiture. — Ground color black to blackish brown. Grey
pruinose on areas as follows: fore part of head; connexivum, and much of
abdominal dorsum, lateral portions of tergite 2; all but median area on tergite
3; median third of tergite 6; broad triangular area of tergite 7; all of genital
segment 1 dorsally. Pronotum broadly testaceous, blackish brown laterally
beyond middle of eyes. Apex of abdominal tergite 6 and genital segments
brown, latter lighter ventrally; underparts of head and rostrum testaceous.
Legs and antenna yellowish to yellow brown. Entire body covered with short,
semi-erect pubescence.

Head. — Length .28 mm; width (including eyes) .40 mm; interocular
space .25 mm. Vertex strongly convex; eyes small, with about 50 ommatidia.
Antennal formula; segments I-IV, 8:7:16:18; segments 1 stout, segment 2
less stout, segments 3 and 4 slender; all segments clothed with long hairs.
Rostrum reaching past front coxae.

Thorax. — Proportional lengths, pronotum/ mesonotum: 6/4. Width across
humeri .57 mm. Posterior margins of pronotum, mesonotum straight; metano-
tum with angles broadly exposed, length .13 mm; mesonotum with small pits

1972

Two New MICROVEL1A

3

behind posterolateral angles of pronotum, widely separated (.35 mm); dorsal
surface of thorax feebly convex.

Legs short, stout, covered with pale hairs, longer on tibia; fore tibia
with short comb. Measurements of legs as follows:

Femur

Tibia

Tarsal 1

Tarsal 2

Anterior

.37 mm

.27

.17

—

Middle

.42

.33

.07

.13

Posterior

.42

.45

.08

.13

Abdomen. — Proportional lengths, abdominal tergites I- VII, 7:5:5: 3:3:
5:9; first genital segment protruding from tergite 7 by .10 mm, rounded
apically; connexiva moderately broad (.10 mm), slightly raised; entire ab-
domen broad, tapering slightly posteriorly, lateral margins of connexiva more
sharply rounded along tergites 6-7. Venter of abdomen broadly raised medi-
ally, extending onto ventrite 5, which is produced slightly medio-caudad and
emarginate; ventrite 6 similarly but more strongly produced and emarginate,
excavate on midline; ventrite 7 not raised medially; broadly, roundly and
deeply excavate medially forming a semicircular depression opening caudad.
Genital segment 1 roundly emarginate ventrally; segment 2 swollen, not
extending beyond tip of segment 1; parameres visible, hooklike, extending
caudad and upward along grooves on the posterolateral margins of genital
segment 1 (Fig. 1 F).

Apterous female, adult:

Very similar to male, except connexivum almost vertical, abdominal
venter unmodified, body somewhat more robust; length 1.33 mm; width
0.65 mm.

MATERIAL

Holotype 8 , Allotype 9 , and Paratypes 6 8 8 , 7 $ $ , Costa Rica, Puntarenas
Province, Boca de Barranca, 9-11 Feb. 1969, Hogue and Bright, LCBA 526,
ex. crabhole Ucides occidentalis. The holotype, allotype and seven paratypes
are deposited in the collections of the Natural History Museum of Los Angeles
County. Six paratypes are in the Polhemus collection.

DIAGNOSIS

The color, extremely small size, larger proportional length of head vs.
remainder of body (17/52 = .328), small eyes, long antenna and modifica-
tion of the male venter distinguish this species from all other Microvelia.
Microvelia laesslei Drake and Hussey and Microvelia distanti Lundblad, with
which inquilina would most likely be confused, are both larger ( laesslei 8
2.28 mm, $ 2.3 mm; distanti 8 1.9 mm, 9 2.3 mm). Neither of them has
ventrite 5 produced or ventrite 6 excavated medially, and their heads are
proportionally smaller than inquilina (length of head/ remainder of body:
distanti, 35/155 = .226; laesslei, 45/185 = .243). Additionally, the colora-

4

Contributions in Science

No. 224

Figure 1. A-B: Microvelia chanita, new species. A, female, dorsal view; B, antenna.
C-F: Microvelia inquilina , new species. C, male, dorsal view; D, male parainere,
dorsal (left) and lateral (right) views; E, antenna; F, male, apical abdominal seg-
ments, ventral view.

1972

Two New MICROVEL1A

5

tion is different, distanti being deep brown with the first two tergites pruinose
and the first three connexiva flavous forming a light transverse band, and
laesslei being deep brown with a rufous pronotum and white wing pads in the
micropterous form (apterous form not known). In both of these species the
pronotum covers the mesonotum, whereas in inquilina the mesonotum is
broadly exposed. The eyes in distanti are not reduced significantly in relation
to the head as in inquilina.

Microvelia chanita Polhemus and Hogue, new species
DESCRIPTION

Apterous female, adult:

Size.— Small, subfusiform. Length, 1.77 mm; width, 0.72 mm.

Coloration and vestiture. — Ground color brown; grey pruinose on fore
part of head, collar, median wedge on abdominal tergite 2, all of tergites 5
and 6; anterior lobe of pronotum white pruinose; median area of head and
pronotum, most of mesonotum and tergite 1, posterior part of each connexival
segment yellowish; venter ochraceous, midventral areas, midlateral spots
brownish; antenna ochraceous to brownish; legs leucine to ochraceous, dorsally
and apically brownish; underparts of head and rostrum ochraceous.

Head. — Length .37 mm, width (including eyes) .43 mm, interocular
space .28 mm. Vertex strongly convex; eyes of moderate size; antennal formula
I-IV, 13:10:16:29; segment 1 stout, 2 less stout, 3-4 slender; all segments
clothed with recumbent hairs of length equal to diameter of segment 2, and
scattered longer hairs. Rostrum reaching beyond fore coxae.

Thorax. — Proportional lengths, pronotum/ mesonotum : 11/3; midline
lengths, anterior pronotal lobe/ posterior pronotal lobe: 5/6. Width across
metanotal angles .72 mm; collar marked by a row of widely spaced pits; lobes
of pronotum separated by a row of deep pits, interrupted medially, as is
transverse row of pits on caudal lobe; caudal margins of pronotum and
mesonotum slightly concave; metanotal angles narrowly exposed, length .83
mm (from postero-lateral angle of mesonotum); mesonotum broadly exca-
vate under posterolateral margins of pronotum; lateral margins of thorax set
with semi-long, curved, bristly hairs; dorsal surface slightly convex, pronotum
depressed below level of mesonotum; propleura depressed along caudal
margin.

Legs of moderate length, covered with short pale hairs, longer on under
surface of femora; measurements of legs as follows:

Femur

Tibia

Tarsal 1

Tarsal 2

Anterior

.47 mm

.40

.23

—

Middle

.52

.43

.12

.15

Posterior

.52

.62

.13

.17

Fore tibia slightly flattened and widened apically, narrowing abruptly just
before apex.

6

Contributions in Science

No. 224

Abdomen. — Proportional lengths of abdominal tergites I- VIII, 9:8:8 :8:
9: 9: 8: 6. Connexiva moderately broad, semi-erect along tergite 1 to vertical at
apex; set with bristly hairs at apex. Venter broadly rounded, feebly flattened
medially, clothed with decumbent hairs visible from above. Shape as in figure
1 A.

Male: Unknown.

MATERIAL

Holotype 9 and 8 9 9 Paratypes, Costa Rica, Puntarenas Prov., Boca de
Barranca, 8 August 1967, Hogue and Bright, LCBA 158, ex. crabhole Cardi-
soma crassum. Deposited in the collection of the Natural History Museum of
Los Angeles County. Three paratypes are in the Polhemus collection.

DIAGNOSIS

Microvelia chanita belongs to the albonotata group including albonotata
Champion, mimula White, tateiana Drake, quieta Drake, novana Drake,
cubana Drake and portoricensis Drake. This group, not previously recognized,
is comprised of those small species (circa 2 mm) which have the pronotum
of medium length, having two distinct lobes separated by a depressed trans-
verse line of pits, but leaving much of the mesonotum exposed. The ratio of
midline length of pronotum/ mesonotum in this group varies from 12/7 = 1.73
( tateiana ) to 11/3 = 3.67 {chanita).

The primary distinguishing characteristics of chanita are extremely long
fourth antennal segment combined with a relatively short thorax (measured
on dorsal midline, thorax/ head: 15/25). M. albonotata, the only other species
with very long fourth antennal segments, has thorax/head: 25/28. Addition-
ally, chanita has narrow apical abdominal tergites, a character state shared
within the albonotata group only by M. portoricensis.

Resumen

Microvelia inquilina, sp. nov., y Microvelia chanita, sp. nov., de la costa
pacifica de Costa Rica son descritos. Estas dos especies, con otra de Costa
Rica, M . or aria Drake (cangrejo no conocido), se encuentran habitando las
cuevas de cangrejos terestres construidas por Car disoma crassum y Ucides
occidentalis, respectivamente, pero la naturaleza de la asociacion no es
conocida.

Literature Cited

Drake, C. J. 1952. Two new Microvelia Westwood (Hemiptera: Veliidae). Bull.
Brooklyn Ent. Soc. 47(1) : 13-15.

Belkin, J. N., C. L. H >gue, P. Galindo, T. H. G. Aitken, R. X. Schick and W. A.
Powder. 1965. Mosquito studies (Diptera, Culicidae) II. Methods for the
collection, rearing and preservation of mosquitoes. Contrib. Amer. Entomol.
Inst. 1(2): 19-78.

Accepted for publication January 18, 1972

So 7 ,73

CzL-ftf

NUMBER 225
MARCH 2, 1972

/

DORSADENA YAQUINAE, A NEW GENUS
AND SPECIES OF MYCTOPHID FISH
FROM THE EASTERN NORTH PACIFIC OCEAN

By Leonard R. Coleman
and Basil G. Nafpaktitis

CONTRIBUTIONS IN SCICNCC

8

NATURAL HISTORY MUSEUM • LOS ANGELES COUNTY

CONTRIBUTIONS IN SCIENCE is a series of miscellaneous technical papers
in the fields of Biology, Geology and Anthropology, published at irregular intervals
by the Natural History Museum of Los Angeles County. Issues are numbered sep-
arately, and numbers run consecutively regardless of subject matter. Number 1 was
issued January 23, 1957. The series is available to scientific institutions and scien-
tists on an exchange basis. Copies may also be purchased at a nominal price. Inquiries
should be directed to Virginia D. Miller, Natural History Museum of Los Angeles
County, 900 Exposition Boulevard, Los Angeles, California 90007.

INSTRUCTIONS FOR AUTHORS

Manuscripts for CONTRIBUTIONS IN SCIENCE may be in any field of
Life or Earth Sciences. Acceptance of papers will be determined by the amount
and character of new information. Although priority will be given to manuscripts
by staff members, or to papers dealing largely with specimens in the collections of
the Museum, other technical papers will be considered. All manuscripts must be
recommended for consideration by the curator in charge of the proper section or
by the editorial board. Manuscripts must conform to those specifications listed
below and will be examined for suitability by an Editorial Committee including
review by competent specialists outside the Museum.

Authors proposing new taxa in a CONTRIBUTIONS IN SCIENCE must
indicate that the primary type has become the property of a scientific institution of
their choice and cited by name.

MANUSCRIPT FORM.— ( 1 ) The 1964 AIBS Style Manual for Biological
Journals is to be followed in preparation of copy. (2) Double space entire manu-
script. (3) Footnotes should be avoided if possible. Acknowledgments as footnotes
will not be accepted. (4) Place all tables on separate pages. (5) Figure legends and
unavoidable footnotes must be typed on separate sheets. Several of one kind may
be placed on a sheet. (6) An abstract must be included for all papers. This will be
published at the head of each paper. (7) A Spanish summary is required for all
manuscripts dealing with Latin American subjects. Summaries in other languages
are not required but are strongly recommended. Summaries will be published at
the end of the paper. (8) A diagnosis must accompany any newly proposed taxon.
(9) Submit two copies of manuscript.

ILLUSTRATIONS. — All illustrations, including maps and photographs, will
be referred to as figures. All illustrations should be of sufficient clarity and in the
proper proportions for reduction to CONTRIBUTIONS page size. Consult the 1964
AIBS Style Manual for Biological Journals in preparing illustration and legend copy
for style. Submit only illustrations made with permanent ink and glossy photo-
graphic prints of good contrast. Original illustrations and art work will be returned
after the manuscript has been published.

PROOF.— Authors will be sent galley proof which should be corrected and
returned promptly. Changes in the manuscript after galley proof will be billed to
the author. Unless otherwise requested, page proof also will be sent to the author.
One hundred copies of each paper will be given free to each author or divided
equally among multiple authors. Orders for additional copies must be sent to the
Editor at the time corrected galley proof is returned. Appropriate order forms will
be included with the galley proof.

Virginia D. Miller
Editor

DORSADENA YAQUINAE, A NEW GENUS AND SPECIES OF
MYCTOPHID FISH FROM THE EASTERN
NORTH PACIFIC OCEAN

By Leonard R. Coleman1 and Basil G. Nafpaktitis2

Abstract: A new genus and species of myctophid fish,
Dorsadena yaquinae, from the eastern north Pacific Ocean is
described. Relationship between the new form and Lampadena
Goode and Bean is suggested by similarities in the structure, size
and position of the supra- and infracaudal luminous glands, in
the arrangement of the body photophores and in otolith mor-
phology. Dorsadena yaquinae, like Lampadena and Taaningich-
thys, seems to be one of the deepest dwelling myctophids. Its
isolated occurrence off Oregon may be attributed to inadequate
sampling of depths exceeding 1500 meters in the central and
western north Pacific. On the other hand, the eastern north
Pacific specimens may represent an expatriate population.

Recent collections of oceanic fishes by the Department of Oceanography,
Oregon State University, have yielded specimens of an undescribed lantern-
fish. This fish is so distinct from any other myctophid as to preclude its place-
ment in any of the approximately thirty genera of the family.

The new species is represented by five specimens, 58.0-101.5 mm in
standard length, collected between latitudes 44°N and 45 °N, and longitudes
134°W and about 139°W where subarctic water predominates in at least the
upper 300 meters.

Counts and measurements were taken according to Nafpaktitis (1968).
Photophore and otolith terminologies follow those of Bolin (1939) and Frizzell
and Dante (1965), respectively. The otoliths are deposited in the collections of
John E. Fitch of the California Department of Fish and Game.

Dorsadena , new genus

Diagnosis: A large, elongate luminous gland immediately in front of adi-
pose fin. Large, undivided supra- and infracaudal luminous glands. Four to five
Prc, in three groups : first two close together and about at level of dorsal mar-
gin of infracaudal luminous gland, third at midlateral line, fourth posterior to,
and about at level of ventral margin of, supracaudal luminous gland; often a
fifth Prc develops close to, and at level of, fourth Prc. Numerous minute sec-
ondary photophores on head, trank and base of caudal fin.

The name Dorsadena [dorsal and adena, from the Greek a$v)v (aden) =
gland] refers to the unique preadipose gland. Type species:

department of Oceanography, Oregon State University, Corvallis, Oregon 97331.
department of Biological Sciences, University of Southern California, Los Angeles,
Calif. 90007; and Research Associate in Ichthyology, Natural History Museum of
Los Angeles County, Los Angeles, Calif. 90007.

1

2

Contributions in Science

No. 225

Dorsadena yaquinae , new species
Figures 1-4

Holotype: Los Angeles County Museum of Natural History (LACM)
30841-1; 77.0 mm, R/V YAQUINA, haul MT-866, between 45°05/N,
138°33'W and 44°44'N, 138°32'W, 0453-1205 hrs, 25 July 1966; 10' Isaacs-
Kidd Midwater Trawl, depth of haul 0-2700 m, 8000 m of wire out; bottom
depth approximately 4207 m.

Paratypes: Oregon State University Department of Oceanography
(OSUDO) 1226, 101.5 mm, and 1227, 58.0 mm. Collection data for both
are the same as for the holotype. U.S. National Museum (USNM) 204869;
87.0 mm, R/V YAQUINA, station NH-450, haul OTB-163, between 44°39'N,
134°34'W and 44°45/N, 134°46'W, 1835-0400 hrs, 1-2 March 1967; 22'
shrimp-type otter trawl, depth of haul 0-3860 m, 6000 m of wire out. Museum
of Comparative Zoology (MCZ) 46681; 62.0 mm, R/V YAQUINA, sta-
tion NH-450, haul MT-1040, between 44°45'N, 134°46'W and 44°43'N,
134°42'W, 0223-0305 hrs, 2 March 1967; 6' Isaacs-Kidd Midwater Trawl,
depth of haul 0-180 m, 800 m of wire out; bottom depth approximately
3800 m.

Diagnosis: As for genus.

Description: D. 14-15; A. 12-14; P. 15-16; V. 8 (9 on one side of one
specimen); gill rakers (4)5+1 + 11, plus 1-3 rudiments on the upper limb
and 3-4 rudiments on the lower limb of the first (right) gill arch; PO 6-8;
VO 3-5; SAO 3; AO 5-7 + 3-5, total 9-11; Prc 2+ 1 + 1-2.

A moderately large myctophid fish. Head large, about 3.3 in standard
length (SL). Eye large, 12.3 (11.6-13.5) in SL, 3.8 (3.7-4.1) in length of
head and 2.5 (2.3-2. 8) in length of upper jaw. Mouth large, terminal, some-
what oblique; length of upper jaw about 5 in SL, 1.5 in length of head, extend-
ing 1.0 to 1.3 times the diameter of eye behind vertical through posterior
margin of orbit. Length of snout 1.4 (1.2-1. 5) in diameter of eye. Posterior
opercular margin forming a blunt point somewhat above base of pectoral fin.
Pterotic spine well developed. Caudal peduncle 10.0 (9.0-11.0) in SL.

Origin of dorsal fin over base of ventral fin. Origin of anal fin on, or
slightly in advance of, vertical through end of base of dorsal fin. Pectoral fin
short, its delicate, fragile rays about as long as diameter of eye. Ventral fins
extending to anus. Base of adipose fin over end of base of anal fin.

Dn absent. A very small, poorly developed Vn immediately above, or in
contact with, dorsal margin of COl (lacrimal) bone. Opj poorly defined,
about at level of angle of mouth and close behind preopercular margin. Op2
twice as large as general body photophores, at least twice its own diameter
above and behind Opx.

Body photophores generally small and ill defined, at least in preserved
specimens. PLO slightly in advance of vertical through upper end of base of
pectoral fin and about its own diameter below lateral line. PVOj under, or
slightly in advance of, PV02, which is located about its own diameter in front

1972

New Genus and Species of Myctophid Fish

3

Figure 1. Dorsadena yaquinae; holotype, 77.0 mm in SL; LACM 30841-1.

4

Contributions in Science

No. 225

Table 1. Measurements of Dorsadena yaquinae

OSUDO

MCZ

OSUDO

LACM

USNM

1226

46681

1227

30841-1*

204869

101.5 mm

62.0mm 58.0mm 77.0mm 87.0mm

Character

Measurements in percent of standard length

Mean

Diameter of eye

7.4

8.4

8.6

8.2

8.0

8.1

Length of upper jaw

20.7

19.8

19.8

20.6

19.5

20.1

Length of head

30.5

30.6

31.6

31.2

30.5

30.9

Depth of caudal peduncle

11.0

11.3

9.7

10.4

9.2

10.3

From tip of snout to base
of pectoral fin

33.5

33.1

33.6

33.0

32.2

33.1

From tip of snout to base
of ventral fin

48.3

48.4

46.9

50.6

47.7

48.4

From tip of snout to
origin of dorsal fin

48.3

48.4

46.6

50.6

48.3

48.4

From tip of snout to
origin of anal fin

63.1

64.5

63.8

63.6

64.4

63.9

From tip of snout to base
of adipose fin

77.8

79.0

75.0

77.9

75.6

77.1

Length of caudal glands

5.4

4.8

5.2

4.2

4.6

4.8

Length of preadipose gland

1 8.4

8.1

6.2

7.8

7.8

7.7

Character

Measurements

in percent of head length

Mean

Length of upper jaw

67.7

64.7

62.8

66.3

64.2

65.1

Diameter of eye

24.2

27.4

27.3

26.3

26.4

26.3

Length of snout

21.0

18.4

19.1

18.8

18.9

19.2

*Holotype

of middle of base of pectoral fin. Six to eight PO, variably spaced on a wavy
line. VLO about 1.5 times its own diameter below lateral line. Three to five,
usually four, VO, level. SAO forming an obtuse angle; SAOi over anus and
slightly raised above level of last VO; distance between SA02 and SAOs 1.5 to
2.0 times as large as that between SAOx and SA02; SAOs somewhat in advance
of, or behind, vertical through center of SA02 and about its own diameter
below lateral line. First and last AOa interspaces sometimes distinctly enlarged;
first AOa, or last, or both slightly raised. Pol behind last AOa, under base of
adipose fin and about its own diameter below lateral line. AOp evenly spaced,
level; last AOp over anterior portion of infracaudal luminous gland. PrCj-Prc.,
interspace less than one photophore diameter; Prc2 slightly higher than Prcx;
Prc3 well behind Prc2 and at level of lateral line; one or two additional Prc

1972

New Genus and Species of Myctophid Fish

5

organs posterior to supracaudal luminous gland and under dorsal procurrent
caudal rays.

Supra- and infra caudal luminous glands undivided, of equal size, their
length 1. 6-2.0 times in diameter of eye, directly apposed to each other, and
framed by darkly pigmented tissue; most luminous tissue found within a
darkly pigmented “hood” at posterior part of each organ.

An undivided luminous gland, about as long as eye diameter, extending
from anterior end of base of adipose fin to about midway between end of base
of dorsal fin and adipose fin; gland outlined by black pigment, with luminous
tissue bulging dorsal.lv.

Large numbers of minute secondary p hot op bores present on head, trunk
and proximal part of caudal fin. Along the lateral line, they appear to be
arranged in a rather regular pattern (Fig. 2).

laws with needlelike teeth, inner ones longer than outer; 5 to 8 broad-
based, hook-like, forward-inclined teeth on posterior part of dentary; a long,
narrow band of slender teeth on each palatine; mesoptery golds with minute,
widely scattered teeth and enlarged, widely spaced ones along periphery and
posterior part of each mesopterygoid; vomer toothless.

The gonads of all five specimens are either poorly developed or regressed.

Circumorbital bones

The circumorbital bones (Fig. 3) show some interesting features. In the
following discussion the terminology is that used by Paxton (in press).

Figure 2. Dorsadena yaquinae ; distribution of secondary photophores on lateral
line scales.

6

Contributions in Science

No. 225

The anterodorsal part of the first circumorbital, COl (lacrimal of some
authors), is folded over to form a large, lateral flap anteroventrad to the eye.
This flap is clearly visible on intact specimens. Paxton (in press) found that
in myctophids “The anterodorsal margin is folded ventrally, so that the ante-
rior end of the COl approaches a closed tube in some forms. In a number of
species, the Vn orbital organ lies on top of the folded edge of the COl.” How-
ever, with the exception of those members of the genus Diaphus with a well
developed Vn (ventronasal) and those of the genus Gymnoscopelus, e. g.,
G. ( Gymnoscopelus ) opisthopterus, G. ( Nasolychnus ) piabilis, with extensive
luminous tissue along the anterior and anteroventral orbital margin, in no
other myctophid form is this flap so extensively developed. It is conceivable
that the ancestral stock from which Dorsadena evolved had a well-developed
Vn. Interestingly, the COl lateral flap appears relatively well developed in
Lampadena anomala, the Vn of which is very small, poorly developed and lies
anterodorsad to the COl.

According to Paxton (op. cit. ) , the lateral margin of the orbital portion
of the third circumorbital, C03 (jugal of some authors), in lantern fishes, is
solid or split. In many forms “A keel or flag of bone projects posteriorly from
the lateral margin at the level of the split. ...” In Dorsadena yaquinae the
lateral margin of the orbital portion of the C03 is split. At the level of
the split, the two parts contribute to the formation of a large, spine-like,
posteroventrally-directed bony process (Fig. 3), also clearly visible on intact
specimens. A relatively well-developed similar process is found also in Lampa-

b.n.

Figure 3. Dorsadena yaquinae; circumorbital bones.

1972

New Genus and Species of Myctophid Fish

7

dena, e.g., L. urophaos and L. anomola, and in some Lampanyctus.

The other circumorbital bones show no marked peculiarities.

Otoliths

Dorsadena yaquinae has a small sagitta (Fig. 4), which is almost as high
as it is long— length to height ratio 1.03:1. It is not notched posterodorsally
and its ventral margin is smooth. The rostrum is well developed; the anti-
rostrum bluntly rounded but distinct. The collum divides the sulcus into two
almost equal sections. The lateral face of the otolith is smooth and some-
what convex.

Nafpaktitis and Paxton (1968) have briefly discussed the trends in otolith
morphology within the genus Lampadena. The sagittae of all the species of
this genus, with the exception of L. anomala, are relatively large and clearly
longer than they are high. Their ventral margins and, in at least two cases,
dorsal margins as well, are scalloped. The rostra are little to moderately devel-
oped and the antirostra are in some cases indistinct. L. anomala has a rela-
tively small otolith with a length to height ratio of 1.2:1, a smooth ventral
margin and a greatly developed rostrum.

The otoliths of L. anomala and D. yaquinae are markedly similar. Fig-
ure 4 shows the otoliths of the two forms and also that of Taaningichthys sp.,
a genus closely related to Lampadena.

Relationships

Until thorough osteological studies are made on cleared and stained
specimens of Dorsadena, interpretations regarding relationships of the new
genus are of necessity based almost solely on external morphology.

There are several morphological similarities between Dorsadena and
Lampadena. The most striking similarity is found in the structure, size and
position of the supra- and infracaudal luminous glands. The body photophores
in both genera are rather poorly developed and similarly arranged. With very

Figure 4. Medial views of left otoliths, anterior end to the right: (A) Dorsadena
yaquinae, otolith 2.40 mm long, specimen 101.5 mm in SL; (B) Taaningichthys sp.,
otolith 1.90 mm long, specimen about 50 mm in SL; (C) Lampadena anomala,
otolith 1.95 mm long, specimen about 48 mm in SL.

8

Contributions in Science

No. 225

few exceptions, the PO and VO series are, in terms of numbers, remarkably
constant within the Myctophidae. In Dorsadena, as in Lampadena and the
closely related Taaningichthys , even the PO and VO vary in numbers. Limited
osteological observations (circumorbital bones) also revealed close similarities.

Nafpaktitis and Paxton (1968) pointed out the marked differences in
otolith morphology between L. anomala and all the rest of the species of Lam-
padena. In fact, the otolith of L. anomala may, in some important respects,
be considered as intermediate between the long, scalloped otoliths of the
rest of the species of Lampadena and the almost round, smooth-edged otolith
of Taaningichthys sp. (Fig. 4). In the same manner, the otolith of Dorsa-
dena yaquinae has features which may be considered intermediate between
L. anomala and Taaningichthys sp., perhaps somewhat closer to the former
than to the latter.

Most species of Lampadena appear to be among the deepest dwelling of
myctophids. The very few known captures of L. anomala with open nets have
been made below 750 meters. Shallow captures of large specimens during the
night, indicative of extensive vertical migration, are known for L. luminosa
and L. urophaos. Young (20-35 mm) specimens of L. speculigera, L. dea and
L. chavesi have been taken during the night in the upper 200 meters.

If we assume that the body photophores of Lampadena, which are poorly
developed (especially in L. anomala) and variable in numbers, reflect deep
mesopelagic, or bathypelagic, existence with limited or nonexisting migratory
habits, then the correlation is stronger in Taaningichthys. The two known
species of this genus show marked degeneration of body photophores and
lateral line. The photophores are small, highly superficial and their numbers
as well as their arrangement vary considerably. A third species (Davy, in
press) appears to have completely lost its body photophores. The lateral line
components are extremely reduced. The eye, in contrast, is very large and well
developed. Members of the genus Taaningichthys are seldom taken above 800
meters and they do not seem to undertake diel vertical migrations.

With the exception of a single specimen (MCZ 46681) taken with a 6'
Isaacs-Kidd Midwater Trawl between the surface and about 200 m, the speci-
mens of Dorsadena yaquinae were captured with larger gear and at depths
exceeding 2000 meters. The possibility does exist that the animals may have
been caught anywhere between the surface and the maximum depth of each
trawl, since the collecting gear used remained open throughout the operation.
However, with the exception mentioned above, Dorsadena has not been taken
in shallower hauls, which greatly outnumber the deep tows. In addition to
capture data, several features of the body photophores suggest that Dorsadena
occurs at depths similar to those occupied by Lampadena and Taaningichthys.
If this is the case, then evolutionary convergence could account for the state
of development of body photophores in the three genera. On the other hand,
a detailed osteological study may support our conclusion that Dorsadena is
closely related to Lampadena.

1972

New Genus and Species of Myctophid Fish

9

Following is a synoptic list of external characters that both relate and
distinguish the three genera, Lampadena, Dorsadena and Taaningichthys, as
they are understood at this time.

Lampadena Goode and Bean, 1896

1. Body moderately robust.

2. Ventral fins inserted under origin of dorsal fin.

3. Teeth on vomer present (absent in L. deal).

4. Lateral line well developed.

5. Preadipose luminous gland absent.

6. Crescent of white tissue on dorsal half of iris present in only one species,
L. chavesi.

7. PO 5-6; VO 3-6; SAO 3; AOa 3-8; AOp 2-5; Prc 2+ 1.

8. Secondary photophores absent or, if present, restricted to head.

Dorsadena, new genus

1. Body moderately robust.

2. Ventral fins inserted under origin of dorsal fin.

3. Teeth on vomer absent.

4. Lateral line well developed.

5. Preadipose luminous gland present.

6. Crescent of white tissue on iris absent.

7. PO 6-8; VO 3-5; SAO 3; AOa 5-7; AOp 3-5; Prc 2+1 + 1-2.

8. Secondary photophores present on head, trunk and proximal part of cau-
dal fin.

Taaningichthys Bolin, 1959

1. Body slender.

2. Ventral fins inserted in advance of origin of dorsal fin.

3. Teeth on vomer absent.

4. Lateral line absent or very poorly developed.

5. Preadipose luminous gland absent.

6. Crescent of white tissue present on posterior half of iris.

7. PO 5-6; VO 2-10; SAO 1; AOa 1-8; AOp 1-5; Prc 2+1; or photophores
absent.

8. Secondary photophores, if present, restricted to head and interradial mem-
brane of caudal fin.

Discussion

Most lanternfishes perform diel vertical migrations of several hundred
meters. During their vertical migrations, these animals cross a wide range of
temperature and salinity. It is therefore difficult to understand how a given
set of physico-chemical factors at a particular depth could limit the horizontal
distribution of these organisms. The answer, or answers, to the puzzle prob-
ably lie in the reproductive physiology on the one hand, and in the tolerance

10

Contributions in Science

No. 225

limits of the early, epipelagic stages on the other. While a large portion of the
epipelagic larvae remain within the ecologically optimum area where they
grow, sink and subsequently metamorphose, many may be transported by
currents to waters of different physico-chemical properties. In this alien envi-
ronment, some young will perish, others will survive, sink and metamorphose.
However, these expatriates are usually unable to reproduce. As Bolin ( 1959b)
points out: “While straggling adults may exist for long periods in waters far
beyond the normal range of the species, permanent populations are restricted
to the proximity of the areas where spawning can be successful.” It is there-
fore necessary to exercise extreme caution in discussing ranges and distri-
butional patterns, especially when we are dealing with oceanic, midwater
organisms with epipelagic larval stages, such as myctophids, because the area
in which a species can exist may be much larger than the area in which it can
spawn. For instance, are the subarctic waters off Oregon within the “normal”
range of Dorsadena yaquinael Does this fish spawn there? The poorly devel-
oped and, in the larger specimens, regressed gonads do not seem to indicate
that spawning takes place in that area. If it does, then the absence of larvae
and young in the California Current System may perhaps be accounted for by
the change in the physico-chemical properties of the subarctic water along
the course of the California Current.

On the basis of the available data, however, it seems more likely that
here we are dealing with an expatriate population originating in deep, seldom
sampled waters either of the Subtropic Region or of the central and western
Subarctic Region.

Acknowledgments

We thank the captain, crew and scientists of the R/V YAQUINA,
Oregon State University, for assisting in the work at sea. We are thankful
also to William G. Pearcy of Oregon State University, Richard H. Rosenblatt
and Robert L. Wisner of Scripps Institution of Oceanography for review-
ing the manuscript, and to John E. Fitch and Jack W. Schott of the Cali-
fornia Department of Fish and Game for supplying comparative otolith
material and for taking the otolith photographs. Financial support by grants
from AEC [AT (45-1) 1750; RLO/56] and NSF (GB-1588) is here grate-
fully acknowledged.

Literature Cited

Bolin, R. L. 1939. A review of the myctophid fishes of the Pacific Coast of the
United States and of lower California. Stan. Ich. Bull. Vol. 1 (4): 89-156.

1959a. Iniomi. Myctophidae from the “Michael Sars” North Atlantic

Deep-Sea Expedition 1910. In Rep. Sci. Res. “Michael Sars” N. Atlantic
Deep-Sea Exped. 1910, Bergen, 4, pt. 2 (7): 1-45.

- 1959b. Differential bipolarity in the Atlantic and Pacific as expressed

by the myctophid fishes. In International Oceanographic Congress, Reprints,
31 August-12 September 1959, Mary Sears, ed., American Association for the
Advancement of Science, Washington, D.C., p. 142-143.

1972

New Genus and Species of Myctophid Fish

11

Davy, B. A review of the lanternfish genus Taaningichthys (family Myctophidae)
with the description of a new species. U.S. Dept. Com., Fish. Bull. Vol. 70 (1),
(in press).

Frizzell, D. L., and J. H. Dante. 1965. Otoliths of some early Cenozoic fishes of
the Gulf Coast. J. Paleontol. 39: 687-718.

Goode, G. B., and T. H. Bean. 1896. Oceanic Ichthyology. U.S. Nat. Mus., Spec.
Bull. 553 p.

Nafpaktitis, B. G. 1968. Taxonomy and distribution of the lanternfishes, genera
Lobianchia and Diaphus, in the north Atlantic. Dana-Rep. 73. Copenhagen.
131 p.

Nafpaktitis, B. G., and J. R. Paxton. 1968. Review of the lanternfish genus
Lampadena with a description of a new species. Los Angeles Co. Mus., Contrib.
Sci. 138: 1-29.

Paxton, J. R. 1972. Osteology and relationships of the lanternfishes (Family
Myctophidae). Natural History Museum, Los Angeles Co., Bull. 13, (in press).

Accepted for publication Nov. 16, 1971

Printer! in T os Aneeles. California by Continental Grap

cJ 0 7' 73

CxL tcf

NUMBER 226
MARCH 21, 1972

THE AMPHINEMURA VENUSTA COMPLEX
OF WESTERN NORTH AMERICA
(PLECOPTERA: NEMOURIDAE)

By Richard W. Baumann
and Arden R. Gaufin

CONTRIBUTIONS IN SCI6NCC

NATURAL HISTORY MUSEUM

LOS ANGELES COUNTY

CONTRIBUTIONS IN SCIENCE is a series of miscellaneous technical papers
in the fields of Biology, Geology and Anthropology, published at irregular intervals
by the Natural History Museum of Los Angeles County. Issues are numbered sep-
arately, and numbers run consecutively regardless of subject matter. Number 1 was
issued January 23, 1957. The series is available to scientific institutions and scien-
tists on an exchange basis. Copies may also be purchased at a nominal price. Inquiries
should be directed to Virginia D. Miller, Natural History Museum of Los Angeles
County, 900 Exposition Boulevard, Los Angeles, California 90007.

INSTRUCTIONS FOR AUTHORS

Manuscripts for CONTRIBUTIONS IN SCIENCE may be in any field of
Life or Earth Sciences. Acceptance of papers will be determined by the amount
and character of new information. Although priority will be given to manuscripts
by staff members, or to papers dealing largely with specimens in the collections of
the Museum, other technical papers will be considered. All manuscripts must be
recommended for consideration by the curator in charge of the proper section or
by the editorial board. Manuscripts must conform to those specifications listed
below and will be examined for suitability by an Editorial Committee including
review by competent specialists outside the Museum.

Authors proposing new taxa in a CONTRIBUTIONS IN SCIENCE must
indicate that the primary type has become the property of a scientific institution of
their choice and cited by name.

MANUSCRIPT FORM.-(l) The 1964 AIBS Style Manual for Biological
Journals is to be followed in preparation of copy. (2) Double space entire manu-
script. (3) Footnotes should be avoided if possible. Acknowledgments as footnotes
will not be accepted. (4) Place all tables on separate pages. (5) Figure legends and
unavoidable footnotes must be typed on separate sheets. Several of one kind may
be placed on a sheet. (6) An abstract must be included for all papers. This will be
published at the head of each paper. (7) A Spanish summary is required for all
manuscripts dealing with Latin American subjects. Summaries in other languages
are not required but are strongly recommended. Summaries will be published at
the end of the paper. (8) A diagnosis must accompany any newly proposed taxon.
(9) Submit two copies of manuscript.

ILLUSTRATIONS. — All illustrations, including maps and photographs, will
be referred to as figures. All illustrations should be of sufficient clarity and in the
proper proportions for reduction to CONTRIBUTIONS page size. Consult the 1964
AIBS Style Manual for Biological Journals in preparing illustration and legend copy
for style. Submit only illustrations made with permanent ink and glossy photo-
graphic prints of good contrast. Original illustrations and art work will be returned
after the manuscript has been published.

PROOF.— Authors will be sent galley proof which should be corrected and
returned promptly. Changes in the manuscript after galley proof will be billed to
the author. Unless otherwise requested, page proof also will be sent to the author.
One hundred copies of each paper will be given free to each author or divided
equally among multiple authors. Orders for additional copies must be sent to the
Editor at the time corrected galley proof is returned. Appropriate order forms will
be included with the galley proof.

VrRGiNiA D. Miller
Editor

THE AMPHINEMURA VENUSTA COMPLEX
OF WESTERN NORTH AMERICA
(PLECOPTERA: NEMOURIDAE)

By Richard W. Baumann1 and Arden R. Gaufin2

Abstract: The Amphinemura venusta complex of Western
North America contains six species where only one was previ-
ously recognized. A comparison of the Amphinemura venusta
(Banks) holotype female with available specimens led to the
re-definition of this species. The male of A. venusta is described
and the species is recorded from Mexico, with the type locality in
Southern Arizona being the northern limit of distribution. Two
species, A. mexicana and A. puebla are described from near
Mexico City. The material from the Rocky Mountains called A.
venusta (Banks), as a result of the Needham and Claassen mono-
graph (1925), is given the name A. banksi. Two species are
named from Southwestern United States: A. apache and A.
mogollonica.

The species in the complex are apparently restricted to per-
manently running waters. In the United States, the flight period
is short, extending from July to September. The data available
for Mexican species indicate that the emergence period is en-
larged and may extend throughout the year.

Members of the genus Amphinemura occur throughout the Holarctic
and Oriental regions (lilies, 1965). This study deals with a species complex
found in the Western United States and Mexico. The complex is characterized
by its peculiar “windowed” forewings. This type of wing, darkly infuscated
with numerous clear spots in the cells (Fig. 1), is also present in some ne-
mourids from the Himalayas. The included species represent the only North
American species possessing this characteristic. Until now, these species were
all included under the specific name Amphinemura venusta (Banks). The
range of A. venusta was recorded by Ricker (1952) as extending from the
Rocky Mountains in Wyoming to the mountains around Mexico City. This
study, which was begun as part of a doctoral thesis by the senior author
(1970), delineates this distributional pattern using the six species presently
known in the complex.

Acknowledgments

The authors are grateful to Dr. Paul H. Arnand, Jr., California Academy
of Sciences (CAS); Dr. William F. Barr, University of Idaho (UI); Dr. C. J.
D. Brown, Montana State University (MSU); Dr. P. J. Darlington, Jr., Mu-
seum of Comparative Zoology, Harvard University (MCZ); Dr. Oliver S.
Flint, Jr., United States National Museum (USNM); Dr. W. J. Hanson and

department of Life Sciences, Southwest Missouri State College, Springfield, Mis-
souri 65802.

department of Biology, University of Utah, Salt Lake City, Utah 84112.

1

2

Contributions in Science

No. 226

Dr. George F. Knowlton, Utah State University (USU); Dr. Charles L.
Hogue, Natural History Museum of Los Angeles County (LACM); Mr.
Stanley G. Jewett, Jr., Portland, Oregon (SGJ); Dr. Richard W. Koss, Johns
Hopkins University; Dr. Carlos Sosa Moss, Escuela Nacional de Agricultura
de Mexico (ENAM); Dr. L. L. Pechuman, Cornell University (CU); Dr.
William E. Ricker, Fisheries Research Board of Canada (WER) ; Mr. Vincent
Roth, Southwest Research Station, Portal, Arizona (SWRS); Dr. Robert C.
Schuster, University of California, Davis (UCD); Dr. Donald W. Webb,
Illinois Natural History Survey (INHS) for making their specimens available
for this study. Abbreviations for collections of the authors are: Richard W.
Baumann (RWB) and University of Utah (UU).

Special thanks are given to Dr. Joachim lilies and Dr. Peter Zwick of the
Max-Planck Limnology Institute, Schlitz, Germany, for their help and the
use of Institute facilities for the preparation of this manuscript. The Spanish
abstract was translated by Luis Benedetto. The drawing of the complete adult
male was done by Michael Miner, a graduate student at the University of Utah.

This work was supported by FWPCA grant No. 1-F-2-WP-26, 393-01,
NSF grant No. GB-7782 and a Sigma Xi grant-in-aid of research to the senior
author.

Amphinemura apache Baumann and Gaufin, new species
Figures 2-5

Male.— Macropterous. Length of forewings 6.0-6.5 mm; length of body
4. 5-5. 5 mm. Body brown; legs yellowish brown, femora dark at tip, tibiae
dark at base. Forewings dusky brown with 35-40 clear rounded spots in cells
distributed regularly over surface; hindwings uniform dusky brown, except
for 1-2 clear spots in costal space. Ninth abdominal tergite produced at pos-
terior margin into blunt raised knob, bearing stout dark spinules. Subgenital
plate rounded, broad at base, tapering gradually to narrow tip, extending to
base of epiproct; lobe at base of 9th sternite four times as long as broad, lateral
margins slightly sinuate, tip rounded. Paraprocts with three sclerotized proc-
esses; inner process narrow, tip blunt, lying alongside and extending slightly
beyond tip of subgenital plate; middle process with large broad base, tapering
abruptly to narrow tip, bearing rows of 16-18 stout spines on anterior third;
outer process long and very thin, bearing 2-4 stout spines at tip (Fig. 4).
Epiproct large and mostly membranous; dorsal aspect rectangular, with deep
narrow sclerotized slit at bilobed tip; lateral aspect narrow at base, becoming
increasingly larger, ending in large bulbous tip, lateral sclerotized band very
narrow, anterior half as dark line; ventral aspect with sclerotized plate, broad
at base, lateral margins sinuate, tapering to narrow tip, bearing 2-4 rows of
short stout spines (Figs. 2, 3a, 3b).

Female.— Macropterous. Length of forewings 7.5 mm; length of body
6.0 mm. Body, appendages and wings similar to male. Seventh sternite very
large and expanded, posterior margin extending over and completely covering

1972

AMPHINEMURA of Western North America

3

middle of 8th sternite, lateral corners formed into large swollen hornlike pro-
jections (Fig. 5). Subgenital plate with median notch and lateral sinuate
margins, posterior-lateral margins produced, with narrow sclerotized band.

Types.™ HOLOTYPE $ and ALLOTYPE $ , Rucker Creek, above Rucker
Lake, Chiricahua Mountains, Cochise Co., Arizona, USA, 18-VII-1968, R. W.
Baumann (LACM). PARATYPES: ARIZONA, Cochise Co., same data as
holotype, 6 $$ (UU) (RWB); Cave Creek, Herb Martyr Campground,
Chiricahua Mountains, 18-VII-1968, R. W. Baumann, 1 $ (dissected from
mature nymph) (RWB); Upper Cave Creek, Chiricahua Mountains,
17-VIII-1970, K. Clarke and D. Sail, 1 A , 1 9 (SWRS); Upper Cave Creek,
below Cave Creek Falls, 23-VIII-1970, V. Roth, 5 $$, 3 2 2 (SWRS)
(RWB).

Amphinemura mogol Ionics

Figure 1. Amphinemura mogollonica, n. sp., adult male.

4

Contributions in Science

No. 226

Figures 2-5. Amphinemura apache, n. sp. 2. Epiproct, lateral view. 3a. Epiproct,
left half, ventral view. 3b. Epiproct, right half, dorsal view. 4. Paraproct, ventral-
lateral view. 5. Female terminalia, ventral view.

Figures 6-9. Amphinemura banksi, n. sp. 6. Epiproct, lateral view. 7a. Epiproct,
left half, ventral view. 7b. Epiproct, right half, dorsal view. 8. Paraproct, ventral-
lateral view. 9. Female terminalia, ventral view.

Figures 10-13. Amphinemura mexicana, n. sp. 10. Epiproct, lateral view. 11a.
Epiproct, left half, ventral view. 1 lb. Epiproct, right half, dorsal view. 12. Paraproct,
ventral-lateral view. 13. Female terminalia, ventral view. (Scale in mm).

1972

AMPH1NEMURA of Western North America

5

Distribution .—Amphinemura apache has been collected only in the
Chiricahua Mountains of Southeastern Arizona. These mountains are known
for their interesting endemic fauna and it is possible that this species is
restricted to this area. The absence of extensive collections from the Amer-
ican Southwest and Mexico, however, makes such an assumption questionable.

Diagnosis. —The male of A. apache has an epiproct which is rectangular
and narrow in dorsal view and enlarged apically in lateral view. It can be
separated from the similar species. A. mexicana and A. venusta, by the broadly
rounded apical portion of the epiproct as seen in lateral view. The apical por-
tion is distinctly angular in A. mexicana and A. venusta and is produced into
a downward directed process. The female can be distinguished by the greatly
expanded seventh abdominal sternite which bears two hornlike processes at
the posterior corners. The females of all other known species have a seventh
sternite which is only slightly expanded with a broadly rounded posterior
margin.

Etymology.— The specific name “apache” is a noun in apposition. It was
chosen because of the importance of the Chiricahua Mountains in the history
of the Apache Indians.

Amphinemura banksi Baumann and Gaufin, new species
Figures 6-9, 23

Nemoura venusta, Needham and Claassen, 1925: 209 (not holotype), 363,
figs. 5-8.

Nemoura ( Amphinemura ) venusta, Ricker, 1952: 27 (in part).

Nemoura ( Amphinemura ) venusta, Gaufin, Nebeker and Sessions, 1966: 34,
35, 37 (distribution); figs. 62, 63, 72.

Amphinemura venusta, lilies, 1966: 189-190 (in part).

Additional references: Nemoura venusta, Dodds and Hisaw, 1925: 382;
Claassen, 1931: 124 (in part); Claassen, 1940: 66 (in part); Gaufin,
1955: 117 (in part); Ricker, 1959: 949 (in part); Gaufin 1964: 222 (in
part); Baumann and Gaufin, 1971: 106 (in part).

Male.— Macropterous. Length of forewings 5. 0-6.0 mm; length of body
5. 0-6.0 mm. Body brown; legs yellowish brown, femora dark at tip, tibiae dark
at base. Forewings dusky brown with 20-25 clear rounded spots in cells beyond
cord, cells between cord and base hyaline, veins brown; hindwings mostly
hyaline, brown area in costal space beyond cord. Ninth abdominal tergite
produced slightly at median-posterior margin, bearing fringe of small dark
spinules. Subgenital plate with broad rounded base, tapering abruptly in
anterior third, extending to base of epiproct, tip rounded; lobe at base of 9th
sternite four times as long as broad, lateral margins parallel, tip rounded.
Paraprocts with three sclerotized processes; inner process fairly broad, bluntly
forked at tip, lying alongside and extending slightly beyond tip of subgenital
plate; middle process with large broad base, apical portion narrow, tip located

6

Contributions in Science

No. 226

on small membranous knob bearing 3-4 stout spines, anterior sclerotized por-
tion with row of 4-5 stout spines; outer process short, base broad, tapering
slightly from angular bend to bluntly rounded tip, bearing 3-5 stout apical
spines (Fig. 8). Epiproct fairly large and mostly membranous; dorsal aspect
as rounded triangle, rounded tip divided by deep narrow sclerotized slit; lateral
aspect quite narrow, width constant throughout, tapering to pointed tip, lateral
parallel sclerotized band narrow, dorsal margin of band at base even with
ventral margin of epiproct; ventral aspect with narrow sclerotized plate,
broad at base, tapering gradually to pointed tip, anterior half bearing triangular
patch of short stout spines (Figs. 6, 7a, 7b).

Female. —Macropterous. Length of forewings 6.5-7. 5 mm; length of
body 6.0-7. 5 mm. Body appendages and wings similar to male. Seventh ster-
nite large, posterior portion broadly rounded and lightly sclerotized, extend-
ing over anterior half of 8th sternite. Subgenital plate with median notch
and rounded lateral sclerotized knob-like projections on posterior margin
(Fig. 9). Vagina with characteristic sclerotized pattern; dorsal aspect almost
square, base slightly broader, lateral-basal corners as lightly sclerotized tri-
angles covering elongate darkly sclerotized areas, apex composed of two
blunt projections which meet at junction of seminal receptacles, projections
ending in blunt tips (Fig. 23).

Types.— HOLOTYPE $ and ALLOTYPE 9, Hidden Valley Creek,
Rocky Mountain National Park, Larimer Co., Colorado, USA, 5-VIII-1953,
A. R. Gaufin (LACM). PARATYPES: ARIZONA, Apache Co., Lukachukai
Creek, Wagon Wheel Campground, 7-VIII-1969, R. W. Baumann, 4 3$,
6 9 9 (RWB). COLORADO, (Rocky Mountain National Park) : Grand Co.,
Onahu Creek, Hwy. 34, 24-VIII-1967, R. W. Baumann, 13,19 (UU).
Larimer Co., Cub Creek, beaver dams, 25-VII-1938, H. H. & J. A. Ross, 1 $
(INHS); Glacier Creek, 17-VIII-1940, T. H. Frison & T. H. Frison, Jr.,
1 $ (INHS); same data as holotype, 10^^,16 9 9 (UU) (RWB); Hidden
Valley Creek, 24-VII-1960, A. R. Gaufin, 1 $ (UU); Fall River, Hwy. 34,
24-VII-1960, A. R. Gaufin, 1 $ ; 24-VIII-1967, R. W. Baumann, 1 $ (UU);
creek, Hwy. 34, near Hidden Valley, 24-VIII-1967, R. W. Baumann, 4 $ 3,
3 9 9 (RWB); Big Thompson River, Moraine Park, 24-VIII-1967, R. W.
Baumann, 1 3,2 9 9 (UU); Mill Creek, near Glacier Basin, 24-VIII-1967,
R. W. Baumann, 29 $ (UU): Glacier Creek, near Bear Lake, 24-VIII-1967,
R. W. Baumann, 1 9 (UU). UTAH, San Juan Co., Pack Creek, Pack Creek
Campground, 8-VIII-1969, R. W. Baumann, 4 $ 3,4 9 9 (RWB). WYOM-
ING, Uinta Co., small creek 2 miles east of Bridger, 21-VII-1967, R. W.
Baumann, 60 $ $, 19 9 9 (RWB).

Additional specimens.— COLORADO, numerous specimens were exam-
ined from the following counties: Boulder, Chaffee, El Paso, Gilpin, Grand,
Jackson, Larimer, Las Animas, Mineral, Rio Blanco, Routt, Summit and
Teller [(MCZ) (USNM) (CU) (UU) (RWB) (INHS) (WER) (CAS).]
IDAHO, Clark Co., 2.5 miles northwest of Kilgore, 15-VII-1956, W. F. Barr,

1972

AMPHINEMURA of Western North America

7

6 $ 8, 15 $ $ (UI). MONTANA, Gallatin Co Hyalite Creek, 9-VIII-1951,
R. Hays and C. I. D, Brown, 1 2 (MSU); West Gallatin River, 9- VIII- 195 1,
R. Hays and C. I. D, Brown, 1 $ (MSU); Beck and Border Canal, 17-VIII-
1951, I. Spindler and W. D. Clothier, 8 $ 8, 6 $ $ (MSU); Allison-Lewis
Ditch, 12- IX- 1951, J. Spindler and W, D. Clothier, 8 8 8, 1 $ (MSU).
Glacier Co Kennedy Creek, 4 miles north of Babb, 13-VII-1963, A. R. Gau-
fin, 1 9 (UU). Judith Basin Co., Martin Creek, 10 miles above Geyser,
7-VII-1966, I. R. Grierson, 1 8 (UU). SOUTH DAKOTA, Lawrence Co.,
Roughlock Falls, near Savoy, Black Hills, 21 -VIII- 1954, M. W. Sanderson,
1 8, 3 2 2 (INKS). UTAH, records checked from the following counties:
Cache, Daggett, Duchesne, San Juan, Summit, Uintah, Utah and Wasatch
i’(WHR) (INHS) (UU) (RWB) (USU) (CAS).] WYOMING, numerous
specimens from the following counties: Albany, Fremont, Johnson, Lincoln,
Park, Sublette, Teton and Uinta [(LACM) (WER) (UU) (CAS) (USNM)
(MSU) (INHS) (UCD).]

Distribution.— yi mphinemura banks i has been recorded from Northern
Montana to Northern Arizona and from Idaho to Colorado. Further collecting
will probably confirm the presence of this species in Northern New Mexico.
A sister species, A. mogollonica, is present in Arizona, New Mexico and
Southwestern Utah but without an overlap in distributional area.

Diagnosis.— Amphinemura banksi is very similar to A. mogollonica. The
males can be separated by the shorter and broader outer lobe of the paraprocts.
The lateral projections of the female subgenital plate are simple and broadly
rounded in A. banksi where they are bilobed and narrowly rounded in A.
mogollonica. The female of A. puebla is also similar but can be distinguished
by the presence of a dark triangular patch over the genital opening.

Remarks.— Needham and Claassen ( 1925) in their Plecoptera monograph
gave descriptions and drawings of a male and female under the name Nemoura
venusta Banks. They included collection records from Colorado, from which
the descriptions and drawings were probably made. These drawings and
descriptions did not agree when compared with the type female of N. venusta
at the Harvard Museum of Comparative Zoology. This left the species figured
without a name.

Etymology .—Amphinemura banksi was chosen in honor of the late Dr.
Nathan Banks, who contributed greatly to the knowledge of the neuropteroid
insects of Western North America.

Amphinemura mexicana Baumann, new species
Figures 10-13

Male.— Macropterous. Length of forewings 6.5-7. 5 mm; length of body
4. 5-6.0 mm. Body brown; legs yellowish brown, femora with 3 dark dorsal
stripes, median stripe short, lateral stripes extending length of femur; tibiae
dark at base; tarsi blackish. Forewings dark brown with 40-45 clear rounded
spots in cells distributed regularly over surface; hindwings uniform dusky

8

Contributions in Science

No. 226

brown, except for 1-2 clear spots in costal space. Ninth abdominal tergite
produced slightly at median-posterior margin, bearing a fringe of small dark
spinules. Subgenital plate with broad oval base, tapering abruptly in anterior
third, extending to base of epiproct, tip broadly rounded; lobe at base of 9th
sternite four times as long as broad, lateral margins parallel, tip rounded.
Paraprocts with three sclerotized processes; inner process fairly broad, forked
at tip, inner prong longer than outer, lying alongside and extending beyond
tip of subgenital plate; middle process with large broad base, tapering to long
narrow anterior portion, tip situated on small membranous knob bearing 1-2
sharp spines, anterior sclerotized portion with 5-7 stout spines; outer process
long and narrow, anterior portion located on broad membranous knob, dorsal
aspect of apex bearing rows of 13-17 stout spines (Fig. 12). Epiproct large
and mostly membranous; dorsal aspect rectangular, with V-shaped notch at
bilobed tip, lobes covered with very small spinules, M-shaped sclerotized
internal structure visible directly behind apex; lateral aspect narrow at base
and greatly enlarged in anterior two-thirds; greatest width near middle, tip
bluntly pointed, lateral sclerotized band narrow, widest at base and apex;
ventral aspect with narrow sclerotized median portion, base broad, tapering
toward apex, with enlargements near middle and slightly behind tip, bearing
large patch of stout spines on anterior two-thirds (Figs. 10, 11a, lib).

Female.— Macropterous. Length of forewings 8. 0-9.0 mm; length of
body 6.0-7. 5 mm. Body, appendages and wings similar to male. Seventh ster-
nite large, lightly sclerotized, posterior portion broadly rounded, extending
over anterior half of eighth sternite. Subgenital plate with median notch,
lateral posterior margins with two sclerotized knoblike lobes, both lobes equal
in size (Fig. 13).

Types.— HOLOTYPE $ and ALLOTYPE $ , La Marquesa, Las Cruces
National Park, Mexico, MEXICO, 5 to 9-VII-1965, Flint and Ortiz (USNM).
PARATYPES: FEDERAL DISTRICT, Desierto de los Leones National Park,
30-VII-1939, 1 $ (SGJ); III to V-1965, N. L. H. Krauss, 1 $ ; X-1965, 1 8,
1 $ (USNM). MEXICO, same data as holotype, 6 $ 8, 10 $ $ (USNM);
La Marquesa, Las Cruces National Park, 13-VII-1966, Flint and Ortiz, 5 8$,

3 9 $ (USNM) (RWB). MORELOS, Laguanas de Zempoala National Park,
18-VIII-1939, 1 $ (SGJ); 10 & ll-VII-1965, Flint and Ortiz, 2^,2$$
(USNM).

Additional specimens.— MICHOACAN, Tuxpan, 8-VII-1965, 6 8 $,

4 $ $ (ENAM). MORELOS, Xochitepec, 14-VII-1965, 1 8,1 9 (ENAM).

Distribution.— Amphinemura mexicana is known only from the moun-
tains of Southern Mexico in the vicinity of Mexico City. The known range
of this species will probably be greatly expanded with intensive collecting
throughout Mexico. Based on present records, this species is the most com-
mon Amphinemura present in Mexico.

Diagnosis.— This species is most similar to A. venucta. The epiproct of
the A. mexicana male has a large angular ventral projection. The ventral proc-

1972

AMPHINEMURA of Western North America

9

ess of the epiproct is narrow and pointed in A. venusta. Females can be sepa-
rated by the shape of the sclerotized knobs on the lateral corners of the
subgenital plate. In A. mexicana, the knobs are equal in size and rounded,
while in A. venusta the inner lobe is large and broadly rounded and the outer
lobe is long and narrow.

Etymology.— The name is derived from Mexico where all specimens have
been collected.

Amphinemura mogollonica Baumann and Gaufin, new species
Figures 1, 14-17, 24

Nemoura venusta, Ricker, 1952: 27 (in part); Ricker 1952: 949 (in part);
Gaufin, 1964: 22 (in part); Gaufin, Nebeker and Sessions, 1966: 35
(in part).

Amphinemura venusta , lilies, 1966: 189-190 (in part).

Male.— Macropterous. Length of forewings 6.0-7. 0 mm; length of body
5. 5-6. 5 mm. Body brown; legs yellowish brown, femora dark at tip, tibiae
dark at base. Forewings dusky brown with 30-35 clear rounded spots in cells
distributed regularly over entire surface; hindwings uniform dusky brown
except for 1-2 clear spots in the costal space (Fig. 1). Ninth abdominal tergite
produced slightly at median-posterior margin, bearing narrow patch of small
dark spinules. Subgenital plate with broad rounded base, tapering abruptly in
anterior third, extending to base of epiproct, tip rounded; lobe at base of 9th
sternite four times as long as broad, lateral margins slightly sinuate, tip
rounded. Paraprocts with three sclerotized processes; inner process fairly
broad, bluntly forked at tip, lying alongside and extending beyond tip of sub-
genital plate; middle process with large base, apical portion narrow, tip
located on small membranous knob bearing 2-4 sharp spines, anterior sclero-
tized portion with row of 4-5 stout spines; outer process long, base broad,
tapering to narrow anterior portion, tip small and rounded, bearing 2-3 stout
apical spines (Fig. 16). Epiproct fairly large and mostly membranous; dorsal
aspect as rounded triangle, rounded tip divided by deep narrow sclerotized
slit; lateral aspect quite narrow, width constant throughout, tapering to
pointed tip, lateral parallel sclerotized band broad at base, dorsal margin of
band at base even with dorsal margin of epiproct; ventral aspect with narrow
sclerotized plate, broad at base, tapering gradually to pointed tip, anterior
two-thirds bearing triangular patch of short stout spines (Figs. 14, 15a, 15b).

Female.— Macropterous. Length of forewings 7. 5-8. 5 mm; length of body
6. 0-8.0 mm. Body, appendages and wings similar to male. Seventh sternite
large, posterior portion broadly rounded and lightly sclerotized, extending
over anterior half of 8th sternite. Subgenital plate with median notch and
bilobed lateral sclerotized projections on posterior margin (Fig. 17). Vagina
with characteristic sclerotized pattern; dorsal aspect with broad base and
broadly rounded apex, lateral-basal corners as lightly sclerotized triangles

10

Contributions in Science

No. 226

partially covering elongate darkly sclerotized areas, apex with two broad
projections, tips rounded above and pointed below, meeting at junction of
seminal receptacles (Fig. 24).

Types.— HOLOTYPE 8 and ALLOTYPE 9 , Christopher Creek, Hwy.
160, Christopher Creek, Gila Co., Arizona, USA, 19-VII-1968, R. W. Bau-
mann (LACM). PARATYPES: ARIZONA, Apache Co., Hall Creek, Hwy.
373, near Greer, 19-VII-1968, R. W. Baumann, 3 ^ (RWB). Cochise Co.,
stream lA mile below Rustler Park spring, Chiricahua Mountains, 27-VIII-
1969, R. & D. Koss 1 8 (RWB). Gila Co., same data as holotype, 21 $ $ ,
7 $ $ (UU) (RWB). Graham Co., Wet Canyon Campground, Graham
Mountains, 13-IX-1952, B. Malkin, 1 $ (CAS); Shannon Campground, Gra-
ham Mountains, 13-IX-1952, B. Malkin, 2 8 8 , 1 $ (CAS). UTAH, Beaver
Co., Birch Creek, below Birch Creek Lake, Kents Lake road, 4-VIII-1969,
R. W. Baumann, 13,19 (RWB). Emery Co., Joes Valley, 6-IX-1945, G. F.
Knowlton, 2 $ $ (WER). Sanpete Co., Ephraim Canyon summit, 6-XX-1945,
G. F. Knowlton, 2 8 8, 3 $ $ (WER). Sevier Co., Fish Lake, 2-IX-1930,
18 (INHS); Seven Mile Creek, above Johnson Valley Reservoir, 24-VIXI-
1962, R. F. Gaufin, 2 $ $ (UU). Washington Co., North Fork Virgin River,
Watchman Campground, Zion National Park, 30-VII-1967, R. W. Baumann,
1 $ (RWB).

Additional specimens.— ARIZONA, Apache Co., 3.8 miles southeast of
Nutrioso, 17-V-1964, S. G. Jewett, Jr., 1 8,299 (dried) (SGJ); Rosey
Creek, Hwy. 373, near Greer, 7-IV-1968, R. W. Baumann, 2 9 9 (dried)
(RWB); 19-V-1970, R. W. Baumann, 1 8,2 9 9 (dried) (RWB). NEW
MEXICO, Grant Co., Pinos Altos, Pinos Altos Mountains, 28-VIII-1951,
E. L. Kessel, 1 $ (CAS).

Distribution.— Amphinemura mogollonica is the most common Arnphine-
mura species in Arizona. It has also been recorded from the Southwestern
parts of New Mexico and Utah.

Diagnosis.— This species is similar to A. banksi but can be distinguished
by the shape of the male paraprocts and the projections on the female sub-
genital plate. The outer lobe of the paraproct is long and thin in A. mogollonica

Figures 14-17. Amphinemura mogollonica, n. sp. 14. Epiproct, lateral view. 15a.
Epiproct, left half, ventral view. 15b. Epiproct, right half, dorsal view. 16. Paraproct,
ventral-lateral view. 17. Female terminalia, ventral view.

Figures 18-21. Amphinemura venusta (Banks). 18. Epiproct, lateral view. 19a.
Epiproct, left half, ventral view. 19b. Epiproct, right half, dorsal view. 20. Paraproct,
ventral-lateral view. 21. Female terminalia, ventral view.

Figure 22. Amphinemura puebla, n. sp., Female terminalia, ventral view.

Figure 23. Amphinemura banksi, n. sp.. Vagina, dorsal view.

Figure 24. Amphinemura mogollonica, n. sp., Vagina, dorsal view.

Figure 25. Amphinemura puebla, n. sp., Vagina, dorsal view. (Scale in mm).

1972

AMPHINEMURA of Western North America

11

and short and blunt in A. banksi. The A. mogollonica female has bilobed
projections and the A. banksi female has single lobed projections. Some vari-
ation exists in the size of the outer lobe in A. mogollonica but usually both
lobes are of similar size.

Etymology.— The name “mogollonica” is taken from the Mogollon Rim
of Arizona.

12

Contributions in Science

No. 226

Amphinemura puebla Baumann, new species
Figures 22, 25

Male.— Unknown.

Female.— Macropterous. Length of forewings 7. 0-8. 5 mm; length of body
6. 0-6. 5 mm. Body brown; legs yellowish brown, femora dark at tip, tibiae
dark at base and tip, tarsi dark. Forewings dusky brown, with 25-30 clear
rounded spots in cells distributed regularly over surface; hindwings uniform
dusky brown. Seventh sternite large, posterior portion rounded and lightly
sclerotized, extending over anterior half of 8th sternite. Subgenital plate with
deep median notch and blunt lateral sclerotized projections on posterior
margin (Fig. 22). Eighth sternite with elongate triangular sclerotized patch
over genital opening. Vagina with characteristic sclerotized pattern; dorsal
aspect short and wide, base broad, apex very broadly rounded, lateral basal
corners as small sclerotized triangles covering elongate oval darkly sclerotized
areas, apex with two narrow projections, tips rounded, meeting at junction of
seminal receptacles (Fig. 25).

Types.— HOLOTYPE 9, 5.2 miles west of Acultzingo (Veracruz), Pue-
bla, MEXICO, 6-VII-1962, J. M. Campbell (INHS). PARATYPES: PUE-
BLA, same data as holotype, 2 9 9 (INHS) (RWB).

Distribution.— Amphinemura puebla is known only from the three type
females from Puebla, Mexico.

Diagnosis.— This species is similar in the female to A. banksi and A.
mogollonica. The shape of the lobes of the subgenital plate is somewhat more
angular in A. puebla but falls within the range of variation found in the above
species. The vagina is, however, quite distinctive and can be recognized by
the ratio of width to length. In A. puebla, the width is nearly twice the length
while in A. banksi and A. mogollonica the width and length are about equal.
The prolonged lobes which meet at the junction of the seminal receptacles are
narrow and of equal length throughout in A. puebla while in A. banksi and
A. mogollonica they are enlarged apically.

Etymology.— The name “puebla” is taken from the Mexican state where
the types were collected.

Amphinemura venusta (Banks)

Figures 18-21

Nemoura venusta Banks, 1911: 337.

Nemoura venusta, Needham and Claassen, 1925: 209 (holotype only).
Nemoura ( Amphinemura ) venusta, Ricker, 1952: 27 (in part).
Amphinemura venusta, lilies, 1966: 189-190 (in part).

Additional references: Nemoura venusta : Claassen, 1940: 66 (in part);

Ricker, 1963: 949 (in part); Ricker, 1950: 205; Gaufin, 1964: 222 (in

part).

1972

AMPH1NEMURA of Western North America

13

Male.— Macropterous. Length of forewings 6.5-7. 0 mm; length of body
5.0-5. 5 mm. Body brown; legs yellowish brown, femora dark at tip, tibiae
dark at base and tip, tarsi black. Forewings deep brown, with 35-40 clear
rounded spots in cells distributed regularly over surface; hindwings uniform
dusky brown, except for 1-2 clear areas in costal space. Ninth abdominal
tergite produced slightly at median-posterior margin, bearing sparse fringe
of small dark hairs, lateral-posterior margins with 2-3 long black hairs. Sub-
genital plate with broad oval base, tapering abruptly in anterior third, extend-
ing nearly to base of epiproct, tip broadly rounded; lobe at base of 9th sternite
four times as long as broad, lateral margins parallel, tip rounded. Paraprocts
with three sclerotized processes; inner process fairly broad, with slight inden-
tation at blunt tip, lying alongside and extending beyond tip of subgenital
plate; middle process broad at base, tapering abruptly to narrow median
portion, tip forked and situated on large membranous bulbous lobe bearing
3-7 sharp spines, anterior sclerotized portion with row of 12-15 stout spines;
outer process fairly long, broad at base, tapering slightly towards apex, with

6- 9 stout spines on blunt tip (Fig. 20). Epiproct large and mostly membra-
nous; dorsal aspect rectangular, deep sclerotized slit at bilobed tip, lobes
bearing few very small dark spinules; lateral aspect narrow at base, tapering
abruptly to slanted angular apex, with large median-ventral projection, lateral
sclerotized band broad at base and tip, narrow medially; ventral aspect with
narrow sclerotized portion, base broad, tapering towards apex, slight enlarge-
ment at anterior third, enlarged area bearing patch of short stout spines
(Figs. 18, 19a, 19b).

Female.— Macropterous. Length of forewings 7. 5-9.0 mm; length of
body 6. 0-8.0 mm. Body appendages and wings similar to male. Seventh
sternite fairly large, lightly sclerotized, broadly rounded and extending over
half of 8th sternite. Subgenital plate with deep median notch, two lateral
knoblike projections on posterior margins, inner projections large and broadly
rounded, outer projections long and very narrow (Fig. 21).

Types.— HOLOTYPE $ , Huachuca Mountains, Cochise or Santa Cruz
Co., Arizona, USA, Oslar (MCZ, #11357). ALLOTYPE $, La Marquesa,
Las Cruces National Park, Mexico, MEXICO, 5 to 9-VII-1965, Flint and
Ortiz (USNM).

Additional specimens.— FEDERAL DISTRICT, St. Rosa Nr., 24-1-1932,
A. Dampf, 1 $ (INHS); Canada Contraras, 14-VI-1947, T. H. Hubbell, 1 8,
1 $ (WER). MEXICO, La Marquesa, Las Cruces National Park, 5 to
9-VII-1965, Flint and Ortiz, 1 4, 2 $ $ (USNM). MICHOACAN, Tuxpan,

7- VIII-1965, 2 $ $ (ENAM).

Distribution.— Amphinemura venusta is known from the United States
by a single record from Southern Arizona (type). The species is recorded
from three Mexican states in the vicinity of Mexico City. The distribution
patterns of Trichoptera species (Flint, 1967) indicate that further collec-
tions in Northern Mexico should fill this distribution gap.

14

Contributions in Science

No. 226

Diagnosis.— The males of this species are easily recognized by the details
of the epiproct and the distinctive bulbous paraprocts. The epiproct of the
most similar species, A. mexicana, has a wide angular ventral projection while
the epiproct of A. venusta is narrow and pointed. Amphinemura venusta is the
only species in this complex which has large membranous enlargements at the
tip of the middle lobe of the paraprocts. The females are similar to A. mexi-
cana but can be distinguished by the thin outer lobes on the median-posterior
margins of the subgenital plate. These lobes are short and broad in A.
mexicana.

Remarks.— Amphinemura venusta was named by Nathan Banks from a
single pinned female. The apparent lack of close examination of the type by
Needham and Claassen and the very general key character by Ricker (1952)
led to the consideration of all specimens of Amphinemura from Western
North America with “windowed” wings under this name.

Key to Species
Males

( puebla unknown)

1. Dorsal aspect of epiproct with broad base and narrow apex; lateral aspect
of epiproct of equal width throughout length; paraprocts bearing 14 or less

spines 2

Dorsal aspect of epiproct with broad base and apex; lateral aspect of
epiproct with narrow base and enlarged apex; paraprocts with 18 or more
spines 3

2. Outer sclerotized process of paraprocts short, broad and blunt at tip

(Fig. 8) banksi

Outer sclerotized process of paraprocts long, narrow and pointed at tip
(Fig. 16) mogollonica

3. Lateral aspect of epiproct broadly rounded at apex, without definite ventral

projection (Fig. 2). apache

Lateral aspect of epiproct angular at apex, with definite ventral projection.

4

4. Ventral projection at apex of epiproct narrow and pointed; middle sclero-
tized process of paraprocts located on large bulbous membranous lobe,
outer sclerotized process broad with large tip (Figs. 18, 20) .... venusta
Ventral projection at apex of epiproct broad and angular; middle sclero-
tized process of paraprocts located on small narrow membranous lobe,
outer sclerotized process narrow with small tip (Figs. 10, 12). . . . mexicana

Females

1. Produced portion of 7th abdominal sternite bluntly forked completely

covering 8th sternite (Fig. 5) apache

Produced portion of 7th abdominal sternite broadly rounded, partially
covering 8th sternite 2

1972

A M PH IN EM URA of Western North America

15

2. Posterior-lateral margin of subgenital plate with one sclerotized projection

on each side (bilobed in mogollonica) 3

Posterior-lateral margins of subgenital plate with two sclerotized projections
on each side 5

3. Projections on subgenital plate with bilobed tip, lobes of about equal size

(Fig. 17) mogollonica

Projections on subgenital plate rounded or slightly angular, sometimes with
small lateral extensions 4

4. Vagina rectangular with broad base, lateral triangles small and dark, ante-

rior projections narrow and rounded at tip; triangular sclerotized patch on
8th sternite over genital opening; projections on subgenital plate angular

(Figs. 22, 25) puebla

Vagina square with broad base and apex, lateral triangles large and light,
anterior projections wide and blunt at tip; sclerotized patch absent from
8th sternite; projections on subgenital plate rounded (Figs. 9, 23). banksi

5. Outer subgenital plate projections equal in size or slightly smaller than

inner projections (Fig. 13) mexicana

Outer subgenital plate projections very narrow, inner projections large

(Fig. 21) venusta

Resumen

El complejo Amphinemura venusta del oeste norteamericano muestra
contener seis especies conocidas de las cuales solo una fue previamente identi-
ficada. La comparacion del holotipo hembra de Amphinemura venusta
(Banks) con los ejemplares disponibles lleva a la rediagnosis de esta especie
en Mexico, siendo la localidad tipica en la parte sur de Arizona el limite boreal
de su distribution. Se describen dos especies A. mexicana y A. puebla de las
cercanias de la ciudad de Mexico. Los especimenes de las Rocalloses llamados
A. venusta (Banks) como resultado de la monografia de Needham y Claassen
( 1925) son llamados A. banksi. Dos especies llamadas A. apache y A. mogol-
lonica provienen del Sudoeste de los Estados Unidos.

Las especies del complejo estan aparentemente restringidas a corrientes
permanentes de agua. En los Estados Unidos, el periodo de vuelo es corto,
extendiendose desde julio a septiembre. Los datos disponibles sobre especies
mexicanas indican que el periodo de emergencia es prolongado y puede exten-
derse durante todo el ano.

Literature Cited

Banks, N. 1911. New species of North American Neuropteroid Insects. Trans.
Amer. Entomol. Soc. 37:335-360.

Baumann, R. W. 1970. The Genus Nemoura (Plecoptera) of the Rocky Mountains.
Ph. D. Thesis. Univ. Utah. 192 p. Univ. Microfilms. Ann Arbor, Mich. (Diss.
Abstr. 3L3068-B).

16

Contributions in Science

No. 226

Baumann, R. W., and A. R. Gaufin. 1971. The Stoneflies (Plecoptera) of the
Wasatch Mountains, Utah. Proc. Utah Acad. Sci., Arts and Lett. 46:106-113
(1969).

Claassen, P. W. 1931. Plecoptera Nymphs of America (North of Mexico). Thomas
Say Found. Entomol. Soc. Amer. 3. 199 p.

1940. A Catalogue of the Plecoptera of the World. Mem. Cornell Agr.

Exp. Sta. 232. 235 p.

Dodds, G. S. and F. L. Hisaw. 1925. Ecological Studies on Aquatic Insects, IV.
Altitudinal Range and Zonation of Mayflies, Stoneflies and Caddisflies in the
Colorado Rockies. Ecology 6:380-390.

Flint, O. S,, Jr. 1967. Studies of Neotropical Caddis Flies, VI: On a collection from
Northwestern Mexico. Proc. Entomol. Soc. Wash. 69:162-176.

Gaufin, A. R. 1955. The Stoneflies of Utah (Checklist). Proc. Utah Acad. Sci., Arts
and Lett. 32:117-120.

1964. Systemic List of Plecoptera of Intermountain Region. Proc. Utah

Acad. Sci., Arts and Lett. 41:221-227.

Gaufin, A. R., A. V. Nebeker and J. Sessions. 1966. The Stoneflies of Utah. Univ.
Utah Publ. Sci. Ser. 14. 93 p.

Illies, J. 1965. Phylogeny and Zoogeography of the Plecoptera. Ann. Rev. Entomol.
10:117-140.

1966. Katalog der rezenten Plecoptera. Das Tierreich, 82. Walter de

Gruyter & Co., Berlin. 632 p.

Needham, J. G. and P. W. Claassen. 1925. A Monograph of the Plecoptera of
America North of Mexico. Thomas Say Found. Entomol. Soc. Amer. 2. 397 p.
Ricker, W. E. 1950. Some Evolutionary Trends in Plecoptera. Proc. Ind. Acad. Sci.
59:197-209.

1952. Systematic Studies in Plecoptera. Ind. Univ. Publ. Sci. Ser. 18.

200 p.

1959. Plecoptera, p. 941-957. In W. T. Edmundson (ed.) Freshwater

Biology. John Wiley and Sons, New York.

Accepted for publication January 14, 1972

NUMBER 227
APRIL 12, 1972

KARYOTYPIC VARIATION AND
EVOLUTION OF THE LIZARDS IN
THE FAMILY XANTUSIIDAE

By Robert L. Bezy

CONTRIBUTIONS IN SCICNCC

8

NATURAL HISTORY MUSEUM • LOS ANGELES COUNTY

CONTRIBUTIONS IN SCIENCE is a series of miscellaneous technical papers
in the fields of Biology, Geology and Anthropology, published at irregular intervals
by the Natural History Museum of Los Angeles County. Issues are numbered sep-
arately, and numbers run consecutively regardless of subject matter. Number 1 was
issued January 23, 1957. The series is available to scientific institutions and scien-
tists on an exchange basis. Copies may also be purchased at a nominal price. Inquiries
should be directed to Virginia D. Miller, Natural History Museum of Los Angeles
County, 900 Exposition Boulevard, Los Angeles, California 90007.

INSTRUCTIONS FOR AUTHORS

Manuscripts for CONTRIBUTIONS IN SCIENCE may be in any field of
Life or Earth Sciences. Acceptance of papers will be determined by the amount
and character of new information. Although priority will be given to manuscripts
by staff members, or to papers dealing largely with specimens in the collections of
the Museum, other technical papers will be considered. All manuscripts must be
recommended for consideration by the curator in charge of the proper section or
by the editorial board. Manuscripts must conform to those specifications listed
below and will be examined for suitability by an Editorial Committee including
review by competent specialists outside the Museum.

Authors proposing new taxa in a CONTRIBUTIONS IN SCIENCE must
indicate that the primary type has become the property of a scientific institution of
their choice and cited by name.

MANUSCRIPT FORM.— ( 1 ) The 1964 AIBS Style Manual for Biological
Journals is to be followed in preparation of copy. (2) Double space entire manu-
script. (3) Footnotes should be avoided if possible. Acknowledgments as footnotes
will not be accepted. (4) Place all tables on separate pages. (5) Figure legends and
unavoidable footnotes must be typed on separate sheets. Several of one kind may
be placed on a sheet. (6) An abstract must be included for all papers. This will be
published at the head of each paper. (7) A Spanish summary is required for all
manuscripts dealing with Latin American subjects. Summaries in other languages
are not required but are strongly recommended. Summaries will be published at
the end of the paper. (8) A diagnosis must accompany any newly proposed taxon.
(9) Submit two copies of manuscript.

ILLUSTRATIONS. — All illustrations, including maps and photographs, will
be referred to as figures. All illustrations should be of sufficient clarity and in the
proper proportions for reduction to CONTRIBUTIONS page size. Consult the 1964
AIBS Style Manual for Biological Journals in preparing illustration and legend copy
for style. Submit only illustrations made with permanent ink and glossy photo-
graphic prints of good contrast. Original illustrations and art work will be returned
after the manuscript has been published.

PROOF.— Authors will be sent galley proof which should be corrected and
returned promptly. Changes in the manuscript after galley proof will be billed to
the author. Unless otherwise requested, page proof also will be sent to the author.
One hundred copies of each paper will be given free to each author or divided
equally among multiple authors. Orders for additional copies must be sent to the
Editor at the time corrected galley proof is returned. Appropriate order forms will
be included with the galley proof.

Virginia D. Miller
Editor

KARYOTYPIC VARIATION AND EVOLUTION
OF THE LIZARDS IN THE FAMILY XANTUSIIDAE

By Robert L. Bezy1

Abstract: Diploid chromosome numbers of ten species of
the family Xantusiidae range from 36 to 40 with 16 to 18 macro-
chromosomes, 18 to 22 microchromosomes, and 50 to 58
chromosome arms. Seven pericentric inversions, the loss of two
pairs of microchromosomes, two centric fusions, and the forma-
tion of satellites on one pair of chromosomes explain the varia-
tion observed. Intraspecific karyotypic variation occurs in Xan-
tusia vigilis and Xantusia henshawi. Chromosomal differences
suggest that Lepidophyma smithi and Lepidophyma occulor are
specifically distinct. Chromosomal similarities are consistent with
the inclusion of ( 1 ) Klauberina riversiana in the genus Xantusia,
and (2) Gaigeia gaigeae in the genus Lepidophyma. Of the sev-
eral groups of lizards that have been considered related to xan-
tusiids, the microteiids have the most similar karyotypes. At pres-
ent, there is no evidence to indicate that hybridization preceded
the evolution of unisexuality in Lepidophyma flavimaculatum
from Panama and Costa Rica, in that (1) the karyotype is
primarily diploid and homomorphic; and (2) there are no plausi-
ble parental species known to occur in the area.

INTRODUCTION

In Camp’s (1923) monumental classification of lizards, the species of
the family Xantusiidae bridged the morphological gap between the two divi-
sions (Ascalabota and Autarchoglossa) of the suborder Sauria, a systematic
dilemma which he resolved by arbitrarily depositing them in the Autarcho-
glossa. Subsequent workers have also found this morphologically ambivalent
family annoying and have shifted it between these divisions. In actuality, these
lizards may well be relicts of the departure point of the two major lines of
saurian evolution and thus might reasonably be placed in a third division, a
taxonomic honor which many systematists might be hesitant to bestow on this
small family.

Not only have xantusiid lizards been troublesome to students of
“higher classification,” but those unforutnate taxonomists who have been
lured into extensive studies of the systematics of the family have suffered
greater torments. Within this handful of species there occurs nearly every
conceivable degree of morphological divergence. Many problems are encoun-
tered by a systematist attempting to define subspecies, species, and genera in

1 Associate Curator of Herpetology, Natural History Museum of Los Angeles
County, Los Angeles, Calif. 90007.

1

2

Contributions in Science

No. 227

this small family because the morphological differences between populations
do not tend to fall into discrete sizes that can be easily assigned rank. In par-
titioning this array of only about 14 species into genera, one must steer
between the Scylla of monotypic genera and the Charybdis of a monotypic
family. Cope (1895) recognized five Recent genera, all of which were mono-
typic except Xantusia , and one of which ( Amoebopsis gilberti) contained
what is currently recognized as only a subspecies ( Xantusia vigilis gilberti ).
Savage (1963) recognized four Recent genera of which two ( Xantusia and
Lepidophyma) were polytypic and two ( Cricosaura and Klauberina ) were
monotypic. In this study, these lizards are treated as two groups: Xantusia
(inclusive of Klauberina ) and Lepidophyma (inclusive of Gaigeia)\ Crico-
saura typica has not yet been studied karyotypically.

Xantusiids have extremely disjunct distributions, a characteristic gen-
erally attributed to primitive, receding groups. Ranges of most of the species
are extremely fragmented and populations are often isolated by hundreds of
miles. Particularly spectacular examples are the occurrence of Xantusia vigilis
and Xantusia henshawi in Durango, Mexico, ca. 400 to 800 air-line miles
southeast of the nearest known populations of these species (Webb, 1965,
1970) and the insular isolation of Xantusia riversiana and Cricosaura typica.
The occurrence of the Eocene fossil, Paleoxantusia fera (Hecht, 1956), in
Wyoming, ca. 300 miles north of the present northern limit of the family,
adds a time dimension to the receding of xantusiids.

Sympatric contacts have been reported for only two pairs of currently
recognized species in the family Xantusiidae: Xantusia henshawi and X.
vigilis in southern California (Klauber, 1931) and Durango, Mexico (Webb,
1970) and Lepidophyma tuxtlae and L. pajapanensis in southern Veracruz
(Werler, 1957). When the lack of sympatry in this family is combined with
extreme variability in morphological divergence at the population level, the
task of defining evolutionarily meaningful (or even morphologically con-
sistent) species becomes difficult (Bezy, 1967b). Further, strong selective
pressure for saxicolous adaptations in highly isolated populations of xantu-
siids has led to morphological convergence at the subspecies level ( Xantusia
vigilis arizonae and X. v. sierrae, Bezy, 1967a, b), at the species level ( Xan-
tusia vigilis arizonae and X. henshawi, Klauber, 1931 ), and at the near-generic
level ( Xantusia and Gaigeia, Smith, 1939).

This analysis of karyotypic variation has been undertaken in the hope
of finding new data to help establish meaningful phylogenetic relationships
in this small but puzzling family. Karyotypes of ten species of xantusiids are
reported and discussed herein: Xantusia henshawi Stejneger, X. vigilis Baird,
X. riversiana Cope, Lepidophyma flavimaculatum A. Dumeril, L. gaigeae
Mosauer, L. micropholis Walker, L. occulor Smith, L. pajapanensis Werler,
L. smithi Bocourt, and L. tuxtlae Werler and Shannon. The biogeographical,
morphological, and karyotypic information indicates that these are all valid
species as will be discussed in a separate paper on the systematics of the genus

1972

Karyotypic Evolution of the Xantusiidae

3

Lepidophyma. Karyotypic data are not yet available for five rare forms of
uncertain status: Cricosaura typica Gundlach and Peters, Lepidophyma don-
tomasi (Smith), L. radula (Smith), L. sylvaticum Taylor, and an undescribed
species of Lepidophyma from Guatemala.

I wish to emphasize that the karyotype data can be meaningfully inter-
preted only by comparison with information from other sources, that is, by
the process which Hennig (1966) dignified with the term “reciprocal illu-
mination.” I consider the comparison of patterns emerging from data of
radically different sources to be a vital step in the establishment of meaning-
ful phylogenetic relationships, and do not accept Sokal and Sneath’s (1963)
view that this is merely circular reasoning. Convergence, for example, can
occur in morphology and in karyotypes, but, because of the radically different
factors governing morphological and karyotypic evolution, the probability
is quite low that convergence between two taxa will occur in both parameters.
For these reasons data on morphological variation are discussed in this paper
where the major focus is on karyotypic evolution. Moreover, the phylogenetic
relationships suggested herein are based not only on an appraisal of data
from both of these sources, but also on biogeographical and ecological field
impressions.

MATERIALS AND METHODS

Chromosomes of cells from bone marrow, spleen, and testicular tissue
were prepared in vivo by Patton’s (1967) modification of the colchicine-
hypotonic citrate technique of Ford and Hamerton (1956) as has been
adapted for lizards by Lowe and Wright (1966) and by Lowe, Wright, and
Cole (1966). The karyotype of Lepidophyma flavimaculatum was also deter-
mined in vitro from lung tissue culture by Dr. T. C. Hsu of the M. D. Ander-
son Hospital and Tumor Institute of Houston.

Good karyotype preparations were especially difficult to obtain from
xantusiid lizards due, in part, to an unusually low level of mitotic activity in
the bone marrow. By increasing the stressing of the peripheral circulatory
system, mitotic activity was increased; unfortunately, this also increased the
mortality among the lizards. The limbs of Xantusia vigilis and Lepidophyma
gaigeae are quite small, and the bone marrow is consequently difficult to
“flush out.” Pooling of the bone marrow from several individuals was neces-
sary to obtain the somatic karyotype of L. gaigeae , while the karyotype of
populations of X. vigilis was derived primarily from study of testicular tissue.

Whenever possible, a minimum of at least ten cells was studied from
each specimen “run.” For each cell, the permanent slide number, the cell
coordinates, the diploid chromosome number (2 n), the number of macro-
chromosomes (macros) and microchromosomes (micros), the occurrence of
secondary constrictions, and the numbers and relative sizes of metacentric
(M), submetacentric (SM), sub telocentric (ST) and telocentric (T) macro-

4

Contributions in Science

No. 227

chromosomes were recorded. The karyotype of the specimen was then deter-
mined on a modal basis.

For the family Xantusiidae the following classification of chromosomes
was found to be the most useful and was employed throughout the study:
metacentric S/L (= ratio of short to long arm of chromosome), 0.76-1.00;
submetacentric S/L, 0.51-0.75; subtelocentric S/L, 0.01-0.50; and telocen-
tric S/L, 0.00. Both pairing and classifying the chromosomes, however, was
done “by eye” rather than by actual measurement. In counting chromosome
arms (CA), metacentric to subtelocentric macrochromosomes were consid-
ered bi-armed, while telocentric macrochromosomes were considered uni-
armed. Because I could not consistently distinguish their centromere positions,
all microchromosomes were considered uni-armed.

KARYOTYPE DESCRIPTIONS

Xantusia vigilis. Study of 525 cells from 30 individuals (29$ , 1 $ ) repre-
senting eleven populations (including X. v. arizonae, X. v. extorris , X. v.
sierrae , and X. v. vigilis ) indicates that the 2 n of this species is 40, with 18
macros and 22 micros (Tables 1 and 2, Fig. 1). The macro pairs were num-
bered from largest to smallest (Fig. 1); the micro pairs were not numbered
as their small size precluded recognition of individual pairs. Pair 1 is by far

II 0 II l! II II U tl

23 456789

10-20

11 li

11 l»

11 •«

4 5

B

10-20

Figure 1. Karyotypes of Xantusia vigilis. A. Karyotype UAZ 24216, $ , 1 1.3 mi
(by Hwy 93) SE Burro Creek, 3200 ft, Yavapai Co., Arizona. Line represents 10 ti.
B. Karyotype /3; UAZ 24861, $ , vie. Yamell, 4750 ft, Yavapai Co., Arizona.

1972

Karyotypic Evolution of the Xantusiidae

5

the largest in the complement and is metacentric to submetacentric. Pair 2 is
about half the size of pair 1 and is consistently metacentric. Pair 3 is only
very slightly smaller than pair 2 and is consistently subtelocentric. On the
basis of size and centromere position these first three pairs are always clearly
distinguishable from one another and are distinctly larger than the remaining
six pairs. Pairs 4 and 5 are larger and more distinctly bi-armed than the last
four pairs (6-9). Pair 4 is subtelocentric and pair 5 is submetacentric. Pairs
6, 7 and 8 are nearly identical in size and are subtelocentric; the largest (6),
however, has only minute short-arms and thus occasionally appears telocentric.

If II VI li li la ** i*

1 2 3 4 5 6789

• • •'# •• ,* * * ,,,,.4, ;

. 10-20

U 41 <» u It

2 3 4 S 6

4« m in

7 8 9

•• *• »•

ft 41

B

10-20

II ii li aft ift aa a*

mm mm

c

• • M m* •• ..

10-20

••

Figure 2. Karyotypes of two species of Xantusia. A. X. riversiana ; UAZ 21688,
$ , N end of San Clemente Island, Los Angeles Co., California. Line represents
10 /a. B. X. henshawi; karyotype .a; LACM 72325, $ , 6.5 mi NE Pedricena, Durango,
Mexico. C. X. henshawi', karyotype /?; UAZ 21694, $ , 2 mi (by rd to Idyllwild)
S Banning, San Jacinto Mts., Riverside Co., California.

6

Contributions in Science

No. 227

The smallest pair (9) varies among the populations of Xantusia vigilis studied.
It appears telocentric (karyotype a. Fig. 1) in individuals from eight popula-
tions ( X . v. sierrae; X. v. vigilis from the Mohave and Sonoran Deserts in
Arizona, California, and Baja California), and subtelocentric (karyotype /3 )
in three populations (X. v. arizonae ; X. v. extorris\ and X. v. vigilis from
Desemboque, Sonora).

Xantusia henshawi. Study of 117 cells from 8 individuals (6$, 2$)

11 II II 12 ftl II Ift a*

A

10-19

ill!

ft! IS ift ftft ii A*

1 2 3 4 5 6 7 8 9

a* .*:*

10—19

II It M 11

Mm

1 2 3 4 5 6 7 8 9

10-19

in

IM Mf IM

1 2 3 4 5 6 7 8 9

D

10-19

Figure 3. Karyotypes of Lepidophyma flavimaculatum. A. Bisexual population;
UAZ 28805, $ , 25 mi (by rd to Malpaso) NW Ocozocoautla, Chiapas, Mexico.
B. Unisexual population; UAZ 27642, $ , 3 mi (air line) SE Achiote, Canal Zone,
Panama. C. Unisexual population. Diploid cell from UAZ 27640, $ , same locality
as UAZ 27642, above; line represents 10 /x. D. Unisexual population. Triploid cell
from UAZ 27640.

1972

Karyotypic Evolution of the Xantusiidae

7

from two populations ( X . h. henshawi and X. h. bolsonae) indicates that the
2 n of this species is 40, with 18 macros and 22 micros (Tables 1 and 2, Fig.
2). The karyotype of X. h. bolsonae (= a) appears identical to the (3 karyo-
type of X. vigilis, while that of X. h. henshawi ( = /3) differs in that pair 7
has longer short-arms and is submetacentric. Matthey (1931) reported that
Xantusia henshawi has a 2 n of 42 with 18 macros and 24 micros. Until his
count can be verified, I prefer to disregard it.

Xantusia riversiana. Study of 135 cells from 9 individuals (4 3,5$) of
one population indicates that the 2 n of this species is 40 with 18 macros and
22 micros (Tables 1 and 2, Fig. 2). The karyotype appears identical to the
/ 3 karyotype of X. vigilis.

Lepidophyma flavimaculation. Study of 276 cells from 10 individuals
(03, 10$) representing three populations (bisexual L. /. flavimaculatum

XK xx M IH ft* A! ft* All

1 2 3 456789

* • ** •• •• •• ** •• ♦ ♦ *• g!<

A 10-19

IS 18 M M s* aa it at to

1 2 34567 89

B -J22S-

ft& II IS iA AA AA AA

1 2 3456789

*• 9A mm mm mm •• .» ..

g 10-19

Figure 4. Karyotypes of three species of Lepidophyma. A. L. tuxtlae. UAZ 28770,
3, 2 mi (by rd) SE Sontecomapan, Veracruz, Mexico. B. L. pajapanensis. UAZ
28810, 3, same locality as L. tuxtlae, above. Line represents 10 n. C. L. gaigeae.
UAZ 28868-73, $ , 2 mi N Durango, Hidalgo, Mexico.

8

Contributions in Science

No. 227

from Chiapas and unisexual L. /. obscurum from Panama and Costa Rica)
indicates that the 2 n of this species is 38 with 18 macros and 20 micros (rather
than 22 as in Xantusia ; Tables 1 and 2, Fig. 3). The macros in this species
appear identical in morphology to those of the a. karyotype of Xantusia vigilis
except that pair 3 bears a distinct terminal satellite. The karyotypes of the
unisexual populations appear to be homomorphic and identical to those of
the bisexual population. However, bone marrow tissue of one individual from
the all-female population in Panama appears to be composed of both diploid
(2 n = 38) and triploid (3n = 57) cells (Fig. 3). Eighty-two diploid and 25
triploid cells were examined from one bone marrow preparation, yielding a
ratio of 3.28 diploid to 1 triploid. This condition was observed in only one of
the 8 individuals studied from this all-female population. The karyotype of
another individual from this same population was also determined in vitro
from lung tissue culture by T. C. Hsu and found to be identical to the diploid
bone marrow cells.

Lepidophyma pajapanensis. Study of 87 cells from 4 individuals ( 1 $ ,
3 9 ) of one population indicates that the 2 n of this species is 3 8 with 1 8
macros and 20 micros (Tables 1 and 2, Fig. 4). The macros appear identical
to those of L. flavimaculatum.

Lepidophyma tuxtlae. Study of 200 cells from 8 individuals (5$, 3$)
representing two populations (Veracruz and Chiapas) indicates that the 2 n
of this species is 38 with 18 macros and 20 micros (Tables 1 and 2, Fig. 4).
The karyotype of this species also appears identical to that of L. flavimacula-
tum. No differences were found between the two populations of L. tuxtlae.

Lepidophyma gaigeae. Study of 77 cells from 4 individuals (2$, 2$)
of one population indicates that the 2 n of this species is 38 with 18 macros
and 20 micros (Tables 1 and 2, Fig. 4). The morphology of the macros
appears identical to that in L. flavimaculatum except that: (1) pair 7 has
longer short-arms, appearing submetacentric more often than subtelocentric;
(2) pair 9 is subtelocentric rather than telocentric.

Lepidophyma micropholis. Study of 83 cells from 3 individuals ( 2$,
1 9 ) of one population indicates that the 2 n of this species is 36 with 16
macros and 20 micros (Tables 1 and 2, Fig. 5). The macros appear identical
to those of L. flavimaculatum, except that: (1) pair 2 A is a large metacentric
that probably was formed by the fusion of pairs 6 and 8; (2) pair 3 lacks
terminal satellites; (3) pair 7 is submetacentric to metacentric, thus resembling
pair 7 of L. gaigeae.

Lepidophyma smithi. Study of 151 cells from 7 individuals (4$, 39)
representing two populations (L. s. smithi and L. s. tehuanae ) indicates that the
2 n of this species is 36 with 16 macros and 20 micros (Tables 1 and 2, Fig. 5).
The macros appear identical to those of L. flavimaculatum except that pair 2A
is a metacentric to submetacentric and probably was formed by centric fusion
of pairs 6 and 9; thus only its long-arms are homologous with pair 2 A of L.
micropholis. That chromosome pair 2A is formed by fusion of pairs 6 and 8

1972

Karyotypic Evolution of the Xantusiidae

9

in L. micropholis and pairs 6 and 9 in L. smithi is conjectured from the
following: (1) pair 2 A appears somewhat more submetacentric in L. smithi
than in L. micropholis ; (2) the smallest chromosome pair in L. micropholis
usually appears slightly smaller than the smallest pair in L. smithi, and is
telocentric in the former and subtelocentric in the latter. All of these differ-
ences could also be explained as resulting from inversions occurring after
one centric fusion, except the difference in the size of the smallest chromo-
some pair. This could be made more concrete by comparing measurements
from photomicrographs of the karyotypes of the two species, but the size

HI XX XX ftX M XA U A*

1 2 6 + 8 3 4 5 7 9

9-18

0 ii ii

It l| in 18 ftft

1

ft *

6+9

9 -18

B

Xt M A-*- Kx M Xi m *x

10-18

c

Figure 5. Karyotypes of three species of Lepidophyma. A. L. micropholis. UAZ
28762, $ , cave at El Pachon, 8 km (by rd) NNE Antigua Morelos, Tamaulipas,
Mexico. Line represents 10 ft. B. L. smithi. UAZ 28812, $, 4 mi NW Mapastepec,
Chiapas, Mexico. C. L. occulor. TCWC 35605, $ , 2.5 mi S Conca, 2000 ft, Quere-
taro, Mexico.

10

Contributions in Science

No. 227

differences involved are so small that truly convincing identification of
homologous chromosomes would probably require observation of synapsis
in artificially produced hybrids.

Lepidophyma occulor. Study of 101 cells from one male indicates that
the 2 n of this species is 36 with 18 macros and 18 micros (Tables 1 and 2,
Fig. 5). The macros are identical to those of L. flavimaculatum, except that
(1) pair 3 lacks terminal satellites; (2) pairs 7 and 8 are submetacentric
instead of subtelocentric; (3) pair 9 is submetacentric instead of telocentric.

DISCUSSION

Construction of the Karyotype Phytogeny:

The special utility of karyotype information in the study of systematics
and evolution lies in three things : ( 1 ) since differences in chromosome num-
ber and form can result in decreased fertility or even sterility of hybrids, detec-
tion of karyotypic differences between two taxa increases the probability that
they are not conspecific; (2) because chromosomal and morphological changes
result from different evolutionary mechanisms, comparisons of the relation-
ships indicated from karyotype analyses with those from other sources of
systematic information (e.g. morphology, behavior, immunology, electropho-
resis) aids in the detection of convergence; and (3) because some chromo-

Figure 6. Phylogeny of the karyotypes of ten species of the family Xantusiidae.
The symbols in the parentheses indicate the derived states occurring in each of the
karyotypes: 18m and 20m = reductions in number of micros; 3 = formation of
satellites on this pair; 6 + 8 and 6 + 9 = centric fusions of macros; 7, 8, 9 = peri-
centric inversions shifting the position of the centromeres on these macros. The
numbers beneath the concentric half circles indicate the total number of derived
states in each of the karyotypes. Data from Tables 1 and 2.

1972

Karyotypic Evolution of the Xantusiidae

11

somal changes appear to be much more common than others, designation of
primitive and derived character states is possible.

Although many cogent criticisms of Hennig’s ( 1966) theory and methods
have been presented (Darlington, 1970), he has, if nothing else, re-empha-
sized the necessity of identifying primitive (plesiomorphic) and advanced
(apomorphic) character states before constructing phylogenies. In the formu-
lation of karyotype phylogenies of lizards, two approaches have been taken
to estimate the direction of evolution. One approach is to regard as primitive
that karyotype which occurs most widely among the families of lizards and
to derive all other karyotypes from this, using whatever cytogenetic mech-
anisms (centric fusion, centric fission, and inversions) are required (Gorman,
Atkins, and Holzinger, 1967; Gorman, Huey, and Williams, 1969; Gorman,
1970).

The second approach to the construction of karyotype phylogenies is
based on the evidence indicating that centric fusions are of much more
common occurrence than fissions (Hsu and Mead, 1969). Earlier cytogenetic
studies of vertebrates, especially lizards, have considered centric fusion (whole
arm translocation or Robertsonian fusions; Matthey, 1951; White, 1954) to
be the predominant mechanism of chromosomal rearrangement. More recently
this approach has been applied to the genus Sceloporus (Lowe, Cole, and
Patton, 1967; Cole, 1970) and Cnemidophorus (Lowe, Wright, Cole, and
Bezy, 1970a). I have elected to utilize this approach in the present study
because: (1) I feel the available evidence indicates that fissions are uncom-
mon, and (2) the small number of taxa and karyotypes in the family Xan-
tusiidae makes it difficult and highly arbitrary to select any one karyotype as
being the most common or widespread in the family.

I thus prefer to consider karyotypes with higher diploid numbers and
higher percentages of acrocentric chromosomes to be primitive, and to derive
karyotypes from these by centric fusion and pericentric inversions, invoking
centric fission only in those specific instances where there is compelling evi-
dence that it has occurred (Lowe, Cole, Wright, and Bezy, 1970b).

In spite of the fact that the paracentric inversions of Drosophila salivary
gland chromosomes form the basis for perhaps the most concrete phylogenies
yet constructed, it is difficult to assign directionality to the unequal pericentric
inversions that are presumed to be responsible for the shifts in centromere
positions of the chromosomes in the karyotypes of xantusiids. However, as
in the case of centric fusions, the general evolutionary trend in karyotypic
evolution is that pericentric inversions tend to convert uni-armed chromo-
somes into bi-armed chromosomes, not vice versa (White, 1954:192). As
with centric fusions, unequal pericentric inversions reduce the number of
acrocentrics and increase the number of subtelocentric to metacentric
chromosomes.

Thus, in constructing the karyotype phylogeny (Fig. 6) for each chromo-
some I have always considered the most nearly acrocentric condition observed

Table 1. Variation in the chromosomes of ten species in the family Xantusiidae. Centromere position (M = metacentric,
SM = submetacentric, ST = subtelocentric, T = telocentric) and presence of satellites (*) for the macrochromo-
some pairs. Centromere positions in parentheses are those observed less frequently for the chromosome pair.

12

Contributions in Science No. 227

C/3

H

H

H

H

H

1

o\

H

oo

C/3

00

00

H

H

H

03

H

1

03

H

H

H

H

H

H

H

H

I

H

s

<x>

00

C/3

00

03

03

03

03

03

03

1

03

03

f-H

H

03

P

03

s

s

H

H

P

H

s

H

H

H

s

s

H

s

t"-

C/3

C/3

oo

00

03

03

03

03

03

03

03

03

P

H

p

H

H

H

H

H

H

H

H

H

H

VO

OO

C/3

oo

oo

00

03

03

03

03

1

1

03

s

s

s

s

S

s

s

s

s

s

s

S

03

C/3

03

03

oo

03

03

03

03

03

03

03

H

H

H

H

H

H

H

H

H

H

H

H

T

C/3

00

C/3

03

03

03

00

03

03

03

03

03

H H H H H

03 CO CO 03 03

H H H

C/3 C/3 C/3

H H H H
w m w w

MISS

s s s s s

s s s s s s s

s s s s s

C/3 C/3 C/3 C/3 C/3

s s s s s

s s s

C/3 C/3 C/3

s s s s

C/3 C/3 C/3 C/3

s s s s s s s

§ 5 5

&0 &Q

a oa.

&

s .<*

« .ft E3

& 1 g
"I s S, j*
11-11

o

-si

ft.

! |

s i

Table 2. Summary of karyotypic variation in ten species of the family Xantusiidae. Diploid chromosome number (2 n)\
number of macrochromosomes (macros); number of microchromosomes (micros); number of pairs of metacentric
(M), submetacentric (SM), subtelocentric (ST), and telocentric (T) macrochromosomes; presence ( + ) or ab-
sence ( — ) of satellites (Sats) on macrochromosome pair 3; number of chromosome arms (CA); and total derived
states (TDS).

1972

Karyotypic Evolution of the Xantusiidae

13

o

CH M M t ri m ^

\c oo oo oo oo
»o

^ in vi vi ^

I I I I

+ + + + I + I

O O O ©

-h — I o -• o o

>n 'O 'C vo »n

in in in in N t t'l

<N

CS CS rt

CS CS CN <N CN

N N M m tn N

M N N r) N
M (S N M (S

O O O O O O 00
<N <N CS CN — 1

OO 00 OO 00 00

OO 00 00 OO IO o oo

o © o o o
^ 'T ^ 'T

00 OO OO oo 'O VO

rn m m m cn m m

a

•S3

.§3

<52.

‘I

<3

-s;

ex

*§

I

>•

•S3

<X>

s*

a £

o, .2

*

a s

o

-s:

cx

b *

occulor

14

Contributions in Science

No. 227

among the various forms to be the primitive condition for that chromosome
and have considered fused chromosomes to be a derived condition. From this
line of reasoning, primitive karyotypic states in the family are: (1) a 2n of
40; (2) 22 micros; (3) 18 macros; (4) pairs 1 and 2, metacentric; (5) pair 5,
submetacentric; (6) pairs 3, 4, 6, 7, and 8, subtelocentric; (7) pair 9, telo-
centric; and (8) no satellites. All of these states are present in the a karyotype
of Xantusia vigil is.

From this primitive condition, the observed karyotypes can be derived
by centric fusions and pericentric inversions using those pathways that would
require the minimum number of chromosomal rearrangements and yet pro-
duce the minimum amount of karyotypic convergence (Fig. 6). A total of
seven pericentric inversions, two fusions of macros, two fusions or losses of
micros, and one instance of satellite formation is required to account for the
chromosomal evolution observed thus far in the family Xantusiidae; a total of
four instances of chromosomal convergence result (chromosomal convergence
occurs when a specific derived state of a given chromosome is independently
evolved in separate lineages). The phytogeny (Fig. 6) is superimposed on a
scale (total derived state or TDS) that is simply the total number of character
states in each karyotype that can be considered to be derived.

Species:

Although recognized species were used to some extent as guides for the
sampling of populations of xantusiids for chromosomal variation, I have
attempted to study as many populations as possible of each of the species.

Two karyotypes (a and /3 ) were observed among the eleven populations
of Xantusia vigil is. The more primitive karyotype (a) occurred in seven
populations of X. v. vigil is from the Mohave and Sonoran Deserts of Califor-
nia, Arizona, and extreme northern Baja California (for localities see Speci-
mens Examined ) and in X. v. sierrae from the foothills of the Sierra Nevada in
the Central Valley of California. The derived karyotype (/ 3 ) was found in
the three most eastern populations sampled: X. v. vigilis from Desemboque,
Sonora, Mexico; X. v. arizonae from Yarnell near the southern edge of the
Colorado Plateau in Arizona; and X. v. extorris from Durango, Mexico.

The similarity of the karyotype of X. v. sierrae to X. v. vigilis rather
than to X. v. arizonae tends to substantiate the hypothesis (Bezy, 1967a)
that the two races specialized for living under granite spalls ( arizonae and
sierrae ) were derived independently from the widespread yucca-dwelling race
( X . v. vigilis). The apparent lack of correspondence of chromosomal races
with morphological subspecies of X. vigilis is interesting, and karyotypic
studies of the other subspecies ( gilberti , utahensis, wigginsi) are planned.

The two populations of Xantusia henshawi studied also had karyotypic
differences that would appear to involve one pericentric inversion. The more
primitive karyotype (a) occurs in X. h. bolsonae from Durango, Mexico,
while the more advanced karyotype (/ 3 ) occurs in the morphologically more

1972

Karyotypic Evolution of the Xantusiidae

15

specialized X. h. henshawi from southern California. Chromosomal differences
of this magnitude have been found in a single population of Sceloporus clarki
(Cole, 1970) and thus may not constitute an effective reproductive barrier.

Two forms that were considered by Walker (1955) to be subspecies of
L. flavimaculatum have different chromosome numbers: L. occulor (2 n of
36 with 18 macros and 18 micros) and L. smithi (2 n of 36 with 16 macros
and 20 micros). The three populations of L. flavimaculatum studied have a
2n of 38 with 18 macros and 20 micros. Such chromosomal differences rarely
occur within species and may constitute genetic isolation mechanisms. Mor-
phological and biographical data that also indicate these are distinct species
will be presented in a separate paper on the systematics of the genus
Lepidophyma.

Genera:

Mayr (1969:92-94) listed several criteria of an “ideal” genus: (1)
monophyly; (2) separation from other genera by a morphological gap, the
size of which is inversely proportional to the number of included species;
(3) reasonable internal homogeneity; and (4) occupation of a distinctive
adaptive zone. Application of these criteria to genera of xantusiids is made
difficult by several factors. Convergence appears to be unusually common
in the family, increasing the difficulty of accessment of monophyly. Because
of the small number of xantusiid species, it is difficult to judge what size of
a morphological gap should delineate a genus. Due to their secretive habits,
little is known of the adaptive zones of xantusiids.

Comparisons of karyotypic phylogenies with those resulting from mor-
phological analyses are quite useful in making decisions about monophyly
and convergence, because radically different factors govern morphological
and chromosomal evolution. However, for this same reason, caution must
be employed in formulating generic classifications based entirely on homo-
geneity and gaps in chromosomal variation. For example, relying exclusively
on the chromosomal data, the 10 species in this study would be partitioned
into the following groupings: (1) X. henshawi, X. river siana, X. vigilis', (2)
L. occulor ; (3) L. micropholis\ (4) L. flavimaculatum , L. tuxtlae, L. paja-
panensis, L. smithy, and (5) L. gaigeae. Although these groupings appear to
be monophyletic on both karyological and morphological grounds, they do
not entirely correspond to morphological clumps and gaps.

I feel that a more reasonable approach to the taxonomic interpretation of
the chromosomal data is to consider the genera that have been proposed on
morphological grounds as hypotheses which are, to varying degrees, testable
by the chromosomal data.

During the last 50 years, a maximum of 5 Recent genera of xantusiids
have been recognized (in parentheses are listed the Recent species that I
consider valid) : Lepidophyma A. Dumeril, 1851 (flavimaculatum, micro pho-
lis, occulor, pajapanensis, smithi, tuxtlae, species novum); Xantusia Baird,

16

Contributions in Science

No. 227

1859 ( henshawi , vigilis)-, Cricosaura Gundlach and Peters, 1863 ( typica );
Gaigeia Smith, 1939 ( dontomasi , gaigeae, radula ); and Klauberina Savage,
1957 {riversiana) . In the most recent review of the genera of the family,
Savage (1963) recognized 4 of these 5, placing the species formerly included
in Gaigeia into the genus Lepidophyma.

No chromosomal data are yet available for Cricosaura typica. This is
especially unfortunate because Savage (1963) considered this species to be
morphologically the most distinctive in the family and placed it in a mono-
typic subfamily, Cricosaurinae, leaving all other species of the xantusiids in
the Xantusiinae. The obtaining of chromosomal data for this species will
allow further testing and comparisons of both the chromosomal and mor-
phological phylogenetic hypotheses.

Among xantusiids the most primitive number of microchromosomes
(22) is found in three of the ten species studied to date: Xantusia henshawi,
X. vigilis, and X. riversiana. The similarity of the karyotypes of the three
species of Xantusia and the consistently lower number of microchromosomes
of the other 7 species xantusiids studied does not support Savage’s (1957)
partioning of X. riversiana into the monotypic genus Klauberina. The chro-
mosomal evidence does not, however, unequivocably support the inclusion of
riversiana in the genus Xantusia for two reasons: (1) the microchromosome
number present in X. henshawi , vigilis, and riversiana is a shared primitive
character state and this increases their phenetic similarity but does not neces-
sarily indicate a close phylogenetic relationship; (2) as was discussed above,
homogeneity and gaps in karyotypic variation do not always correspond with
those of other data (morphological, ecological, behavioral, etc.). What can be
said is simply that the chromosomal data lacks the pattern that Savage (1957)
has reported for the morphological data, in that X. henshawi and X. vigilis do
not share any chromosomal state that could be considered derived from a
primitive state occurring in X. riversiana.

In addition to the pattern present in the chromosomal data, there are
several other reasons why I prefer not to recognize the genus Klauberina.
Genera are predictive hypotheses based on monophyly, similarities, and
gaps. Monotypic genera are often the result of classifications in which there
has been an overemphasis of differences. One increasingly popular solution
to this problem is to use numerical techniques for quantifying species differ-
ences and then to compare these differences with standards for the minimum
acceptable size of generic gaps. Short of such an analysis, I can argue against
the partioning of the genus Xantusia only by pointing out the many similar-
ities of the three species (X. henshawi, riversiana, and vigilis ) and their differ-
ences from other xantusiids. This has already been done for the chromosomal
data. The morphological evidence indicated that Xantusia riversiana ( =
Klauberina ) is more closely related to X. vigilis and X. henshawi than any
of these three species are to any of the other xantusiid (Savage, 1963). The
Eocene Wyoming fossil Paleoxantusia ferra has been considered intermediate

1972

Karyotypic Evolution of the Xantusiidae

17

between X. riversiana ( Klauberina ) on the one hand and X. vigilis and hen-
shawi on the other (Savage, 1963:34), suggesting that these lines diverged
later than did Lepidophyma, Cricosaura, and Xantusia. The distributions of
the species of the family suggest that each of the above three genera also
occupies a somewhat consistent and distinctive adaptive zone. Species of the
genus Lepidophyma occur primarily in wet tropical forests; Cricosaura typica
is isolated in the Cabo Cruz area of Cuba apparently occurring under rocks
and decaying leaves in forest (Barbour and Ramsden, 1919:178); while the
three species of Xantusia have largely allopatric ranges in the arid and semi-
arid southwestern U.S. and northwestern Mexico. I am not trying to ignore
such distinctive species ecologies as the montane limestone cap-rock habitat
of L. gaigeae or the less restricted microhabitat enjoyed by Xantusia riversiana
in its insular isolation, but wish simply to point out the biogeographical con-
sistency of the three Recent genera that I feel should be recognized. Regal
(1968) has recently pointed out that the pupils of some members of the
genus Lepidophyma (perhaps exclusive of L. gaigeae ) are round while those
of other xantusiids are elliptical, an observation originally made by Cope
(1900) but apparently overlooked by Savage ( 1963). This is a morphological
observation that has broad ecological and evolutionary implications in that
Regal (1968:85-86) presents the viewpoint that in xantusiids the elliptical
pupil is a derived condition associated with the evolution of basking behavior.
It may, then, be a derived character state shared by Cricosaura typica, Xan-
tusia henshawi, X. vigilis, X. riversiana , and perhaps L. gaigeae. Further
studies of pupil shape and retina structure in xantusiids are needed to deter-
mine the direction and degree of convergence in the evolution of eyes in this
family.

I feel that the chromosomal, morphological, and biogeographical infor-
mation summarized above indicates that the evolutionary relationships of the
three species of Xantusia ( henshawi , riversiana , and vigilis ) are best reflected
by their inclusion in one genus Xantusia, with two subgenera, Xantusia ( X .
henshawi and X. vigilis ) and Klauberina (X. riversiana) .

Smith (1939) proposed the monotypic genus Gaigeia in which he
placed Lepidophyma gaigeae. He considered the genus to be intermediate
between Lepidophyma and Xantusia in scale characters, having three of the
distinctive character states of each of these genera, plus one unique scale
character and a unique habitat. Because he felt that ( 1 ) three subsequently
described species (L. dontomasi, L. radula, and L. sylvaticum, considered by
Smith, 1942, as species of Gaigeia ) bridged the gap in scalation between the
two genera ( Lepidophyma and Gaigeia) and (2) “the two supposed genera
are practically identical in their skeletons,” Savage (1963:33) placed all
these species in Lepidophyma, a conclusion that was anticipated by Hecht
(1956:2). Although I have karyotypic data for only one (L. gaigeae) of the
four species that Smith ( 1942) considered to be in the genus Gaigeia, it is per-
haps the most distinctive one of this group. The chromosomal information is

18

Contributions in Science

No. 227

more conclusive in this instance than it is in the case of Xantusia riversiana, in
that L. gaigeae shares one definitely derived chromosomal state (loss of
one pair of microchromosomes) with all other species of Lepidophyma stud-
ied. It also shares one character state that is probably derived (the presence
of secondary constrictions on chromosome pair 3) with four other species
of Lepidophyma ( flavimaculatum , pajapanensis, smithi, and tuxtlae). The
karyotype of L. gaigeae is one of the most highly derived in the genus Lepi-
dophyma (Tables 1 and 2, Fig. 6). Interestingly enough, the karyotype of
L. gaigeae shares two derived chromosomal states with the /3 karyotype of
Xantusia henshawi in that chromosome pair 7 is submetacentric and chromo-
some pair 9 is subtelocentric. However, the pattern existing in the number of
microchromosomes and the occurrence of secondary constrictions on the
third pair of chromosomes make the conclusion inescapable that these two
derived karyotypic states shared by Xantusia h. henshawi and Lepidophyma
gaigeae must be the result of a certain amount of chromosomal convergence
that has accompanied their morphological convergence. In this case I feel that
the chromosomal data largely agree with the osteological information (Savage,
1963:33), and that L. gaigeae (and thus perhaps the other three species of
Gaigeia recognized by Smith, 1942) should be included in the genus
Lepidophyma.

Two species, L. micropholis and L. occulor, share ( 1 ) the loss of at
least one pair of micros, a derived state characteristic of other species of
Lepidophyma-, (2) the absence of satellites on pair 3, a primitive state char-
acteristic of the species of the genus Xantusia ; and (3) submetacentric pair
7, a derived state also present in X. henshawi and L. gaigeae. Chromosomally
L. occulor and L. micropholis thus appear to form a distinct species group
in the genus Lepidophyma, a hypothesis which is to be tested by morpho-
logical data.

Inter-familial Relationships:

The evolutionary relationships of the Xantusiidae remain obscure. Cope
(1900) placed the xantusiids in the suborder Leptoglossa within which he
considered them to be most closely allied to the lacertids. Camp (1923)
pointed out the similarities of xantusiids to both (1) the gekkonids (of the
division Ascalabota) and (2) the scincids, teiids, and especially the lacertids
(all of the section Scincomorpha of the division Autarchoglossa). Although
the family Xantusiidae bridged the morphological gap between his two major
divisions of the Sauria, Camp (1923) placed it in the Autarchoglossa, of
which he considered it to be the most primitive family. McDowell and Bogert
(1954) anticipated that future workers would refer the Xantusiidae to the
Gekkota. Underwood (1957) placed the xantusiids in the Ascalabota; Savage
(1963) referred them to the Gekkota. More recent morphological evidence
has been presented which ally the family with both Gekkota (St. Girons,
1967) and Scincomorpha (Miller, 1966; Etheridge, 1967).

1972

Karyotypic Evolution of the Xantusiidae

19

Available karyotype data for xantusiids, scincids, lacertids, teiids, and
gekkonids are summarized in Table 3. Although there is overlap in both
chromosome number and number of chromosome arms, gekkonid karyotypes
differ from those of xantusiids in (1) usually being composed entirely of
telocentric chromosomes; and (2) having a smooth gradation in chromosome
size, thus precluding a distinction between marcros and micros. Scincid
karyotypes differ in having (1) usually fewer micros, and (2) fewer chro-
mosome arms. Those of lacertids differ in having (1) fewer micros, (2)
more macros, and (3) fewer chromosome arms. Teiid karyotypes overlap
those of xantusiids in all regards (numbers of chromosomes, macros, micros,
and chromosome arms).

Derivation of the primitive xantusiid karyotype from known gekkonid
karyotypes would require the fusion of telocentric chromosomes to form
longer bi-armed macrochromosomes and the retention of the centromeres
(devested of most of their euchromatin) as microchromosomes, thus increas-
ing the number of chromosome arms while chromosome number remains
approximately constant. However, because they have many primitive states,
the karyotypes of gekkonids could be considered ancestral to those of most
families of lizards.

Among the lizard families thought by various workers to be closely
related to xantusiids, teiids appear to be karyotypically the most similar. That
these two families may be closely related is suggested by: (1) the existence
of macroteiids having primitive (unfused) karyotypes with numbers of chro-
mosome arms approximating those of xantusiids; and (2) the complementary
geographical distribution and the similarities in macrochromosome configura-
tion, external morphology, and ecology of microteiids and xantusiids. I must
stress that I present this simply as a phylogenetic hypothesis that should be
tested by further comparisons (anatomical, karyotypic, serological, etc.)
between xantusiids and other lizards, especially microteiids.

Table 3. Diploid chromosome number (2 n), numbers chromosome arms (CA),
macrochromosomes (Macros), and microchromosomes (Micros), and
literature source (Reference) for five families of lizards.

Family

In

CA

Macros

Micros

Reference

Xantusiidae

36-40

50-58

16-18

18-22

This paper

Gekkonidae

32-63

32-63

32-63

Kluge and Eckardt, 1969

Scincidae

24-32

36-46

10-32

0-18

Dutt, 1969

Lacertidae

24-38

38

24-36

0-3

Gorman, 1969

Teiidae

34-56

46-66

12-32

22-26

Gorman, 1970

20

Contributions in Science

No. 227

Origin of Unisexuality in the Genus Lepidophyma:

Telford and Campbell (1970) reported an all-female population of
Lepidophyma flavimaculatum in the Canal Zone (3 miles SE Achiote, Colon
Province) of Panama. To help elucidate the evolutionary origin of unisexual
xantusiids, I have studied karyotypes of specimens from this population and
have analysed variation in sex ratio in the genus Lepidophyma.

As was pointed out above (see Karyotype Descriptions) the karyotypes
of specimens from this all-female population of L. flavimaculatum are, with
one exception, diploid and appear identical to those of individuals of this
species from a bisexual population in Chiapas. This same karyotype was also
found in recently obtained material from a unisexual population of L. fla-
vimaculatum in southeastern Costa Rica. Thus, this case of presumed
parthenogensis appears generally not to involve polyploidy. The pos-
sibility that this population is allodiploid, however, cannot be ruled out by

Table 4. Sample size (N), number of males ( $ ), number of females ( $ ), and
percent females ( % 9 ) for ten species samples of Lepidophyma and 13
populations of L. flavimaculatum. Asterisk (*) indicates a sex distribu-
tion that is significantly different (.05 level) from that of L. gaigeae (see
text).

N

$

$

% $

dontomasi

1

0

1

100

gaigeae

260

110

150

58

micropholis

10

6

4

40

occulor

6

3

3

50

pajapanensis

13

4

9

69

radula

1

0

1

100

smithi

144

63

81

56

tuxtlae

53

24

29

55

species novum

5

1

4

80

flavimaculatum

174

29

145

83*

Tamaulipas

15

2

13

87

Queretaro

9

0

9

100*

Nuevo Leon

2

1

1

50

San Luis Potosi

1

1

0

0

Veracruz

3

0

3

100

Oaxaca

3

1

2

67

Tobasco

3

1

2

67

Chiapas

12

5

7

58

Guatemala

18

5

13

72

Honduras

17

10

7

41

Nicaragua

5

1

4

80

Costa Rica

49

2

47

96*

Panama

37

0

37

100*

1972

Karyotypic Evolution of the Xantusiidae

21

the evidence at hand, since at least two other species, L. tuxtlae and L. pajap-
anensis, have karyotypes identical to the one under consideration. Hybridiza-
tion between any of these species could result in an allodiploid in which the
two separate chromosomal complements, although not distinguishable mor-
phologically, are sufficiently different genetically to reduce the efficiency
of meiosis and thereby increase the selective advantage of parthenogenetic
reproduction.

Both triploid (3 n = 57) and diploid (2 n = 38) cells were observed in
the karyotype slides from one of the eight individuals that was analysed from
the Panama population (see Karyotype Descriptions above). It is difficult
to hypothesize a reasonable mechanism for the origin of these two levels
of ploidy that were observed in this one bone marrow preparation. Although
the triploid and diploid cells were found in a bone marrow preparation, some
type of mosaic may be involved and the two levels of ploidy may represent
different types of leukocytes derived from different embryonic tissue lines. I
am not aware of any really comparable phenomena among vertebrates,
except perhaps the tissue mosaics involving centric fusions in Salmo irideus,
reported by Ohno, Stenius, Fiast, and Zenges (1965) and the exparabiotic
diploid-triploid leukocyte chimeraras of Rana pipiens reported by Volpe and
Gebhardt (1966).

To survey the genus Lepidophyma for the occurrence of unisexuality, the
sex of 666 adult specimens of the 10 recognized species was determined by
examination of gonads (Table 4). Because many of the samples are small
and most have greater than 50 per cent female, statistical tests were used
to determine which samples have significantly different sex ratios. Choice of
the appropriate test was somewhat difficult because the per cent female is
greater than 50 in 9 of the 10 species. These observed deviations from the
50 per cent female (that would be theoretically expected to occur at birth
in a bisexual species) may be due to: (1) chance; (2) alteration of sex ratio
by a basic genetic mechanism ( e.g . meiotic deive); (3) differences in sur-
vivorship of the sexes; or (4) differences in the “collectability” of the sexes.
Since chi-square analysis ordinarily requires the use of a theoretical value, it
does not aid in the task of distinguishing between ( 1 ) sex ratio deviations
resulting from a basic genetic mechanism and (2) those of non-genetic
origin (differential sampling and survivorship). The other available statistical
test, the contingency test (Simpson, Roe, and Lewontin, 1960:186-191),
requires the selection of one of the samples as a standard with which the
other samples are to be compared. Although this procedure has several pitfalls
of its own, it does maximize the probability of making correct distinctions
between genetic and non-genetic deviations in sex ratio, if it is accepted
that the samples and the standard have a similar collecting bias.

The sample of Lepidophyma gaigeae was chosen as the standard because
it (1) is the largest available species sample; (2) was drawn from a relatively
small geographic area (mountains of Queretaro and Hidalgo, Mexico); and

22

Contributions in Science

No. 227

(3) was collected throughout the year. Using a 2x2 contingency test, with
Yates’ correction where applicable (see Simpson, Roe, Lewontin, 1960:186-
191), the number of males and females in each species sample was tested
against that of L. gaigeae. For only L. flavimaculatum was the per cent female
found to be statistically different (.05 level) from that of L. gaigeae. As this
polytypic species ranges from Tamaulipas, Mexico, to Panama, the species
sample was divided into 13 geographical samples (based on the states of
Mexico and the countries of Central America). When the number of males
and females in each of these geographical samples was compared with that
in L. gaigeae , only Panama (100% female), Costa Rica (96% female), and
Queretaro (100% female) were found to be significantly different; Tamaulipas
(87% female) almost reached the accepted level of significance (.05). The
only other geographical samples large enough to allow reasonable estimates
of sex ratio (Chiapas, Guatemala, and Honduras) do not differ significantly
from L. gaigeae. Twenty of the 29 known males of L. flavimaculatum occur
among the samples of these apparently bisexual populations. Thus L. fla-
vimaculatum appears to be a polytypic species composed of (1) a central
diploid bisexual population, L. f. flavimaculatum, in Chiapas (58% female),
Guatemala (72% female), and Honduras (41% female); (2) a northern
all-female or nearly all-female population (of unknown level of ploidy), L. f.
tenebrarum, in Tamaulipas (87% female) and Queretaro (100% female);
and (3) a southern all-female or nearly all-female diploid population, L. /.
obscurum, in Costa Rica (96% female) and Panama (100% female). Sam-
ples are inadequate to determine the sex ratios of the intervening populations
with any degree of accuracy.

Analysis of large samples from local populations throughout the exten-
sive range of the polytypic L. flavimaculatum is required to determine whether
changes in sex ratio and morphology are gradual or abrupt, and to allow an
appraisal of the taxonomic status of the included forms. The two known male
specimens from Costa Rica are among the northernmost available from that
country, suggesting that the occurrence of males in “highly female” popula-
tions in Costa Rica might be nothing more than an artifact resulting from
the accidental grouping of samples from bisexual and unisexual populations.
In Tamaulipas, on the other hand, there is better evidence that males may
actually occur in quite low frequency in local populations, since among the
10 adult specimens available from the Gomez Farias region, only one male
was found. Comparison of sex ratios in several age classes could help to
determine the relative importance of pre- and post-natal mechanisms in
altering the sexual composition of the population. Before any of these ques-
tions can be addressed, adequate samples must be collected. This task is made
both difficult and urgent as the devastation of the lowland tropical forests of
Middle America approaches completion.

Unisexuality in the genus Lepidophyma appears to be similar to that of
the lizards of the saxicola group of Lacerta in that (a) all forms are diploids

1972

Karyotypic Evolution of the Xantusiidae

23

with two identical sets of chromosomes, (b) there are forms intermediate
between bisexual and unisexual; (c) the formation of small isolated popula-
tions appears to have been an important factor in the evolution of partheno-
genesis (Darevsky, 1966). Known unisexual gekkos (Kluge and Eckardt,
1969) and agamids (Hall, 1970) are triploid rather than diploid. In the
genus Cnemidophorus diploid unisexuality has been reported for C. neomex-
icanus and some C. tesselatus, but these, however, have been convincingly
demonstrated to be allodiploids resulting from inter-specific hybridization
(Lowe and Wright, 1966; Wright and Lowe, 1967), while karyotypic hetero-
morphism is not apparent in the unisexual L. flavimaculatum (Fig. 3).
Vanzolini (1970) recently reported an apparently rapid shift from bisexuality
to unisexuality in some Amazonian populations of Cnemidophorus lemniscatus
and suggests that such a shift is probably not the result of inter-specific
hybridization. However, Denise Peccinini (1971) reported that although
these unisexual populations are diploid, they have one to three pairs of heter-
omorphic chromosomes and “it is possible, therefore, that the hybridization
has been between subspecies of C. lemniscatus or even intraspecific poly-
morphic variants.” For Lepidophyma flavimaculatum there is, at present, no
morphological, cytogenetic, or biogeographical evidence that hybridization
preceded the evolution of unisexuality. However, the paucity of the data
leaves the question still open and it is certainly not unfeasible that the diploid
unisexual population in Panama arose by hybridization between forms that
are karyotypically similar but sufficiently different genetically to impair synap-
sis and thus add selective pressures for the evolution of unisexual reproduction.

During my approximately 10 years of experience with xantusiids, a
number of field impressions have been formed about their ecology and prob-
able evolutionary history. Although it is perhaps somewhat premature, I
wish to here present those impressions that may help to explain the evolution
of unisexuality in the family.

Xantusiids characteristically occur in localized but frequently dense
populations. This distributional pattern is dictated by their narrow micro-
environmental requirements. The ecological conditions to which the family
is adapted were probably more widespread in the early Tertiary. This group
of lizards appears to have responded to the increasingly arid continental
climates of the middle and late Tertiary by becoming increasingly specialized
for, and restricted to, specific limited ecological situations (e.g., under cap
rocks of boulders, under bark, beneath yucca-like plants, in caves) in which
their unaltered microenvironmental requirements could be met. These stresses
have produced a disjunct relictual pattern of distribution. Moreover, the
resulting isolated populations are frequently under tremendous pressure for
colonization of new areas because of fluctuations in climate, vegetation, and
habitat availability.

For example, the narrow ecological requirements of Xantusia vigilis
result in a disjunct geographical range and in “clumped” distributions within

24

Contributions in Science

No. 227

any given area. These local “clumps” appear to occur in areas having optimal
edaphic and microclimatic conditions and relatively large numbers of yuccas
or other suitable plants. Because of climatic and vegetational changes, the
concurrence of all these conditions is not only a rare condition, but probably
also an extremely transitory one.

Field experience with Lepidophyma flavimaculatum leads me to believe
that these generalizations are particularly valid for this species. The popula-
tion located by Telford and Campbell near Achiote appears highly localized
and rather dense. To date approximately 50 individuals have been collected
from this population while only ca. 20 are known from the rest of Panama.
My efforts to locate other individuals of this species even short distances from
this population were unsuccessful (see also Telford and Campbell, 1970).
Optimal conditions of forest canopy, humidity, and soil, as well as the presence
of a number of extremely large logs in the proper state of decay appear to be
involved; all of these factors may be related to a particular stage in the suc-
cession of this nearly mature secondary forest. Judging from the large number
of Lepidophyma found around them, each of these logs would appear to form
a “colony.” As forest maturation and log decay continue, the individuals of
this population are under considerable selective pressure to establish new
colonies, perhaps at great distances, where the soil, humidity, forest canopy,
and logs are livable.

These selective pressures would favor the evolution of unisexuality,
thereby facilitating colonization by allowing each individual to reproduce in
isolation and by doubling the reproductive potential. The occurrence of
unisexual populations at the northern and the southern periphery of the range
of L. flavimaculatum is thus probably indicative of a continuing contraction
rather than expansion of its range. This is in marked contrast to the situation
in the genus Cnemidophorus in which the evolution of unisexuality appears
to have resulted from interspecific hybridization and expansion into new
habitats (Wright and Lowe, 1968).

SPECIMENS EXAMINED

The following specimens were used in the karyotypic analysis and are
deposited in the Herpetological Collection, Department of Biological Sciences,
the University of Arizona (UAZ); the Natural History Museum of Los
Angeles County (LACM); and the Texas Cooperative Wildlife Collection
(TCWC), Texas A & M University.

Lepidophyma flavimaculatum : MEXICO: Chiapas : 25 mi (by rd to
Malpaso) NW Ocozocoautla (UAZ 28805-06). PANAMA: Canal Zone : 3
mi (air line) SE Achiote (8 mi NNW Escobal) (UAZ 27637-42, 27644,
28826). COSTA RICA: Puntarenas Prov .: 6 km S San Vito de Java (LACM
72323).

Lepidophyma gaigeae : MEXICO: Hidalgo : 2 mi N Durango, 13 mi

1972

Karyotypic Evolution of the Xantusiidae

25

(by Hwy 85) S Jacala (UAZ 28868-72); Durango, 15 mi (by Hwy 85) S
Jacala (UAZ 28880-84, 28895-905).

Lepidophyma micropholis : MEXICO: Tamaulipas : Cave at El Pachon,
8 km (by rd) NNE Antigua Morelos (UAZ 28762, 28767, 28769).

Lepidophyma occulor : MEXICO: Queretaro : 2.5 mi S Conca, 2000 ft
(TCWC 35605).

Lepidophyma pajapanensis : MEXICO: Veracruz : Coyame, 9 mi SE
Catemaco (UAZ 28804); 2 mi (by rd) SE Sontecomapan, 14 mi (by rd.)
NE Catemaco (UAZ 28808-10).

Lepidophyma smithi : MEXICO: Chiapas : ca. V2 mi (by Hwy 200)
NW Escuintla (UAZ 28788); 9 mi (by Hwy 200) NW Escuintla (UAZ
28797); 4 mi NW Mapastepec, 24 mi (by Hwy 200) NW Escuintla (UAZ
28812-15); Oaxaca : IV2 mi (by Hwy 190) E Tapanatepec (UAZ 28794).

Lepidophyma tuxtlae : MEXICO: Chiapas : 25 mi (by rd to Malpaso)
NW Ocozocoautla (UAZ 28780, 28782); Veracruz: 2 mi (by rd) SE Sonte-
comapan, 14 mi (by rd) NE Catemaco (UAZ 28770-76).

Xantusia henshawi : MEXICO: Durango : 6.5 mi NE Pedricena (13.7
mi by rd SE Chocolate) (LACM 72324-25). UNITED STATES: California :
Riverside Co.: 2 mi (by rd to Idyllwild) S Banning, San Jacinto Mts. (UAZ
21653, 21694, 21700); 3 mi (by rd to Idyllwild) S Banning, San Jacinto Mts.
(UAZ 21690, 21692).

Xantusia riversiana : UNITED STATES: California : Los Angeles Co.:
N end of San Clemente Island (UAZ 21679-81, 21683-84, 21686-89).

Xantusia vigilis: MEXICO: Baja California del Norte : ca. 14 mi (by rd)
E La Trinidad, Valle de La Trinidad (UAZ 28961-62); Durango : 6.5 mi NE
Pedricena (13.7 mi SW Chocolate) (LACM 72326-331); Sonora : 1-2 mi (by
rd) S Desemboque del Rio San Ignacio (UAZ 24858, 24860, 24868, 24894).
UNITED STATES: Arizona : Yavapai Co.: 11.3 mi (by Hwy 93) SE Burro
Creek, ca. 3200 ft (UAZ 24210, 24216, 24231); vie. Yarnell, 4750 ft (UAZ
24184, 24196, 24227, 24854, 24861); Yuma Co.: E end of Palm Canyon,
Kofa Mts. (UAZ 24215, 24240); California : Kern Co.: 0.5 mi (by rd) E
Granite Station (LACM 72332-33); 0.9 mi (by Hwy 178) SE of the summit
of Walker Pass (LACM 72334); 6 mi W Mojave (LACM 72335); Los
Angeles Co.: 1.8 mi (by Hwy 14) N Palmdale (LACM 72336); Riverside
Co.: 1 mi S, % mi W Whitewater (LACM 72337-338).

ACKNOWLEDGMENTS

This paper is an expansion and revision of a part of a dissertation
submitted to the University of Arizona. During my graduate work, several
individuals have been exceptionally generous with their time, ideas, and con-
structive criticisms. I am particularly grateful to my dissertation advisor, Dr.
Charles H. Lowe, for imparting an ecophysioevolutionary perspective to my
interest in reptiles and amphibians; to Dr. C. Jay Cole, for his patience in

26

Contributions in Science

No. 227

helping me to find my way out of corn fields and in tutoring me in all aspects
of cytotaxonomy including the art of making a pastie; to Dr. James L. Patton
for discussing ideas and principles of cytogenetics; to Dr. Philip J. Regal for
discussing with me his thoughts on the ecology and evolution of spectacled
lizards; and to Dr. David S. Hinds for sparing me from the tragedy of the
“double-nested do-loop.” Dr. John Wright has helped greatly with the diffi-
cult task of revising and preparing the paper for publication, by offering
encouragement, advice, and constructive criticism.

Xantusiid lizards are collected with crowbars and hard work; I am
indebted to many blister-handed field friends: Kathryn Bolles, Eldon Braun,
Duke Campbell, Jay Cole, Steven Goldberg, Charles Lowe, Roy McDiarmid,
Philip Regal, Michael Robinson, Wade Sherbrooke, Sam Telford, David
Whistler, and John, Brian, and Keith Wright. I am especially grateful to James
Dixon for making available a live specimen of the rare Lepidophyma occulor
for chromosomal analysis.

I express my thanks also to those who have read and criticized this manu-
script: Drs. L. A. Carruth, L. A. Crowder, H. K. Gloyd, W. B. Heed, C. H.
Lowe, J. L. Patton, F. G. Werner, and J. W. Wright. The skillful editorial
efforts of Dr. Gloyd are especially appreciated.

The following persons permitted me to examine specimens under their
care: Dr. James R. Dixon, Texas Cooperative Wildlife Collection, Texas
A & M University; Dr. Charles L. Douglas, Texas Natural History Collection,
University of Texas; Dr. William E. Duellman, University of Kansas Museum
of Natural History; Dr. Donald F. Hoffmeister, University of Illinois Museum
of Natural History; Dr. Charles H. Lowe, University of Arizona; Dr. James
A. Peters, United States National Museum; Dr. William F. Pyburn, University
of Texas, Arlington; Dr. Douglas H. Rossman, Louisana State University
Museum of Zoology; Dr. Jay M. Savage, University of Southern California;
Dr. Sam R. Telford, Florida State Museum; Dr. Charles F. Walker, University
of Michigan Museum of Zoology; Dr. Richard G. Zweifel, American Museum
of Natural History.

Dr. T. C. Hsu of the M. D. Anderson Hospital and Tumor Institute of
Houston generously determined the karyotype of an individual of Lepido-
phyma flavimaculatum by lung tissue culture.

The Computer Center of the University of Arizona facilitated the data
reduction.

This study was partially supported by a NASA traineeship at the Uni-
versity of Arizona.

1972

Karyotypic Evolution of the Xantusiidae

27

Literature Cited

Barbour, T., and C. T. Ramsden. 1919. The herpetology of Cuba. Mem. Mus.
Comp. Zool. 47(2): 77-166.

Bezy, R. L. 1967a. A new night lizard (Xantusia vigilis sierrae) from the southern
Sierra Nevada in California. J. Ariz. Acad. Sci. 4:163-167.

1967b. Variation, distribution, and taxonomic status of the Arizona night

lizard (Xantusia arizonae). Copeia 1967:653-661.

Camp, C. L. 1923. Classification of the lizards. Bull. Amer. Mus. Nat. Hist. 48:
289-481.

Cole, C. J. 1970. Karyotypes and evolution of the spinosus group of lizards in the
genus Sceloporus. Amer. Mus. Nov. 2431:1-47.

Cope, 1895. The genera of Xantusiidae. Amer. Nat. 29:757-758.

Cope, E. D. 1900. The crocodilians, lizards, and snakes of North America. Ann.
Rep. U. S. Nat. Mus. for 1898, part 11:151-1294.

Darevsky, I. S. 1966. Natural parthenogenesis in a polymorphic group of Cauca-
sian rock lizards related to Lacerta saxicola Eversmann. J. Ohio Herp. Soc.
5:115-152.

Darlington, P. S. 1970. A practical criticism of Hennig-Brundin “Phylogenetic
Systematics” and Antarctic biogeography. Syst. Zool. 19:1-18.

Dutt, K. 1969. Study of chromosomes in two species of Indian lizards. Microscope
17:213-218.

Etheridge, R. 1967. Lizard caudal vertebrae. Copeia 1967:699-721.

Ford, C. E., and J. L. Hamerton. 1956. A colchicine, hypotonic citrate, squash
sequence for mammalian chromosomes. Stain Tech. 31:247-254.

Gorman, G. C. 1969. New chromosome data for 12 species of lacertid lizards. J.
Herpetol. 3:49-54.

1970. Chromosomes and the systematics of the family Teiidae (Sauria,

Reptilia). Copeia 1970:230-245.

Gorman, G. C., L. Atkins, and T. Holzinger. 1967. New karyotypic data on 15
genera of lizards in the family Iguanidae, with a discussion of taxonomic and
cytological implications. Cytogenetics 6:286-299.

Gorman, G. C, R. B. Huey, and E. E. Williams. 1969. Cytotaxonomic studies on
some unusual iguanid lizards assigned to the gerera Chamaeleolis, Polychrus,
and Phenacosaurus, with behavioral notes. Breviora 316:1-17.

Hall, W. P. III. 1970. Three probable cases of parthenogenesis in lizards (Agami-
dae, Chamaeleontidae, Gekkonidae). Experientia 26:1271-73.

Hecht, M. K. 1956. A new xantusiid lizard from the Eocene of Wyoming. Amer.
Mus. Nov. 1774:1-8.

Hennig, W. 1966 Phylogenetic systematics. Univ. 111. Press, Urbana.

Hsu, T. C., and R. A. Mead. 1969. Mechanisms of chromosomal changes in mam-
malian speciation, p. 8-17. In Comparative Mammalian Cytogenetics (Kurt
Benirschke, ed.). Springer-Verlag, N. Y.

Klauber, L. M. 1931. A new species of Xantusia from Arizona, with a synopsis of
the genus. Trans. San Diego Soc. Nat. Hist. 7:1-16.

Kluge, A. G., and M. J. Eckardt. 1969. Hemidactylus garnoti Dumeril and Bibron,
a triploid all-female species of gekkonid lizard. Copeia 1969:651-664.

28

Contributions in Science

No. 227

Lowe, C. H., and J. W. Wright. 1966. Evolution of parthenogenetic species of
Cnemidophorus (whiptail lizards) in western North America. J. Ariz. Acad.
Sci. 4:81-87.

Lowe, C. H., C. J. Cole, and J. L. Patton. 1967. Karyotype evolution and specia-
tion in lizards (genus Sceloporus ) during evolution of the North American
Desert. Syst. Zool. 16:296-300.

Lowe, C. H., J. W. Wright, and C. J. Cole. 1966. Chromosomes and karyotypes
of sceloporine iguanid lizards in the North American Southwest. Mamm.
Chrom. Newsletter 22:201-203.

Lowe, C. H., J. W. Wright, C. J. Cole, and R. L. Bezy. 1970a. Chromosomes and
evolution of the species groups of Cnemidophorus (Reptilia: Teiidae). Syst.
Zool. 19:128-141.

1970b. Natural hybridization between the teiid lizards Cnemidophorus

sonorae (parthenogenetic) and Cnemidophorus tigris (bisexual). Syst. Zool.
19:114-127.

Matthey, R. 1931. Chromosomes de Reptiles Sauriens, Ophidiens et Cheloniens.
L’evolution de la formule chromosomiale chez les Sauriens. Revue Suisse de
Zoologie 38:117-184.

Matthey, R. 1951. The chromosomes of the vertebrates. Adv. Genet. 4:159-180.

Mayr, E. 1969. Principles of systematic zoology. McGraw-Hill, New York.

McDowell, S. M., and C. M. Bogert. 1954. The systematic position of Lantha-
notus and the affinities of the anguinomorphan lizards. Bull. Amer. Mus. Nat.
Hist. 105:1-142.

Miller, M. R. 1966. The cochlear duct of lizards. Proc. Calif. Acad. Sci. 23:
255-359.

Ohno, S., C. Stenius, E. Faisst, and M. T. Zenzes. 1965. Post-zygotic chromo-
somal rearrangements in rainbow trout ( Salmo irideus Gibbons). Cytogenetics
4:117-129.

Patton, J. L. 1967. Chromosome studies of certain pocket mice, genus Perognathus
(Rodentia: Heteromyidae). J. Mamm. 48:27-37.

Peccinini, D. 1971. Chromosome variation in populations of Cnemidophorus lem-
niscatus in the Amazon valley. Ciencia e Cultura 23(2) : 133-136.

Regal, P. J. 1968. An analysis of heat-seeking in a lizard. Ph.D. dissertation, UCLA.

St. Girons, H. 1967. Morphologie comparee de Fhypophyse chez les Squamata:
Donnees complementaires et apport a la phylogenie des reptiles. Ann. des Sci.
Nat. Zool. (Paris) 12:229-308.

Savage, J. M. 1957. Studies on the lizard family Xantusiidae. III. A new genus for
Xantusia riversiana Cope, 1883. Zoologica 42:83-86.

1963. Studies of the lizard family Xantusiidae IV. The genera. Los

Angeles Co. Mus., Contrib. Sci. 71:1-38.

Simpson, G. G., A. Roe, and R. C. Lewontin. 1960. Quantitative Zoology. Har-
court, Brace, New York.

Smith, H. M. 1939. Notes on Mexican reptiles and amphibians. Zool. Ser., Field
Mus. Nat. Hist. 24:15-35.

Smith, H. M. 1942. Mexican herpetological misellany. Proc. U. S. Nat. Mus. 92:
349-395.

1972

Karyotypic Evolution of the Xantusiidae

29

Sokal, R. R., and P. H. A. Sneath. 1963. Principles of numerical taxonomy. W. H.
Freeman, San Francisco.

Telford, S. R., and H. W. Campbell. 1970. Ecological observations on an all-
female population of the lizard Lepidophyma flavimaculatum (Xantusiidae)
in Panama. Copeia 1970:379-381.

Underwood, G. 1957. On the lizards of the family Pygopodidae. A contribution
to the morphology and phylogeny of the Squamata. J. Morphol. 100(2):
207-268.

Vanzolini, P. E. 1970. Unisexual Cnemidophorus lemniscatus in the Amazonas
Valley: a preliminary note (Sauria, Teiidae). Papeis avulsos de Zoologia, S.
Paulo, 23(7) :63-68.

Volpe, E. P., and B. M. Gebhardt. 1966. Evidence from cultured leucocytes of
blood cells chimerism in ex-parabiotic frogs. Science 154:1197-1199.

Walker, C. F. 1955. Two new lizards of the genus Lepidophyma from Tamaulipas.
Occ. Pap. Mus. Zool., Univ. Mich. 564:1-10.

Webb, R. G. 1965. A new night lizard (genus Xantusia ) from Durango, Mexico.
Amer. Mus. Nov. 2231:1-16.

1970. Another new night lizard ( Xantusia ) from Durango, Mexico. Los

Angeles Co. Mus., Contrib. Sci. 194:1-10.

Werler, J. E. 1957. A new lizard of the genus Lepidophyma from Volcan San
Martin Pajapan. Herpetologica 13:223-226.

White, M. J. D. 1954. Animal cytology and evolution. 2nd ed., Cambridge Univ.
Press, Cambridge.

Wright, J. W., and C. H. Lowe. 1967. Evolution of the allodiploid parthenospecies
Cnemidophorus tesselatus (Say). Mamm. Chrom. Newsletter 8:95-96.

1960. Weeds, polyploids, parthenogenesis, and the geographical and eco-
logical distribution of all-female species of Cnemidophorus. Copeia 1968(1):
128-138.

Accepted for publication July 12, 1971

-

' SpSj

NUMBER 228
JUNE 7, 1972

son 3

C 2 Lur

TYPE SPECIMENS OF AVIAN FOSSILS
IN THE COLLECTIONS OF THE
NATURAL HISTORY MUSEUM
OF LOS ANGELES COUNTY

By Hildegarde Howard

CONTRIBUTIONS IN SC16NCC

0

NATURAL HISTORY MUSEUM • LOS ANGELES COUNTY

CONTRIBUTIONS IN SCIENCE is a series of miscellaneous technical papers
in the fields of Biology, Geology and Anthropology, published at irregular intervals
by the Natural History Museum of Los Angeles County. Issues are numbered sep-
arately, and numbers run consecutively regardless of subject matter. Number 1 was
issued January 23, 1957. The series is available to scientific institutions and scien-
tists on an exchange basis. Copies may also be purchased at a nominal price. Inquiries
should be directed to Virginia D. Miller, Natural History Museum of Los Angeles
County, 900 Exposition Boulevard, Los Angeles, California 90007.

INSTRUCTIONS FOR AUTHORS

Manuscripts for CONTRIBUTIONS IN SCIENCE may be in any field of
Life or Earth Sciences. Acceptance of papers will be determined by the amount
and character of new information. Although priority will be given to manuscripts
by staff members, or to papers dealing largely with specimens in the collections of
the Museum, other technical papers will be considered. All manuscripts must be
recommended for consideration by the curator in charge of the proper section or
by the editorial board. Manuscripts must conform to those specifications listed
below and will be examined for suitability by an Editorial Committee including
review by competent specialists outside the Museum.

Authors proposing new taxa in a CONTRIBUTIONS IN SCIENCE must
indicate that the primary type has become the property of a scientific institution of
their choice and cited by name.

MANUSCRIPT FORM.— ( 1 ) The 1964 AIBS Style Manual for Biological
Journals is to be followed in preparation of copy. (2) Double space entire manu-
script. (3) Footnotes should be avoided if possible. Acknowledgments as footnotes
will not be accepted. (4) Place all tables on separate pages. (5) Figure legends and
unavoidable footnotes must be typed on separate sheets. Several of one kind may
be placed on a sheet. (6) An abstract must be included for all papers. This will be
published at the head of each paper. (7) A Spanish summary is required for all
manuscripts dealing with Latin American subjects. Summaries in other languages
are not required but are strongly recommended. Summaries will be published at
the end of the paper. (8) A diagnosis must accompany any newly proposed taxon.
(9) Submit two copies of manuscript.

ILLUSTRATIONS. — All illustrations, including maps and photographs, will
be referred to as figures. All illustrations should be of sufficient clarity and in the
proper proportions for reduction to CONTRIBUTIONS page size. Consult the 1964
AIBS Style Manual for Biological Journals in preparing illustration and legend copy
for style. Submit only illustrations made with permanent ink and glossy photo-
graphic prints of good contrast. Original illustrations and art work will be returned
after the manuscript has been published.

PROOF.— Authors will be sent galley proof which should be corrected and
returned promptly. Changes in the manuscript after galley proof will be billed to
the author. Unless otherwise requested, page proof also will be sent to the author.
One hundred copies of each paper will be given free to each author or divided
equally among multiple authors. Orders for additional copies must be sent to the
Editor at the time corrected galley proof is returned. Appropriate order forms will
be included with the galley proof.

Virginia D. Miller
Editor

TYPE SPECIMENS OF AVIAN FOSSILS IN THE COLLECTIONS OF
THE NATURAL HISTORY MUSEUM OF LOS ANGELES COUNTY

By Hildegarde Howard1

Abstract: Fossil bird types in the collections of the Natural
History Museum of Los Angeles County are listed with their
catalog numbers under the original published names. Included,
in addition to the type series, are subsequently described or
figured specimens that provide information concerning skeletal
elements not included in the original description. Bibliographic
references and locality data are provided throughout.

The International Code of Zoological Nomenclature (1964, Art. 72D)
recommends not only that each institution mark and carefully preserve all
type specimens deposited therein, but that it publish a list of all such material
in its possession. Accordingly, the following catalog of avian fossil types in the
collections of the Natural History Museum of Los Angeles County (LACM)
is presented. The help of Pierce Brodkorb in reviewing the completed manu-
script is gratefully acknowledged.

Included are holotypes, syntypes, paratypes, and lectotypes as defined by
the International Code (op. cit., Arts. 73 and 74) as well as casts of specimens
in these categories designated with the prefix plasto.

In Avian Paleontology, specimens remaining after designating the holo-
type are usually listed as “referred.” The term paratype (or in older publica-
tions, cotype) is reserved for outstanding specimens in the type series. How-
ever, in strict adherence to recommendation 73D of the International Code,
all specimens (other than the holotype) listed in the original description of a
species should be known as paratypes. This catalog follows the Code recom-
mendation, but the term will appear in quotes (“paratype”) unless it is also
used by the original describer.

Also included are described or figured specimens, recorded subsequent
to the original type description, that provide additional information regarding
the species. This material falls within the definition of the hypotype (Zullo
and Hertlein, 1970:3) and is listed under this term. As complete fossil skele-
tons are rarely found, paratypes and hypotypes, which often represent different
skeletal elements than the holotype, are of particular importance in Avian
Paleontology. Tentatively identified paratypes and hypotypes are included if
figured.

Species are grouped according to Order and Family and arranged alpha-
betically by genus as first described. The following information is included for
each entry: author, bibliographic reference, type category, skeletal element
(and portion thereof if incomplete), geologic age, Formation (if known) and

Research Associate in Palaeornithology, Natural History Museum of Los Angeles
County, Los Angeles, Calif. 90007.

1

2

Contributions in Science

No. 228

locality. Specimen numbers refer to the latest LACM Vertebrate Paleontology-
catalog. Specimens described from the former California Institute of Tech-
nology collection (now incorporated with the LACM collections) are prefixed
(CIT). Each holotype bears an individual catalog number; a few “paratypes”
or hypotypes have been grouped under one number. Wherever possible, ele-
ments described without designation of catalog numbers have been traced and
their numbers included here. However, hypotypes in this category are omitted
unless figured. For plastotypes, the catalog number of the original institution
is provided as well as the LACM cast number, and all data included in the
entry refer to the specimen from which the cast was made. In a few instances,
the specimens for which we have plastotypes were not illustrated in the orig-
inal description; reference to a review by a later author is, therefore, included.

The catalog includes 53 holotypes, 3 syntypes, 525 paratypes, 214 hypo-
types (168 figured) and 46 plastotypes of 112 species and two subspecies.
In a few instances a specimen is listed with more than one species, owing to
reidentification. Parenthetical reference to the most recent assignment is given
under the earliest listing. In the alphabetical species index at the end of the
catalog, the latest taxonomic designations are given in brackets.

Holotypes, syntypes and plastotypes are housed in a separate case in the
Department of Vertebrate Paleontology apart from the general collections.
Paratypes and hypotypes are filed by locality within the Vertebrate Paleon-
tology collections, except that some Rancho La Brea hypotypes have been
used in the composite mounts of the several species from that locality, and
“paratypes” and hypotypes of Mancalla from the San Diego Formation are
included in the composite mount of that flightless bird. See Howard (1962,
figs. 8, 10-21) for illustrations of the mounted specimens.

Avian fossils have been recorded from 50 LACM collecting areas, 33 of
which contain the material listed herein. Broken down into separate localities,
the number is considerably greater, as for example, the various pits at Rancho
La Brea and the separate street roadcuts in San Diego where the San Diego
Formation was accessible. Another 20 or more LACM collecting sites contain
unrecorded avian fossils.

List of Abbreviations

AMNH

ANSP

BM

CAS

CIT

CM

FGS

LACM

American Museum of Natural History
Academy of Natural Sciences, Philadelphia
British Museum

California Academy of Sciences
California Institute of Technology
Canterbury Museum, Christchurch, New Zealand
Florida Geological Survey

Natural History Museum of Los Angeles County
(formerly Los Angeles County Museum)

1972

Type Specimens of Avian Fossils

3

MCZ

SBMNH

SDSM

SU

UCLA/VP

UCMP

UF

USNM

YPM

(dist.)

(prox.)

(frag.)

(tent.)

Museum of Comparative Zoology, Harvard
Santa Barbara Museum of Natural History
South Dakota School of Mines
Stanford University

University of California, Los Angeles, Vertebrate Paleontology
Department

University of California Museum of Paleontology (Berkeley)

University of Florida

United States National Museum

Yale Peabody Museum

distal end preserved

proximal end preserved

fragmentary specimen

tentative identification

GAVIIFORMES : GAVIIDAE

Gavia concinna Wetmore

WETMORE, 1940: 25, figs. 1-4.

Plastoholotype ulna (prox.) USNM 16160; cast C681 Early Pliocene,
Etchegoin Formation, Sweetwater Canyon near King City, Monterey
County, California.

HOWARD, 1949b: 185-187, pi. 3, figs. 5, 6, 6a.

Hypotypes (tent.) upper mandible 2110 (figs. 6, 6a), and humerus 2133
(fig. 5) (see G. howardae Brodkorb); Pliocene, San Diego Formation,
San Diego, California.

BRODKORB, 1953: 211.

Hypotypes: cranium, rostrum and mandible 2109, rostrum 2110 (figured
tentatively, Howard 1949b, pi. 3, figs. 6, 6a), humerus (prox.) 2444;
Pliocene, San Diego Formation, San Diego, California.

Gavia howardae Brodkorb

BRODKORB, 1953: 212-213, fig. IB.

Holotype humerus (dist.) 2111 (fig. IB); “paratypes” humeri (dist.)
2133, 2175; Pliocene, San Diego Formation, San Diego, California.
MILLER and BOWMAN, 1958: 4, fig. 1 (p. 11).

Hypotype tibiotarsus (dist.) 2314; Pliocene, San Diego Formation, San
Diego, California.

PODICIPEDIFORMES: PODICIPEDIDAE

Colymbus sub parvus Miller and Bowman

MILLER and BOWMAN, 1958: 6, figs. 5a, 5b (p. 11).

Holotype femur (dist.) 2568 (figs. 5a, 5b); paratype femur (dist.) 2118;
“paratypes” tibiotarsus (prox.) 2129, coracoid 2354; Pliocene, San Diego
Formation, San Diego, California.

4

Contributions in Science

No. 228

PROCELLARIIFORMES : DIOMEDEIDAE
Diomedea calijornica Miller

HOWARD, 1966d: 2, fig. 1-1.

Hypotype humerus (dist.) 16468; Middle Miocene, Temblor Formation,
Sharktooth Hill, Kern County, California; Loc. 1625.

Diomedea milleri Howard

HOWARD, 1966d: 2-4, fig. 1C.

Holotype ulna (prox.) 7319 (fig. 1C); “paratype” tarsometatarsus
(prox.) 16474; Middle Miocene, Temblor Formation, Sharktooth Hill,
Kern County, California; Locs. 1655 and 1625.

PROCELLARIIFORMES : PROCELLARIIDAE
Fulmarus hammeri Howard

HOWARD, 1968b: 9, figs. 2F, 2K (p. 4).

Holotype carpometacarpus (prox.) 18262 (figs. 2F, 2K); “paratype”
humerus (dist.) 18263; Late Miocene, Leisure World, Laguna Hills,
Orange County, California; Loc. 1945.

Puffinus cal ho uni Howard

HOWARD, 1968b: 6, figs. 2A-2E (p. 4).

Holotype humerus (dist.) 17508 (figs. 2A, 2E); “paratypes” tarso-
metatarsus (prox.) 17582 (figs. 2B, 2C, 2D), humerus (dist.) 17539,
ulna (prox.) 17530; Late Miocene, Leisure World, Laguna Hills, Orange
County, California; Loc. 1945.

Puffinus conradi Marsh

MARSH, 1870: 212 (figured SHUFELDT, 1915, pi. 8, figs. 63-64).
Plastoholotype humerus (dist.); plasto “paratype” ulna (dist.), both
ANSP 13360; cast C688; Middle Miocene, Calvert Formation, Calvert
County, Maryland.

Puffinus diatomicus Miller

MILLER, 1925b: 111, pis. 1-2.

Plastoholotype complete skeletal impression UCMP 26541 (pi. 1), 2
casts (one in relief, one impressed) C692; plastoparatype impression left
wing bones SU 1 (pi. 2), cast (in relief) C693; Miocene, diatomaceous
shales, Lompoc, Santa Barbara County, California.

Puffinus felthami Howard

HOWARD, 1949b: 194, pi. 2, figs. 4, 6.

Holotype humerus (dist.) 2037 (fig. 6); paratype tarsometatarsus (prox.)
2038 (fig. 4); Early Pliocene, Repetto Formation, 3 miles north of
Corona del Mar, Orange County, California; Loc. 1067.

Puffinus inceptor Wetmore

WETMORE, 1930: 86, figs. 1-3.

Plastoholotype humerus (dist.) CAS 5223; cast C678; Middle Miocene,
Temblor Formation, Sharktooth Hill, Kern County, California.

1972

Type Specimens of Avian Fossils

5

Puffinus kanakoffi Howard

HOWARD, 1949b: 187, pi. 2, figs. 3, 5.

Holotype tarsometatarsus 2122 (fig. 3); paratypes, humerus (dist.) 2120
(fig. 5), femur 2124; “paratypes” tarsometatarsus 2126, tibiotarsus
(prox.) 2123, 4 humeri 2114, 2116, 2146, 2160; Pliocene, San Diego
Formation, San Diego, California.

Puffinus mitchelli Miller

MILLER, 1961: 400, fig. 1.

Plastoholotype humerus (dist.) UCMP 58184; cast C684; Middle Mio-
cene, Temblor Formation, Sharktooth Hill, Kern County, California.
Puffinus priscus Miller

MILLER, 1961: 399, fig. 1.

Plastoholotype humerus (dist.) UCMP 58185; cast C683; Middle Mio-
cene, Temblor Formation, Sharktooth Hill, Kern County, California.
Puffinus tedfordi Howard

HOWARD, 1971: 2, figs. 1A, IB, IE, IF.

Holotype tarsometatarsus (prox.) 15386 (figs. IB, IE) ; paratype tarso-
metatarsus 15387 (figs. 1A, IF); Early Pliocene, Almejas Formation, SE
corner Cedros Island, Baja California, Mexico; Loc. 65151.

PELECANIFORMES: ELOPTERYGIDAE
Elopteryx nopcsai Andrews

ANDREWS, 1913: 195, figs. 1-2.

Plastoholotype femur (prox.) BM A 1234 (fig. 1); cast C699; plasto-
paratype tibiotarsus (dist.) BM A1234 (fig. 2); cast C700; Late Creta-
ceous (Maestrichtian) Szentpeterfalva near Hatszeg, Transylvania,
Rumania.

LAMBRECHT, 1929: 1266, figs. 2-10 (p. 1263).

Plastohypotypes 2 tibiotarsi (dist.) BM A1588 (figs. 2, 6, 9, 10); cast
C702; BM A1528 (figs. 3, 5, 7, 8); cast C701; Late Cretaceous (Mae-
strichtian) Szentpeterfalva near Hatszeg, Transylvania, Rumania.

PELECANIFORMES: CYPHORNITHIDAE
Palaeochenoides mioceanus Shufeldt
SHUFELDT, 1916: 347, pi. 15.

Plastoholotype femur (dist.) YPM 2176; cast C742; Early Miocene,
Hawthorne Formation, Stono River, Charleston County, South Carolina.
HOPSON, 1964: 8, fig. 2.

Plastohypotype (tent.) tarsometatarsus (dist.) MCZ 2514; cast C741;
Early Miocene, Hawthorne Formation, Ashley River, Charleston County,
South Carolina.

PELECANIFORMES : PSEUDODONTORNITHIDAE
Osteodontornis orri Howard

HOWARD, 1957a: 3, figs. 2-8.

6

Contributions in Science

No. 228

Plastoholotype nearly complete skeleton in shale SBMNH 309 (skull,
figs. 4, 6; foot bones, figs. 7, 8; complete skeleton, figs. 2, 3); casts (sec-
tions from skeleton) C703-C714; Miocene, flagstone quarry, west side
Tepusquet Creek, Santa Barbara County, California.

HOWARD and WHITE, 1962: 4-11, figs. 2, 3, 5.

Hypotypes upper and lower jaw fragments 2707B and 2707A (figs. 2, 3),
atlas vertebra 2707D (fig. 5); Miocene diatomaceous shales, Del Gado
Drive near Sepulveda and Ventura Blvds., Sherman Oaks, Los Angeles
County, California; Loc. 1267.

Pseudodontornis stirtoni Howard and Warter

HOWARD and WARTER, 1969: 348, pis. 1-3.

Plastoholotype incomplete skull and jaws CM AV20569; cast C690;
?Pliocene Greta Siltstone, Waitotaran Stage; concretion found on Motu-
nau Beach, 36 miles north of Christchurch, New Zealand.

PELECANIFORMES: SULIDAE

Miosula media Miller

MILLER, 1925b: 114, pi. 5.

Plastoholotype impression of incomplete skeleton UCMP 26543; cast (in
relief) C696; Miocene diatomaceous shales, Lompoc, Santa Barbara
County, California.

Miosula recentior Howard

HOWARD, 1949b: 190, pi. 2, figs. 1, 2.

Holotype tibiotarsus 2117 (pi. 2, figs. 2, 2a) partype ulna (prox.) 2112
(pi. 2, fig. 1) (see Sula humeralis ); Pliocene, San Diego Formation, San
Diego, California; Loc. 1071.

Moris reyana Howard

HOWARD, 1936: 213, figs. 37a-b.

Holotype coracoid 991 (figs. 37a-b); “paratype” pedal phalanx 996;
Late Pleistocene, Lincoln Blvd., Del Rey Hills, northeast of Playa del Rey,
Los Angeles County, California; Loc. 1024.

HOWARD, 1949a: 21, 24.

Hypotypes tarsometatarsus (prox.) 2052, radius 2043; Late Pleistocene,
Newport Bay Mesa, Orange County, California; Loc. 1066.

Moris vagabundus Wetmore

HOWARD, 1966d: 5, figs. 1A, IB, 1J.

Hypotypes humerus 7432 (figs. 1A, 1 J ) , ulna (prox.) 16473 (fig. IB),
3 humeri (incomplete) 16467, 13980, 16471, 2 ulnae (prox.) 16472,
16470; Middle Miocene, Temblor Formation, Sharktooth Hill, Kern
County, California.

Sula humeralis Miller and Bowman

MILLER and BOWMAN, 1958: 9.

“Paratypes” femur 2522, ulna (prox.) 2112; Pliocene, San Diego Forma-
tion, San Diego, California.

1972

Type Specimens of Avian Fossils

7

Sula lompocana Miller

MILLER 1925b: 114, pi. 4.

Plastoholotype impression of incomplete skeleton UCMP 26544; cast (in
relief) C697; Miocene, diatomaceous shales, Lompoc, Santa Barbara
County, California.

Sula pohli Howard

HOWARD, 1958: 4, fig. 1.

Holotype wing bones on slab 2674 (fig. 1); “paratype” humerus 2532;
Middle Miocene, Ventura Blvd. between Whitsett and Coldwater Canyon
Road, Studio City, Los Angeles County, California; Loc. 1229.

Sula stocktoni Miller

MILLER, 1935: 75, fig. 2.

Plastoholotype part skeleton in shale UCMP 32105; cast C743; Miocene,
Lomita diatomite, Los Angeles County, California.

HOWARD, 1958: 12, fig. 3.

Hypotype humerus 2533; Miocene, Round Drive near Chester St., El

Sereno, Los Angeles County, California; Loc. 6455.

Sula willetti Miller

MILLER, 1925b: 112, pi. 3.

Plastoholotype impression of nearly complete skeleton UCMP 26542;
cast (in relief) C698; Miocene, diatomaceous shales, Lompoc, Santa
Barbara County, California.

PELECANIFORMES: PLOTOPTERIDAE
Plotopterum joaquinensis Howard
HOWARD, 1969a: 68, fig. 1.

Holotype coracoid (dist.) 8927; Early Miocene, Vaqueros Formation,
Pyramid Hill, Kern County, California; Loc. 1626.

PELECANIFORMES: PHALACROCORACIDAE
Graculus macropus Cope

COPE, 1878: 386 (figured, SHUFELDT, 1892, pi. 15, figs. 7, 8; lecto-
type selected, HOWARD, 1946: 153).

Plastolectotype tarsometatarsus AMNH 3555; cast C665; Late Pleisto-
cene, Fossil Lake, Oregon.

Phalacrocorax femoralis Miller
MILLER, 1929: 167, fig. 58.

Plastoholotype posterior skeletal impression UCLA/ VP 2754; cast C736;
Late Miocene, Modelo Formation, Poyer quarry, near Calabasas, Los
Angeles County, California.

Phalacrocorax goletensis Howard

HOWARD, 1965a: 51, figs. 1A-1D.

Holotype coracoid 4632 (figs. 1A-1D); “paratype” humerus (dist.)

8

Contributions in Science

No. 228

3166; Pliocene, Goleta Formation, Morelia lacustrine basin near La
Goleta, Morelia, Michoacan, Mexico; Loc. 1136.

Phalacrocorax kennelli Howard

HOWARD, 1949b: 188, pi. 3, figs. 7-8.

Holotype coracoid (dist.) 2127 (pi. 3, figs. 7, 7a); “paratype” humerus
(prox.) 2121 (pi. 3, figs. 8, 8a); Pliocene, San Diego Formation, San
Diego, California; Loc. 1080.

MILLER and BOWMAN, 1958: 12, fig. 3.

Hypotypes tibiotarsus 2566 (prox.) (fig. 3), femur 2528, ulna 2529;
Pliocene, San Diego Formation, San Diego, California.

Ardea paloccidentalis Shufeldt

SHUFELDT, 1892: 411, pi. 17, fig. 31.

Plastoholotype tarsometatarsus (dist.) AMNH 3484; cast C670; Late
Pleistocene, Fossil Lake, Oregon.

ARDEIFORMES : CICONIIDAE
Ciconia maltha Miller

MILLER, 1932: 215, fig. 23C.

Hypotype lower mandible (CIT)293; Late Pleistocene, McKittrick
asphalt deposits, Kern County, California; Loc. (CIT)138.

MILLER, 1938: 458, pi. 37B.

Hypotype cranium (CIT)1894; Late Pleistocene, McKittrick asphalt
deposits, Kern County, California; Loc. (CIT)138.

HOWARD, 1942: 193-195, figs. 1, la.

Hypotype rostrum (CIT)1894; Late Pleistocene McKittrick asphalt de-
posits, Kern County, California; Loc. (CIT)138.

Jabiru? weillsi Sellards

SELLARDS, 1916: 146, pi. 26, fig. 1.

Plastoholotype humerus USNM (FGS) 5961; cast C682; Late Pleisto-
cene, stratum 2, canal bank, Vero, Florida.

Mycteria wetmorei Howard

HOWARD, 1935b: 253, fig. 47.

Holotype lower mandible (frag.) K3527 (fig. 47, 1 and 2); “paratype”
tarsometatarsus (prox.) K3528 (fig. 47, 3 and 4); Late Pleistocene,
Rancho La Brea, Los Angeles, California.

PHOENICOPTERIFORMES : PALAELODIDAE
Megapaloelodus connectens A. Miller
A. MILLER, 1944: 86, figs. 1-2.

Plastoholotype tarsometatarsus (dist.) UCMP 37367; cast C689; Early
Miocene, Upper Rosebud Formation, Flint Hill, Bennett County, South
Dakota; UCMP Loc. V3417.

Megapaloelodus opsigonus Brodkorb

HOWARD, 1971: 6, figs. IK, 1M, IN.

1972

Type Specimens of Avian Fossils

9

Hypotype (tent.) tarsometatarsus (dist.) 15423; Early Pliocene, Almejas
Formation, SE corner Cedros Island, Baja California, Mexico; Loc.
65148.

PHOENICOPTERIFORMES : PHOENICOPTERIDAE
Phoenicopterus minutus Howard

HOWARD, 1955b: 202, pi. 50, figs. 1-7.

Holotype tibiotarsus (fig. 3-7) and associated (prox.) tarsometatarsus
(figs. 1, 2) 2445; “paratype” tarsometatarsus (prox.) 2473; Pleistocene,
Manix Lake, Mohave Desert, California; Loc. 1093.

Phoenicopterus stocki Miller

MILLER, 1944b: 77, figs. 1-2.

Holotype tibiotarsus (dist.) (CIT)3245 (fig. 1); “paratypes” (cata-
logued subsequent to publication) tibiotarsi (prox.) 4623 (fig. 2), (dist.)
4624, 4626, humeri (dist.) 4629, 4630, ulna (prox.) 4627, carpometa-
carpus 4628, tarsometatarsus (dist.) 4625; Middle Pliocene, Rincon-
Yepomera area, Chihuahua, Mexico; Loc. (CIT)289.

HOWARD, 1966a: 3.

Hypotypes scapula (frag.) 9731, radii (prox.) 9732, (dist.) 9733; Middle
Pliocene, Rincon- Yepomera area, Chihuahua, Mexico; Locs. (CIT)289
and 276.

ANSERIFORMES : ANATIDAE: CYGNINAE
Cygnus paloregonus Cope

HOWARD, 1946: 162, 164.

Plastohypotypes furcula AMNH 3536, carpometacarpus AMNH 3554;
casts C666 and C664; Late Pleistocene, Fossil Lake, Oregon.

Olor matthewi Shufeldt

SHUFELDT, 1913: 151, pi. 35, fig. 422.

Plastosyntype, carpometacarpus AMNH 3554 (see Cygnus paloregonus ) ;
cast C664; Late Pleistocene, Fossil Lake, Oregon.

ANSERIFORMES: ANATIDAE: ANSERINAE
Anser condoni Shufeldt

SHUFELDT, 1892: 406, pi. 16, fig. 19.

Plastoholotype furcula AMNH 3536 (see Cygnus paloregonus Cope);
cast C666; Late Pleistocene, Fossil Lake, Oregon.

Brant a dickey i Miller

MILLER, 1944a: 27, fig. 6.

Hypotype coracoid (CIT)3236; Pliocene, Owyhee, east side Dry Creek,
Malheur County, Oregon; Loc. (CIT)62.

Branta minuscula Wetmore

WETMORE, 1924: 6, figs. 3-4.

Plastoholotype humerus (prox.) USNM 10548; cast C679; Early Pleisto-
cene (late Pliocene?) 2 miles south of Benson, Arizona.

10

Contributions in Science

No. 228

Branta propinqua Shufeldt

SHUFELDT, 1892: 407, pi. 15, fig. 17.

Plastoholotype humerus AMNH 3547; cast C667; Late Pleistocene,
Fossil Lake, Oregon.

Eremochen russelli Brodkorb

HOWARD, 1966a: 4, fig. 1J.

Hypotype (tent.) scapula 9734; Middle Pliocene, Rincon- Yepomera
area, Chihuahua, Mexico; Loc. (CIT)289.

Presbychen abavus Wetmore

HOWARD, 1966d: 8, figs. 1D-1F (p.3).

Hypotype tarsometatarsus (prox.) 16466; Middle Miocene, Temblor
Formation, Sharktooth Hill, Kern County, California; Loc. 1625.

ANSERIFORMES : ANATIDAE: TADORNINAE
Anabernicula gracilenta Ross
ROSS, 1935: 107, fig. 6.

Holotype tarsometatarsus (CIT)1169 (fig. 6); paratypes two tarsometa-
tarsi (CIT)1168, (CIT)1170; “paratypes” tarsometatarsi (CIT)1171-
1175; Late Pleistocene, McKittrick asphalt deposits, Kern County,
California; Loc. (CIT)138.

HOWARD, 1964b: 286, pi. 7A-H.

Hypotype humerus 27349 (pi. 7A, 7B); Late Pleistocene, McKittrick
asphalt deposits, Kern County, California; Loc. (CIT)138. Hypotypes
carpometacarpus K4744 (pi. 7C, 7D), femur K4789 (pi. 7E, 7F),
tarsometatarsus K4797 (pi. 7G, 7H); Late Pleistocene, Rancho La Brea,
Los Angeles, California.

Anabernicula oregonensis Howard

HOWARD, 1964d: 5, figs. 1A, IB.

Plastoholotype humerus AMNH 3548; cast C676 (figs. 1A, IB), “para-
type” coracoid (CIT)3279; Late Pleistocene, Fossil Lake, Oregon.
Brantadorna downsi Howard

HOWARD, 1963: 8, pi. 1, figs. G, H, I.

Holotype humerus (prox.) 3911 (fig. G); paratype coracoid (dist.)
3910 (figs. H, I); “paratype” humerus (dist.) 3911; Middle Pleistocene,
Upper Palm Spring Formation, Mesquite Oasis, Vallecito Creek, Anza-
Borrego Desert, San Diego County, California; Loc. 1323.

ANSERIFORMES: ANATIDAE: ANATINAE
Nettion bunkeri Wetmore

HOWARD, 1966a: 7, figs. IF, 1G.

Hypotype coracoid 4621; Middle Pliocene, Rincon- Yepomera area, Chi-
huahua, Mexico; Loc. (CIT)289.

Wasonaka yepomerae Howard

HOWARD, 1966a: 5, figs. 1A-1E, 1H, 1-1.

1972

Type Specimens of Avian Fossils

11

Holotype humerus 4620 (figs. 1A, IB); paratypes furcula 4618 (figs. 1H,
1-1), ulna 4619 (figs. 1C, ID, IE); Middle Pliocene, Arroyo de las Bar-
rancas Blancas, Va mile east of Yepomera, Chihuahua, Mexico; Loc.
(CIT)286.

ANSERIFORMES : ANATIDAE: MERGINAE
Bucephala fossilis Howard

HOWARD, 1963: 11, pi. 1, figs. A-C.

Holotype carpometacarpus (prox.) 2787 (figs. A, B); paratype humerus
(prox.) 2885 (fig. C); “paratypes” two carpometacarpi (prox. 2886,
2887); Middle Pleistocene, Upper Palm Spring Formation, Arroyo
Tapiado, Vallecito Creek, Anza-Borrego Desert, San Diego County,
California; Loc. 1430.

Chendytes lawi Miller

HOWARD, 1947: 76, fig. 15.

Hypotypes coracoid (dist.) 2042 (fig. 15), humerus 2030; Late Pleisto-
cene, Newport Bay Mesa, Orange County, California; Loc. 1066.
HOWARD, 1949a: 21 and 25.

Hypotypes pelvis and synsacrum (frag.) 2055, 3 pedal phalanges 2025;
Late Pleistocene, Newport Bay Mesa, Orange County, California; Loc.
1066.

HOWARD, 1955a: 136, figs, lb, lc, lh, 2a, 2d.

Hypotypes humerus 2455 (figs, lb, lc), premaxilla 2059, femur 2015
(figs. 2a, 2d); Late Pleistocene Newport Bay Mesa, Orange County,
California; Loc. 1066. Hypotype scapula 2006 (fig. lh) ; Late Pleistocene,
Lincoln Blvd., Del Rey Hills, northeast of Playa del Rey, Los Angeles
County, California; Loc. 1024.

MILLER, MITCHELL and LIPPS, 1961: 4-10, pis. 1-2.

Hypotypes coracoid 2697 (pi. 1, fig. b), humerus 2698 (pi. 1, fig. c),
cranium and part lower jaw (missing) (pi. 1, figs, a, d) and associated
atlas, axis and cervical vertebra 2699, pelvis 2696 (pi. 2, figs, a, b), eight
associated vertebrae 2702; Late Pleistocene, north shore of east end of
West Anacapa Island, California.

HOWARD, 1964c: 372-376, fig 1.

Hypotypes sternum 2725 (figs, la, lh), humerus 4868 (figs, lb, lc),
ulnae 2736 (fig. Id), 2764, carpometacarpus 5536 (figs, le, If, Ig),
scapulae 2713, 2733, 2733a, 5538, coracoid 2730; Late Pleistocene,
north shore of east end of West Anacapa Island, California.

Chendytes milleri Howard

HOWARD, 1955a: 137, figs. 1-3.

Holotype humerus 2364 (figs, la, Id); paratypes femur 2378 (figs. 2b,
2c), ulna 2387 (figs. If, lg), scapula 2386 (figs, le, li); “paratypes”
incomplete coracoids, scapulae, humeri, ulnae, pelvis, femora, tibiotarsi,
fibula, tarsometatarsi, phalanges and vertebrae 2379-2385, 2388-2390,

12

Contributions in Science

No. 228

2392-2415 (including illustrated pelvis 2395, figs. 3a, 3b); Early?
Pleistocene, north side San Nicolas Island, California; Loc. 1085.

ANSERIFORMES: ANATINAE: OXYURINAE
Oxyura bessomi Howard

HOWARD, 1963: 13, pi. 1, figs. D, E.

Holotype carpometacarpus 2785 (figs. D, E); “paratypes” ulna (dist.)
2784, coracoids 2535 and 4966, carpometacarpus (prox.) 2888; Middle
Pleistocene, Upper Palm Spring Formation, Vallecito Creek, Anza-
Borrego Desert, San Diego County, California.

FALCONIFORMES: TERATORNITHIDAE
Cathartornis gracilis Miller

MILLER, 1910: 14, figs. 4a, 4b (p. 9).

Plastoholotype tarsometatarsus UCMP 12598 (figs. 4a, 4b) ; cast C686;
plastocotype tarsometatarsus UCMP 12600; cast C687; Late Pleistocene,
Rancho La Brea, California.

Teratornis incredibilis Howard
HOWARD, 1952: 51, pi. 10.

Holotype cuneiform (07)5067; Late Pleistocene, Smith Creek Cave,
White Pine County, Nevada; Loc. (07)251.

HOWARD, 1963: 16, pi. 2 A, 2C.

Hypotype radius (dist.) 3803; Middle Pleistocene, Upper Palm Spring
Formation, Vallecito Creek, A nza Borrego Desert, San Diego County,
California; Loc. 1318.

HOWARD, 1972: (in press).

Hypotype (tent.) incomplete rostrum 26697; Late Pliocene (Blancan),
Fish Creek, Anza-Borrego Desert, San Diego County, California; Loc.
6747.

Teratornis merriami Miller

MILLER, 1925a: 87, pis. 1-4.

Hypotypes skull B1380 (pi. 1), furcula B1366 (pi. 2A-B), coracoid
B1369 (pi. 2C), sternum B1365 (pi. 2D-E, and pi, 3A), partial pelvis
B1368 (pi. 3B); humerus B1370 (pi. 3C-F), carpometacarpus B1373
(pi. 4A), femur B1374 (pi. 4C-D), tarsometatarsus D542 (pi. 4G-H),
wing phalanx B1376 (pi. 4B), tibiotarsus B1372 (pi. 4E-F); Late
Pleistocene, Rancho La Brea, Los Angeles, California.

FALCONIFORMES: VUL7URIDAE
Coragyps occidentalis mexicanus Howard
HOWARD, 1968a: 124.

Holotype tarsometatarsus 20455; paratypes 21 tarsometatarsi 3358 and
20307-20326, 38 coracoids 3354 and 20327-20363, 15 humeri 3352 and
20364-20377, 20 ulnae 3356 and 20378-20396, 21 carpometacarpi 3355

1972

Type Specimens of Avian Fossils

13

and 20397-20416, 23 femora 3353 and 20417-20438, 17 tibiotarsi 3357
and 20439-20454; Late Pleistocene, San Josecito Cave, Nuevo Leon,
Mexico; Loc. (CIT)192.

Gymnogyps amplus Miller

FISHER, 1944: 290, figs. 43, 45, 46.

Hypotypes cranium B5415 (figs. 43, 45, 46), rostrum B6513, mandible
B7591; Late Pleistocene, Rancho La Brea, Los Angeles, California.
Sarcorhamphus clarki Miller

MILLER and HOWARD, 1938: 171, pi. 2a-c.

Hypotypes cranium and rostrum K3158 (pi. 2a), cranium B2148 (pi. 2b,
2c) ; Late Pleistocene, Rancho La Brea, Los Angeles, California.
HOWARD, 1969b: 5.

Hypotype axis vertebra 4638; Late Pleistocene, Tequixquiac, Mexico;
Loc. (CIT)310.

Vultur kernensis Miller

MILLER, 1931: 70, fig. 16.

Holotype humerus (dist.) (CIT)454; Pliocene, Pozo Creek, Kern River
Divide, Kern County, California; Loc. (CIT)49.

FALCONIFORMES: ACCIPITRIDAE: BUTEONINAE

Aquila pliogryps Shufeldt

SHUFELDT, 1892: 416, p. 17, fig. 33.

Plastoholotype pedal phalanx 1, digit 1 AMNH 3471; cast C668; Late
Pleistocene, Fossil Lake, Oregon.

Aquila sodalis Shufeldt

SHUFELDT, 1892: 417, pi. 15, fig 5.

Plastoholotype tarsometatarsus (prox.) AMNH 3470; cast C663; Late
Pleistocene, Fossil Lake, Oregon.

Buteo typhoius Wetmore

WETMORE, 1923: 489, figs. 3, 4.

Plastoholotype tarsometatarsus (dist.) AMNH 1754; cast C680; Late
Miocene, Snake Creek beds, Sioux County, Nebraska.

Geranoaetus fragilis Miller

HOWARD, 1932: 16-25, pis. 1-6.

Hypotypes cranium D1184 (pi. 1, figs. 1, la), rostrum D1142 (pi. 1,
figs. 2, 2a), mandible D2029 (pi. 1, fig. 3), furcula C8184 (pi. 1, figs. 4,
4a), scapula C5485 (pi. 1, figs. 5, 5a, 5b), sternum C7929 (pi. 2, figs. 1,
la), coracoid E4079 (pi. 2, figs. 2, 2a, 2b), humerus C8735 (pi. 3, figs.
1, la), carpometacarpus E1091 (pi. 3, fig. 3), ulna C5261 (pi. 4, figs.
1, la, lb), radius D8354 (pi. 4, figs. 2, 2a), pelvis C6481 (pi. 5, figs. 1,
la, lb), femur C684 (pi. 5, figs. 2, 2a, 2b), tibiotarsus C7332 (pi. 6,
figs. 2, 2a, 2b, 2c), tarsometatarsus E893 (pi. 6, figs. 1, la, lb); Late
Pleistocene, Rancho La Brea, Los Angeles, California.

14

Contributions in Science

No. 228

Geranoaetus grinnelli Miller

HOWARD, 1932: 33-43, pis. 14-19.

Hypotypes crania E3600 (pi. 14, fig. 1) and D4284 (pi. 14, figs, la, lb),
rostrum F3071 (pi. 14, figs. 2, 2a), mandible C5852 (pi. 14, fig. 3),
furcula C2508 (pi. 14, figs. 4, 4a), sternum D5981 (pi. 15, figs. 1, la),
coracoid C5842 (pi. 15, figs. 2, 2a, 2b, 2c), scapula C4450 (pi. 15, figs.
3, 3a, 3b), humerus D2365 (pi. 16, figs. 1, la), carpometacarpus C1587
(pi. 16, figs. 2, 2a), ulna C1937 (pi. 17, figs. 1, la, lb), radius D9637
(pi. 17, figs. 2, 2a), pelvis C1036 (pi. 18, figs. 2, 2a, 2b), femur C1028
(pi. 18, figs. 1, la), tibiotarsus C3103 (pi. 19, figs. 1, la, lb), tarsometa-
tarsus C6804 (pi. 19, figs. 2, 2a, 2b); Late Pleistocene, Rancho La Brea,
Los Angeles, California.

Miohierax stocki Howard

HOWARD, 1944: 236, fig. 40.

Holotype tarsometatarsus (dist.) metatarsal 1 and 9 phalanges (CIT)
1396; Miocene, Tick Canyon Formation, Vasquez Canyon, Los Angeles
County, California; Loc. (CIT) 201.

Morphnus daggetti Miller

MILLER, 1915: 179, fig. 63.

Holotype tarsometatarsus K3114 (old no. A380); Late Pleistocene,
Rancho La Brea, Los Angeles, California.

MILLER, 1925a: 97, pi. 5, fig. F.

Hypotype tibiotarsus J9744; Late Pleistocene, Rancho La Brea, Los
Angeles, California.

HOWARD, 1932: 16 (footnote), text figs. 1A, IB.

Hypotype coracoid D1217; Late Pleistocene, Rancho La Brea, Los
Angeles, California.

Morphnus woodwardi Miller

HOWARD, 1932: 25-30, pis. 7-12.

Hypotypes cranium F3172 (pi. 7, figs. 1, la), rostrum C6846 (pi. 7, figs.
2, 2a), mandibular symphysis D1019 (pi. 7, fig. 3), furcula D3056
(pi. 7, figs. 4, 4a), coracoid D4676 (pi. 7, figs. 5, 5a, 5b), sternum
D2398 (pi. 8, figs. 1, la), scapula D4816 (pi. 8, figs. 2, 2a, 2b), humerus
D6743 (pi. 9, figs. 1, la), radius (prox.) C4224 (pi. 9, fig. 2), carpo-
metacarpus D1702 (pi. 9, fig. 3), ulnae (prox.) C9264 (pi. 10, figs. 1,
la, lb), and G7554 (pi. 10, fig. 2), (dist.) D5177 (pi. 10, figs. 3, 3a, 3b),
pelvis C8858 (pi. 11, figs. 1, la), tibiotarsus D1974 (pi. 11, figs. 2, 2a,
2b), femur Cl 111 (pi. 12, figs. 1, la, lb), tarsometatarsus C6644 (pi.
12, figs. 2, 2a, 2b); Late Pleistocene, Rancho La Brea, Los Angeles,
California.

Spizaetus willetti Howard

HOWARD, 1935a: 207, fig. 40.

Holotype tarsometatarsus (dist) (CIT) 1791; Late Pleistocene, Smith
Creek Cave, White Pine County, Nevada; Loc. (CIT) 251.

1972

Type Specimens of Avian Fossils

15

FALCONIFORMES: ACCIPITRIDAE: PALAEOPLANCINAE

Palaeoplancus sternbergi Wetmore
WETMORE, 1933: 7, figs. 15-16.

Plastoholotype (part) tarsometatarsus (dist.) from skeleton USNM
12479; cast C677; Middle Oligocene, Brule Formation, Plum Creek,
Niobrara County, Wyoming.

FALCONIFORMES: ACCIPITRIDAE: GYPAETINAE

Arikarornis macdonaldi Howard

HOWARD, 1966c: 2, figs. 1A-1D.

Holotype tarsometatarsus (dist.) 9357; Early Miocene, Middle Sharp’s
Formation, Sharp’s Cut-off Road, Shannon County, South Dakota; Loc.
1821.

Neogyps errans Miller

HOWARD, 1932: 45-62, pis. 20-25.

Hypotypes cranium C2053 (pi. 20, figs. 1, la), rostrum D4615 (pi. 20,
figs. 2, 2a), mandible C694 (pi. 20, fig. 4), furculae D6522 (pi. 20,
fig. 3), B8633 (pi. 20, fig. 3a), sternum Cl 118 (pi. 21, figs. 1, la), cora-
coid C5467 (pi. 21, figs. 2, 2a), scapula C7922 (pi. 21, figs. 3, 3a),
humerus C2946 (pi. 22, figs. 1, la), ulna C4049 (pi. 23, figs. 1, la, lb),
radii (prox.) C3849 (pi. 23, figs. 2, 2a), (dist.) C1528 (pi. 23, fig. 3),
carpometacarpus D3374 (pi. 22, fig. 2), pelvis Cl 3 14 (pi. 24, figs. 1,
la, lb), femur J7555 (pi. 25, figs. 2, 2a, 2b), tibiotarsus C4982 (pi. 25,
figs. 1, la, lb), tarsometatarsus F2017 (pi. 24, figs. 2, 2a); Late Pleisto-
cene, Rancho La Brea, Los Angeles, California.

Neophrontops americanus Miller

HOWARD, 1932: 62-70, pis. 26-29.

Hypotypes cranium D7752 (pi. 26, figs. 1, la), rostrum J9068 (pi. 26,
figs. 2, 2a), mandible C7398 (pi. 26, fig. 3), sternum E2033 (pi. 26,
figs. 4, 4a), furcula E3859 (pi. 26, figs. 5, 5a), coracoid E2661 (pi. 27,
fig. 1), scapula E3453 (pi. 27, figs. 2, 2a), humerus G1987 (pi. 27, figs.
3, 3a), ulna D8188 (pi. 28, figs. 1, la, lb), radius D7841 (pi. 28, figs. 2,
2a), carpometacarpus H2477 (pi. 27, fig. 4), pelvis E2051 (pi. 29,
figs. 1, la), femur D9765 (pi. 29, figs. 2, 2a, 2b), tibiotarsus F1958
(pi. 29, figs. 3, 3a, 3b), tarsometatarsus E2159 (pi. 29, figs. 4, 4a); Late
Pleistocene, Rancho La Brea, Los Angeles, California.

Neophrontops vallecitoensis Howard
HOWARD, 1963: 17, pi. 3, fig. B.

Holotype tarsometatarsus (dist.) (pi. 3B) with associated metatarsal
1 and 8 phalanges 2866; paratype tarsometatarsus (dist.) 3769; Middle
Pleistocene (Irvingtonian), Upper Palm Spring Formation; Vallecito
Creek, Anza-Borrego Desert, San Diego County, California; Locs. 1299
and 1356.

16

Contributions in Science

No. 228

FALCONIFORMES: FALCONIDAE
Polyborus prelutosus Howard

HOWARD, 1938: 226, pis. 1-3.

Holotype humerus E4398 (pi. 1, fig. 1, pi. 2, fig. 3); “paratypes”
humeri E3927 (pi. l,fig. 3, pi. 2, fig. 1),E4356 (pi. l,fig. 2, pi. 2, fig. 4),
E9852 (pi. 1, fig. 4), E3255 (pi. 1, fig. 5), E1318 (pi. 2, fig. 2), E1804
(pi. 2, fig. 5), rostrum E4485 (pi. 3, fig. 2), carpometacarpus E3556
(pi. 3, figs. 4, 4a), femora E1210 (pi. 3, fig. 6), E4012 (pi. 3, fig. 7), E651
(pi. 3, fig. 8), tarsometatarsi E681 (pi. 3, fig. 10), E3446 (pi. 3, fig. 11),
and 747 specimens not listed by catalog numbers. It is impossible to
trace all of these specimens. However, the following, derived from the
author’s notes, are representative of the unillustrated elements described :
coracoids E3080, E9884, H4545, H4606, ulnae E905, E1339, E1583,
E3367, pelves D9083, D9619, E4678, E9617, tibiotarsi E3954, E4267,
E4327, E4493; Late Pleistocene, Rancho La Brea, Los Angeles,
California.

Polyborus prelutosus grinnelli Howard
HOWARD, 1940: 41.

Holotype tarsometatarsus (CIT)2709; “paratypes” 10 tarsometatarsi
(CIT) 27 10-27 19, 3 humeri (CIT) 2720-2722, 3 ulnae (CIT) 2723-2725,
5 carpometacarpi (CIT) 2726-2730, 4 femora (CIT) 273 1-2734, 2 tibio-
tarsi (CIT) 2735-2736, coracoid (CIT)2737; Late Pleistocene, San Jose-
cito Cave, Nuevo Leon, Mexico; Loc. (CIT) 192.

GALLIFORMES : CRACIDAE
Procrax brevipes Tordoff and Macdonald

TORDOFF and MACDONALD, 1957: 179, pi. 10, fig. 1.
Plastoholotype incomplete skeleton in matrix SDSM 511; cast C538;
Early Oligocene, top of Chadron Formation, Pennington County, South
Dakota.

GALLIFORMES: TETRAONIDAE
Palaeotetrix gilli Shufeldt

SHUFELDT, 1892: 415, pi. 17, fig. 34.

Plastoholotype carpometacarpus AMNH 3474; cast C672; Late Pleisto-
cene, Fossil Lake, Oregon.

Pediocaetes lucasi Shufeldt

SHUFELDT, 1892: 414, pi. 17, fig. 30.

Plastoholotype ulna AMNH 3476; cast C675; Late Pleistocene, Fossil
Lake, Oregon.

GALLIFORMES: PHASIANIDAE
Miortyx aldeni Howard

HOWARD, 1966c: 5, fig. IE.

1972

Type Specimens of Avian Fossils

17

Holotype humerus (prox.) 9388; Early Miocene, Middle Sharp’s Forma-
tion, gully beside Sharp’s Cut-off Road, Shannon County, South Dakota;
Loc. 1982.

GALLIFORMES : MELEAGRIDIDAE
Agriocharis anza Howard

HOWARD, 1963: 19, pi. 3, fig. A.

Holotype humerus 3753 (pi. 3, fig. A); paratypes humerus (prox.),
sternum (frag.), sacrum and ulna collected with type 3753; Middle
Pleistocene, Upper Palm Spring Formation, Vallecito Creek, Anza-
Borrego Desert, San Diego County, California; Loc. 1358.

Meleagris eras sipes Miller

MILLER, 1940: 154, fig. 45 A.

Holotype tarsometatarsus (CIT)2708; Late Pleistocene, San Josecito
Cave, Nuevo Leon, Mexico; Loc. (CIT)192.

Pavo calif ornicus Miller

HOWARD, 1927: 3-27, pis. 1-13.

Hypotypes cranium E5226 (pi. 1, fig. 3, and pi. 2, fig. 1), two sterna
E5173 (pi. 3, fig. 1), E5691 (pi. 4, fig. 1), furcula J6535 (pi. 5, figs. 4
and 7), scapula E5445 (pi. 7, fig. 5, and pi. 8, fig. 1), coracoid E7239
(pi. 6, figs. 1, 5, and pi. 7, fig. 3), humerus E7108 (pi. 2, fig. 5), ulna
E6192 (pi. 8, fig. 5), radius D9790 (pi. 7, figs. 9, 13), carpometacarpus
E6666 (pi. 9, fig. 1), femur old no. 3 + 4 (pi. 9, fig. 5), tibiotarsus F6993
(pi. 10, fig. 1, pi. 11, fig. 1), tarsometatarsus E6839 (pi. 12, fig. I, pi. 13,
fig. 1 ) ; Late Pleistocene, Rancho La Brea, Los Angeles, California.
HOWARD, 1928: 90.

Hypotypes five beaks K2474-2478; Late Pleistocene, Rancho La Brea,
Los Angeles, California.

HOWARD, 1945: 597, pi. 25.

Hypotypes tarsometatarsi (age stages) K1681 (upper fig. a), K8364
(upper fig. b), G6282 (upper fig. c), E7224 (upper fig. d), E8569
(upper fig. e), E6732 (upper fig. f), E6697 (upper fig. g), E7737 (upper
fig. h), E6793 (lower fig. a), E5075 (lower fig. b), E6801 (lower fig. c),
E6173 (lower fig. d), E5333 (lower fig. 3); Late Pleistocene, Rancho La
Brea, Los Angeles, California.

GRUIFORMES: PHORUSRHACIDAE
Titanis walleri Brodkorb

BRODKORB, 1963: 113, fig. 2.

Plasto“paratype” pedal phalanx 1, digit 3 UF 4109; cast C427; Late
Pleistocene, Santa Fe River, Gilchrist/ Columbia County line, Florida.

18

Contributions in Science

No. 228

GRUIFORMES: RALLIDAE

Epirallus natator Miller

MILLER, 1942: 43, fig. la.

Holotype tarsometatarsus (CIT)2943; Late Pleistocene, San Josecito
Cave, Nuevo Leon, Mexico; Loc. (CIT) 192.

Fulica hesterna Howard

HOWARD, 1963: 22, pi. 1, fig. F (p. 10).

Holotype tibiotarsus (dist.) 2873 (pi. 1, fig. F) ; par atype tarsometatarsus
and 5 pedal phalanges 2873; “paratype” tibiotarsus (dist.) 2875; Middle
Pleistocene, Upper Palm Spring Formation, Vallecito Creek, Anza-
Borrego Desert, San Diego County, California; Locs. 1433 and 1299.
Fulica minor Shufeldt

SHUFELDT, 1892: 412, pi. 17, fig. 32.

Plastoholotype humerus AMNH 3480 (name preoccupied; see Fulica
shufeldti Brodkorb); cast C673; Late Pleistocene, Fossil Lake, Oregon.
Fulica shufeldti Brodkorb

BRODKORB, 1964b: 186.

Plastoholotype humerus AMNH 3480; cast C673; Late Pleistocene,
Fossil Lake, Oregon.

CHARADRIIFORMES : SCOLOPACIDAE

Palnumenius victima Miller

MILLER, 1942: 45, fig. lb.

Holotype tarsometatarsus (CIT) 2944; Late Pleistocene, San Josecito
Cave, Nuevo Leon, Mexico; Loc. (CIT) 192.

CHARADRIIFORMES: LARIDAE

Larus oregonus Shufeldt

SHUFELDT, 1892: 398, pi. 15, fig. 3.

Plastoholotype humerus (prox.) AMNH 3494; cast C662; Late Pleisto-
cene, Fossil Lake, Oregon.

Larus robustus Shufeldt

SHUFELDT, 1892: 398, pi. 15, figs. 1-2.

Plastoholotype coracoid AMNH 3497; cast C674; Late Pleistocene,
Fossil Lake, Oregon.

CHARADRIIFORMES: STERCORARIIDAE

Stercorarius shufeldti Howard

HOWARD, 1946: 184, pi. 2, figs. 1, 2.

Plastoholotype humerus AMNH 3491; cast of proximal end only C671;
Late Pleistocene, Fossil Lake, Oregon.

1972

Type Specimens of Avian Fossils

19

CHARADRIIFORMES: ALCIDAE: ALCINAE
Aethia rossmoori Howard

HOWARD, 1968b: 16, figs. 2-1, 2-J (p. 4).

Holotype ulna 18948 (fig. 2-J); “paratype” humerus (dist.) 18949 (fig.
2-1); Late Miocene, Leisure World, Laguna Hills, Orange County, Cali-
fornia; Loc. 1945.

Brachyramphus pliocenus Howard

HOWARD, 1949b: 191, pi. 3, figs. 1, 2.

Holotype humerus 2119 (pi. 3, fig. 2); “paratypes” cranium 2166 (pi. 3,
fig. 1), humerus (dist.) 2152, articular end mandible 2172; Pliocene, San
Diego Formation, San Diego, California.

Cerorhinca dubia Miller

MILLER, 1925b: 115, pi. 2.

Plastoholotype impression of leg bones in shale UCMP 26546; cast
(in relief) C695; Late Miocene, diatomaceous shales, Lompoc, Santa
Barbara County, California.

Cerorhinca minor Howard

HOWARD, 1971: 9, figs. ID, 1G, 1H, 1J.

Holotype humerus (prox.) 15408 (fig. 1 J ) ; “paratypes” ulna (prox.)
15406 (fig. 1G), tarsometatarsus 15407 (fig. ID), humerus (prox.)
15420, coracoid (dist.) 15421 (fig. 1H); Early Pliocene, Almejas For-
mation, SE corner Cedros Island, Baja California, Mexico; Locs. 65153
and 65148.

CHARADRIIFORMES: ALCIDAE: MAN C ALLIN AE
Alcodes ulnulus Howard

HOWARD, 1968b: 18, figs. 2G, 2H, 2L (p. 4).

Holotype ulna 18277 (fig. 2H) ; “paratypes” ulna (dist.) 18279, carpo-
metacarpus (prox.) 18278 (figs. 2G, 2L); Late Miocene, Leisure World,
Laguna Hills, Orange County, California; Loc. 1945.

Mancalla californiensis Lucas

LUCAS, 1901: 133, figs. 1, 2.

Plastoholotype humerus (prox.) USNM 4976; cast C685; Early Pliocene,
Repetto Formation, Third Street Tunnel, Los Angeles, California.
HOWARD, 1949b: 196, pi. 3, figs. 3, 3a, 4, 4a.

Hypotypes carpometacarpus (prox.) 2033 (figs. 3, 3a), tarsometatarsus
(shaft) 2034 (figs. 4, 4a), vertebra 2035; Early Pliocene, Repetto For-
mation, 3 miles north of Corona del Mar, Orange County, California;
Loc. 1067.

HOWARD, 1970: 2.

Hypotypes humeri (prox.) 2576, (dist.) 2577, coracoid 2581, ulna 2580,
radius 2579, carpometacarpus 2578, tibiotarsus 2424, tarsometatarsus
2250; Early Pliocene, Repetto Formation, 3 miles north of Corona del
Mar, Orange County, California; Loc. 1067.

20

Contributions in Science

No. 228

Mancalla cedrosensis Howard

HOWARD, 1971: 11, figs. 1L and 2A-K.

Holotype incomplete skeleton 15373 including essentially complete right
scapula (fig. 2H), coracoid (figs. 2C, 2G, 2J), humerus (fig. 2B), ulna,
femur (fig. 2K), and tibiotarsus (figs. 2A and 2D), left ulna (fig. 2E),
radius (fig. 2F) , carpometacarpus (fig. 21) , distal end right radius, carpo-
metacarpus and left tibiotarsus, proximal end right tarsometatarsus and
left femur, and fragmentary vertebrae; “paratypes” complete right tarso-
metatarsus (fig. 1L) and associated fragmentary pelvis, femur, tibiotarsi
and vertebrae 15425; associated leg bones 23739; associated furcula,
sternum, scapulae, coracoids, carpometacarpus 15410; and 50 separate
elements nos. 15364-15372, 15374-15385, 15388-15402, 15408-15409,
15411-15412, 15415-15419, 15424, 15427; Early Pliocene, Almej as For-
mation, Cedros Island, Baja California, Mexico.

Mancalla milleri Howard

HOWARD, 1970: 7, figs. 1A-1C.

Holotype femur 2185 (figs. 1A, IB; figured Miller and Howard, 1949,
pi. 5, fig. 3 as Mancalla diegense ); paratype humerus 2813 (fig. 1C);
“paratypes” (Miller and Howard, 1949, pis. 1-6, figured as M. diegense',
see Pliolunda diegense ) scapula 2070 (pi. 4, fig. 4), humeri 2066, 2096
(pi. 2, figs. 2, 3), ulnae 2069 (pi. 1, fig. 1) 2082 (pi. 1, fig. 2, pi. 3,
fig. 8), 2079 and 2101 (pi. 3, figs. 9-10), 2179 (pi. 5, fig. 2) , carpometa-
carpus 2068 (pi. 4, fig. 15), femur 2097 (pi. 4, fig. 1), tibiotarsi 2083,
2100, 2108, 2134 (pi. 4, figs. 8, 10, 13, 14), sterna 2063 (pi. 2, fig. 7),
2180 (pi. 6, fig. 1), pelvis 2182 (pi. 6, fig. 2) ; “paratypes” cranium 2204,
sterna 2325, 2661, scapulae 2167, 2252, 2257, 2278, 2297, 2506, 2536,
coracoids 2208, 2229, 2243, 2276, 2338, 2498a, 2555, 2559, 2627,
humeri 2096a and b, 2206, 2219, 2292, 2303, 2326, 2427, 2442, 2480a,
b, and c, 2504, 2553, 2679, 2850, ulnae 2342, 2439, 2484, 2497a and b,
2552, 2558, 6426, radii 2335, 2632, carpometacarpus 2825, femora
2508, 2848, tibiotarsi 2088, 2209, 2286, 2478, 2549, 2628, tarsometa-
tarsi 2327, 2488, 2548, 2682, 6454; Pliocene, San Diego Formation,
San Diego, California.

Pliolunda diegense Miller

MILLER and HOWARD, 1949: 201-228, pi. 1-6.

Hypotypes scapulae 2176, 2049 (pi. 4, figs. 5, 7), coracoids 2087, 2067
(pi. 3, figs. 2, 3), ulna 2064 (pi. 3, fig. 7), carpometacarpus (prox.)
2068 (pi. 4, fig. 16), tibiotarsi 2125 (pi. 4, fig. 9), 2177 (pi. 5, fig. 1),
tarsometatarsi 2178, 2177 (pi. 5, figs. 4-5); (see Mancalla milleri for
reassignment of other figured specimens); Pliocene, San Diego Forma-
tion, San Diego, California.

HOWARD, 1970: 7, fig. ID.

Hypotype humerus 2670; Pliocene, San Diego Formation, San Diego,
California.

1972

Type Specimens of Avian Fossils

21

Praemancalla lagunensis Howard

HOWARD, 1966b: 4, figs. 1A, C, D, E, G.

Holotype humerus (dist.) 15288 (fig. IE, 1G); paratype carpometa-
carpus (prox.) 15287 (fig. 1A); “paratypes” carpometacarpus (prox.)
15290, coracoid (dist.) 15289 (figs. 1C, ID), scapula 15294, lower
mandible (articular end) 15428; Late Miocene, Leisure World, Laguna
Hills, Orange County, California; Loc. 1945.

CUCULIFORMES : CUCULIDAE
Geococcyx conklingi Howard

HOWARD, 1931: 208, fig. 50.

Syntypes humerus (dist.) 118 (figs. 50c, 50c'), ulna 119 (fig. 50b),
femur 113 (figs. 50a, 50a'); “paratypes” femora (shafts) 114, 115,
humerus (shaft) 117, tibiotarsus (dist.) 116; Late Pleistocene, Conkling
Cavern, Dona Ana County, New Mexico; Loc. 1009.

STRIGIFORMES: PROTOSTRIGIDAE
Protostrix californiensis Howard

HOWARD, 1965b: 350, pi. 49, figs. 1, 3.

Holotype humerus 6171; Eocene, Poway Formation, 300 yards north of
intersection of Lake Shore and Jackson drives, San Diego, California;
Loc. 1723.

STRIGIFORMES: STRIGIDAE

Asio priscus Howard

HOWARD, 1964a: 28, fig. 1.

Holotype tibiotarsus 4712; Late Pleistocene, Arlington Canyon, Santa
Rosa Island, California; Loc. (CIT) 106.

Strix brea Howard

HOWARD, 1933: 66, fig. 15.

Holotype tarsometatarsus E9379 (fig. 15); “paratypes” rostra C7125,
K2713, sterna E2477, D9615, coracoids E9273, E9687, H4850, H4872,
H4881, H4889, H4904, H4911, H4923, scapulae E2720, H6610, H6613,
H6629, H6636, H6656, H6659, H6673, humeri E8911, E9051, E9425,
E9804, F9305, G1229, carpometacarpi H3096-H3098, H3107, H3126.
femora E9439, E9647, E9909, F4884, tibiotarsi El 139, E9267, E9414,
E9545, E9606, E9758, E9888, E9919, E9932, E9942, F7456, tarso-
metatarsi E9416, E9417, E9575, E9892, E9911, G3933, G3957, G3958;
Late Pleistocene, Rancho La Brea, Los Angeles, California.

PASSERIFORMES: PALAEOSCINIDAE
Palaeoscinis turdirostris Howard

HOWARD, 1957b: 6, figs. 1-2.

Holotype complete skeleton on two slabs of matrix 2604; Miocene,
Tepusquet Creek, Santa Barbara County, California; Loc. 1127.

22

Contributions in Science

No. 228

PASSERIFORMES: ICTERIDAE
Pandanaris convexa A. Miller

A. MILLER, 1947: 22, fig. 4a-d.

Holotype upper mandible K7278 (figs. 4a, 4b, 4c); “paratype” lower
mandible K7279 (figs. 4b, 4d); Late Pleistocene, Rancho La Brea, Los
Angeles, California.

Pyelorhamphus molothroides A. Miller
A. MILLER, 1932: 39, pi. 4, figs. 1-5.

Holotype lower mandible 320 (pi. 4, figs. 1, 2, 3); “paratype” upper
mandible 338 (pi. 4, figs. 2, 4, 5); Quaternary (?Late Pleistocene),
Shelter Cave, Dona Ana County, New Mexico; Loc. 1010.

PASSERIFORMES: FRINGILLIDAE

Pipilo angelensis Dawson

DAWSON, 1948: 59, fig. 16.

Holotype upper mandible K7291 (fig. 16); paratype upper mandible
K7292; “paratypes” six upper mandibles, all K7293; Late Pleistocene,
Rancho La Brea, Los Angeles, California.

Literature Cited

Andrews, C. W. 1913. On some bird remains from the Upper Cretaceous of Tran-
sylvania. Geol. Mag. decade 5, 10 (5): 193-196, figs. 1-2.

Brodkorb, P. 1953. A review of the Pliocene loons. Condor 55:211-214, fig. 1.

1963a. Catalogue of Fossil Birds, Part 1. Bull. Fla. St. Mus., Biol. Sci. 7

(4): 179-293.

1963b. A giant flightless bird from the Pleistocene of Florida. Auk

80:111-115, figs. 1-2.

1964a. Catalogue of Fossil Birds, Part 2. Bull. Fla. St. Mus., Biol. Sci. 8

(3): 195-335.

— . 1964b. A new name for Fulica minor Shufeldt. Quart. J. Fla. Acad. Sci.,

27 (3): 186.

_. 1967. Catalogue of Fossil Birds, Part 3. Bull. Fla. St. Mus., Biol. Sci. 11

(3) : 99-220.

1971. Catalogue of Fossil Birds, Part 4. Bull. Fla. St. Mus., Biol. Sci.

15 (4): 163-266.

Cope, E. D. 1878. Descriptions of new extinct vertebrata from the Upper Tertiary
and Dakota formations. Bull. U.S. Geol. and Geog. Surv. Terr. 4 (2):379-396.
Dawson, W. R. 1948. Records of fringillids from the Pleistocene of Rancho La Brea.
Condor 50:57-63, fig. 16.

Fisher, H. I. 1944. The skulls of the cathartid vultures. Condor 46:272-296, figs.
42-47.

Hopson, J. A. 1964. Pseudodontornis and other large marine birds from the Mio-
cene of South Carolina. Postilla, Peabody Mus. Nat. Hist., Yale Univ. No. 83,
19 p., figs. 1-3.

Howard, H. 1927. A review of the fossil bird, Parapavo calif ornicus (Miller) from
the Pleistocene asphalt beds of Rancho La Brea. Univ. Calif. Publ. Bull. Dept.
Geol. Sci. 17:1-56, pis. 1-13.

1928. The beak of Parapavo californicus (Miller). Bull. So. Calif. Acad.

Sci. 27:90-91.

1972

Type Specimens of Avian Fossils

23

1931. A new species of road-runner from Quaternary cave deposits in

New Mexico. Condor 33:206-209, figs. 48-50.

1932. Eagles and eagle-like vultures of the Pleistocene of Rancho La Brea.

Carnegie Inst. Wash., Publ. 429, 82 p., 3 figs, in text, 29 pis.

1933. A new species of owl from the Pleistocene of Rancho La Brea,

California. Condor 35:66-69, fig. 15.

1935a. A new species of eagle from a Quaternary cave deposit in eastern

Nevada. Condor 37:206-209, fig. 40.

1935b. The Rancho La Brea Wood Ibis. Condor 37:251-253, fig. 47.

1936. A new fossil bird locality near Playa del Rey, California, with

description of a new species of sulid. Condor 38:211-214, fig. 37.

1938. The Rancho La Brea Caracara, a new species. Carnegie Inst. Wash.,

Publ. 487:217-240, 3 pis.

1940. A new race of Caracara from the Pleistocene of Mexico. Condor

42:41-44.

1942. A review of the American fossil storks. Carnegie Inst. Wash., Publ.

530:187-203, 2 text figs., 1 pi.

1944. A Miocene hawk from California. Condor 46:236-237, fig. 40.

1945. Observations on young tarsometatarsi of the fossil turkey, Parapavo

calif ornicus (Miller). Auk 62:596-603, 1 text fig., pi. 25.

1946. A review of the Pleistocene birds of Fossil Lake, Oregon. Carnegie

Inst. Wash., Publ. 551:141-195, 2 pis.

1947. Wing elements assigned to Chendytes. Condor 49:76-77, fig. 15.

1949a. Avian fossils from the marine Pleistocene of southern California.

Condor 51:20-28.

1949b. New avian records for the Pliocene of California. Carnegie Inst.

Wash., Publ. 584:177-199, 3 pis.

1952. The prehistoric avifauna of Smith Creek Cave, Nevada, with a

description of a new gigantic raptor. Bull. So. Calif. Acad. Sci. 51 (2):50-54,
pi. 10.

1955a. New records and a new species of Chendytes, an extinct genus

of diving geese. Condor 57:135-143, figs. 1-3.

1955b. Fossil birds from Manix Lake, California. U.S. Geol. Surv. Prof.

Paper No. 264-1:199-205, pi. 50, and text fig. 42.

1957a. A gigantic “toothed” marine bird from the Miocene of California.

Santa Barbara Mus. Nat. Hist., Bull. Dept. Geol. 1 : 1-23, figs. 1-8.

1957b. A new species of passerine bird from the Miocene of California.

Los Angeles Co. Mus., Contrib. Sci. 9:1-16, figs. 1-2.

1958. Miocene sulids of southern California. Los Angeles Co. Mus.,

Contrib. Sci. 25:1-15, figs. 1-3.

1962. Fossil Birds. Los Angeles Co. Mus., Sci. Ser. 17, Paleontol. 10, 44 p.,

22 figs.

1963. Fossil birds from the Anza-Borrego Desert. Los Angeles Co. Mus.,

Contrib. Sci. 73:1-33, 1 fig., 3 pis.

1964a. A fossil owl from Santa Rosa Island, California, with comments on

the eared owls of Rancho La Brea. Bull. So. Calif. Acad. Sci. 63(1) : 27-3 1 , 1 fig.

1964b. Fossil Anseriformes. In “Waterfowl of the World,” by Jean Dela-

cour, vol. 4, p. 233-326, 5 text figs., 10 pis. Country Life Ltd., London.

1964c. Further discoveries concerning the flightless “diving geese” of the

genus Chendytes. Condor 66:372-376, 1 fig.

1964d. A new species of the “Pigmy Goose,” Anabernicula, from the

Oregon Pleistocene, with a discussion of the genus. Amer. Mus. Novitates
2200:1-14, 2 figs.

24

Contributions in Science

No. 228

1965a. A new species of cormorant from the Pliocene of Mexico. Bull. So.

Calif. Acad. Sci. 64 (l):50-55, 1 fig.

- 1965b. First record of avian fossils from the Eocene of California.

J. Paleontol. 39(3) :350-354, pi. 49.

- 1966a. Pliocene birds from Chihuahua, Mexico. Los Angeles Co. Mus.,
Contrib. Sci. 94:1-12, 1 fig.

— 1966b. A possible ancestor of the Lucas Auk (Family Mancallidae) from

the Tertiary of Orange County, California. Los Angeles Co. Mus., Contrib. Sci.
101:1-8, 1 fig.

1966c. Two fossil birds from the Lower Miocene of South Dakota. Los

Angeles Co. Mus., Contrib. Sci. 107:1-8, 1 fig.

1966d. Additional avian records from the Miocene of Sharktooth Hill,

California. Los Angeles Co. Mus., Contrib. Sci. 114:1-11, 1 fig.

. 1968a. Limb measurements of the extinct vulture, Coragyps occidentalis,

with a description of a new subspecies. Papers Archaeol. Soc. New Mex.
1:115-128.

... 1968b. Tertiary birds from Laguna Hills, Orange County, California. Los

Angeles Co. Mus., Contrib. Sci. 142:1-21, 2 figs.

- 1969a. A new avian fossil from Kern County, California. Condor

71:68-69, 1 fig.

1969b. Avian fossils from three Pleistocene sites in central Mexico. Los

Angeles Co. Mus., Contrib. Sci. 172:1-11, 1 fig.

1970. A review of the extinct avian genus, Mancalla. Los Angeles Co.

Mus., Contrib. Sci. 203:1-12, 1 fig.

1971. Pliocene avian remains from Baja California. Los Angeles Co. Mus.,

Contrib. Sci. 217:1-17, 2 figs.

1972. The Incredible Teratorn again. Condor 74: (in press).

Howard, H., and S. L. Warter. 1969. A new species of bony-toothed bird ( Family
Pseudodontornithidae) from the Tertiary of New Zealand. Rec. Canterbury
Mus. 8(4) : 345-357, 4 pis.

Howard, 11., and J. A. White. 1962. A second record of Osteodontornis, Miocene
“toothed” bird. Los Angeles Co. Mus., Contrib. Sci. 52:1-12, 5 figs.

International Commission on Zoological Nomenclature. 1964. International
code of zoological nomenclature, adopted by the XV International Congress of
Zoology. Int. Trust Zool. Nomenclature, London. 176 p.

Lambrecht, K. 1928. Mesozoische und tertiare Vogelreste aus Siebenburgen.
Proc. X International Congr. Zool., Budapest, 1927, Sect. VIII, Paleozoologie
et Zoogeographic, p. 1262-1275, figs. 1-16.

Lucas, F. A. 1901. A flightless auk, Mancalla calif orniensis, from the Miocene of
California. Proc. U.S. Nat. Mus. 24(1245) : 133-134, figs. 1-2.

Marsh, O. C. 1870. Notice of some fossil birds from the Cretaceous and Tertiary
formations of the United States. Amer. J. Sci. ser. 2, 49(46) : 205-2 17.

Miller, A. H. 1932. An extinct icterid from Shelter Cave, New Mexico. Auk
49:38-41, pi. 4.

... .... 1947. A new genus of icterid from Rancho La Brea. Condor 49:22-24,

fig. 4.

Miller, L. H. 1910. The condor-like vultures of Rancho La Brea. Univ. Calif. Publ.
Bull. Dept. Geol. 6:1-19, figs. 1-5.

1915. A walking eagle from Rancho La Brea. Condor 17:179-181, fig. 63.

— . 1925a. The birds of Rancho La Brea. Carnegie Inst. Wash., Publ. 349:

63-106, pis. 1-6, 20 text figs.

1925b. Avian remains from the Miocene of Lompoc. Carnegie Inst.

Wash., Publ. 349:107-117, pis. 1-9, 1 text fig.

1929. A new cormorant from the Miocene of California. Condor 31:

167-172, figs. 58-59.

1972

Type Specimens of Avian Fossils

25

1931. Bird remains from the Kern River Pliocene of California. Condor

33:70-72, fig. 16.

1932. The Pleistocene storks of California. Condor 34:212-216, fig. 23.

1935. New bird horizons in California. Publ. Univ. Calif. Los Angeles,

Biol. Sci. 1(5) :73-80, 2 figs.

1938. A study of the skull of the Pleistocene stork, Ciconia maltha Miller.

Trans. San Diego Soc. Nat. Hist. 8:455-462, pi. 37.

1940. A new Pleistocene turkey from Mexico. Condor 42: 154-156, fig. 45.

1942. Two new bird genera from the Pleistocene of Mexico. Univ. Calif.

Publ. Zool. 47:43-46, fig. 1.

1944a. Some Pliocene birds from Oregon and Idaho. Condor 46:25-32,

fig. 6.

, 1944b. A Pliocene flamingo from Mexico. Wilson Bull. 56:77-82, figs. 1-2.

1961. Birds from the Miocene of Sharktooth Hill, California. Condor

63:399-402, fig. 1.

Miller, L., and R. I. Bowman. 1958. Further bird remains from the San Diego
Pliocene. Los Angeles Co. Mus., Contrib. Sci. 20:1-15, figs. 1-5.

Miller, L., and H. Howard. 1938. The status of the extinct condorlike birds of
the Rancho La Brea Pleistocene. Publ. Univ. Calif. Los Angeles, Biol. Sci. 1 :
169-176, pi. 2, 2 text figs.

1949. The flightless Pliocene bird Mancalla. Carnegie Inst. Wash. Publ.

584(7) :201-228, pis. 1-6.

Miller, L., E. D. Mitchell, and J. H. Lipps. 1961. New light on the flightless
goose, Chendytes lawi. Los Angeles Co. Mus., Contrib. Sci. 43:1-11, pis. 1-2.
Ross, R. 1935. A new genus and species of pigmy goose from the McKittrick
Pleistocene. Trans. San Diego Soc. Nat. Hist. 8(15) : 107-1 14, figs. 1-6.
Sellards, E. H. 1916. Human remains and associated fossils from the Pleistocene
of Florida. 8th Ann. Rept. Fla. Geol. Surv., p. 121-160, pis. 15-31, text figs.
1-15.

Shufeldt, R.W. 1892. A study of the fossil avifauna of the Equus beds of the
Oregon desert. J. Acad. Nat. Sci. Phila. 9(3) :389-425, pi. 15-17.

1913. Review of the fossil fauna of the desert region of Oregon, with a

description of additional material collected there. Bull. Amer. Mus. Nat. Hist.
32(6): 123-178, pis. 9-43.

1915. Fossil birds in the Marsh collection of Yale University. Trans. Con-
necticut Acad. Arts and Sci. 19:1-110, 15 pis.

1916. New extinct bird from South Carolina. Geol. Mag. n.s. 3(8): 343-

347, pi. 15.

Tordoff, H. B., and J. R. Macdonald. 1957. A new bird (family Cracidae) from
the early Oligocene of South Dakota. Auk 74:174-184, pi. 10, and text fig. 1.
Wetmore, A. 1923. Avian fossils from the Miocene and Pliocene of Nebraska.
Bull. Amer. Mus. Nat. Hist. 48( 12) :483-507, figs. 1-20.

1924. Fossil birds from southeastern Arizona. Proc. U.S. Nat. Mus.

64(5): 1-18, figs. 1-9.

1930. Fossil bird remains from the Temblor Formation near Bakersfield,

California. Proc. Calif. Acad. Sci. 19(8) : 85-93, 7 text figs.

1933. An Oligocene eagle from Wyoming. Smithsonian Misc. Coll. 87(19) :

1-9, figs. 1-19.

1940. Fossil bird remains from Tertiary deposits in the United States.

J. Morphol. 66:25-37, figs. 1-14.

Zullo, V. A., and L. G. Hertlein. 1970. Catalog of specimens in the type collec-
tion of the Department of Geology, California Academy of Sciences. Cephalo-
poda. Occas. Papers Calif. Acad. Sci. 82:1-130.

26

Contributions in Science

No. 228

Species Index

Names in brackets indicate latest taxonomic assignments (see Brodkorb, 1963a,

1964a, 1967, 1971).

abavus, Presbychen, p. 10
aldeni, Miortyx, p. 16
americanus, Neophrontops, p. 15
amplus, Gymnogyps, p. 13
angelensis, Pipilo, p. 22
anza, Agriocharis, p. 17
bessomi, Oxyura, p. 12
brea, Strix, p. 21
brevipes, Procrax, p. 16
bunkeri, Nettion, p. 10
calhouni, Puffinus, p. 4
californica, Diomedea, p. 4
californicus, Pavo [Parapavo], p. 17
californiensis, Mancalla, p. 19
californiensis, Protostrix, p. 21
cedrosensis, Mancalla, p. 20
clarki, Sarcorhamphus [Breagyps], p. 13
concinna, Gavia, p. 3
condoni, Anser [Cygnus paloregonus], p. 9
conklingi, Geococcyx, p. 21
connectens, Megapaloelodus, p. 8
conradi, Puffinus, p. 4
convexa, Pandanaris, p. 22
crassipes, Meleagris [Agriocharis], p. 17
daggetti, Morphnus [Wetmoregyps], p. 14
diatomicus, Puffinus, p. 4
dickeyi, Branta, p. 9
diegense, Pliolunda [Mancalla], p. 20
downsi, Brantadorna, p. 10
dubia, Cerorhinca, p. 19
errans, Neogyps, p. 15
felthami, Puffinus, p. 4
femoralis, Phalacrocorax, p. 7
fossilis, Bucephala, p. 1 1
fragilis, Geranoaetus [Buteogallus], p. 13
gilli, Palaeotetrix [Dendragapus], p. 16
goletensis, Phalacrocorax, p. 7
gracilenta, Anabernicula, p. 10
gracilis, Cathartornis, p. 12
grinnelli, Geranoaetus [Spizaetus], p. 14
grinnelli, Polyborus prelutosus
[Caracara], p. 16
hammeri, Fulmarus, p. 4
hesterna, Fulica, p. 18
howardae, Gavia, p. 3
humeralis, Sula, p. 6
inceptor, Puffinus, p. 4
incredibilis, Teratornis, p. 12
joaquininsis, Plotopterum, p. 7

kanakoffi, Puffinus, p. 4

kennelli, Phalacrocorax, p. 8

kernensis, Vultur [Sarcoramphus], p. 13

lagunensis, Praemancalla, p. 21

lawi, Chendytes, p. 11

lompocana, Sula [Morns], p. 7

lucasi, Pediocaetes [Dendragapus], p. 16

macdonaldi, Arikarornis, p. 15

maltha, Ciconia, p. 8

macropus, Graculus [Phalacrocorax], p. 7

matthewi, Olor [Cygnus paloregonus], p. 9

media, Miosula, p. 6

merriami, Teratornis, p. 12

mexicanus, Coragyps occidentalis, p. 12

milleri, Chendytes, p. 11

milled, Diomedea, p. 4

milleri, Mancalla, p. 20

minor, Cerorhinca, p. 19

minor, Fulcia [F. shufeldti], p. 18

minuscula, Branta [Anabernicula], p. 9

minutus, Phoenicopterus, p. 9

mioceanus, Palaeochenoides, p. 5

mitchelli, Puffinus, p. 5

molothroides, Pyelorhamphus, p. 22

natator, Epirallus, p. 18

nopcsai, Elopteryx, p. 5

opsigonus, Megapaloelodus, p. 8

oregonensis, Anabernicula, p. 10

oregonus, Larus, p. 18

orri, Osteodontornis, p. 5

paloccidentalis, Ardea

[Botaurus lentiginosus], p. 8
paloregonus, Cygnus, p. 9
pliocenus, Brachyramphus, p. 19
pliogryps, Aquila [Spizaetus], p. 13
pohli, Sula, p. 7

prelutosus, Polyborus [Caracara], p. 16

priscus, Asio, p. 21

priscus, Puffinus, p. 5

propinqua, Branta, p. 10

recentior, Miosula, p. 6

reyana, Moris [Morus reyanus], p. 6

robustus, Larus, p. 18

rossmoori, Aethia, p. 19

russelli, Eremochen, p. 10

shufeldti, Fulica, p. 18

shufeldti, Stercorarius, p. 1 8

sodalis, Aquila [Hypomorphnus], p. 13

sternbergi, Palaeoplancus, p. 15

1972

Type Specimens of Avian Fossils

27

stirtoni, Pseudodontornis, p. 6
stocki, Miohierax, p. 14
stocki, Phoenicopterus, p. 9
stocktoni, Sula [Palaeosula], p. 7
subparvus, Colymbus [Podiceps], p. 3
tedfordi, Puffinus, p. 5
turdirostris, Palaeoscinis, p. 21
typhoius, Buteo, p. 13
ulnulus, Alcodes, p. 19
vagabundus, Moris [Morus], p. 6

Accepted for publication March 24, 1972

vallecitoensis, Neophrontops, p. 15
victima, Palnumenius, p. 18
walleri, Titanis, p. 17
weillsi, ?Jabiru [ciconia maltha], p. 8
wetmorei, Mycteria, p. 8
willetti, Spizaetus, p. 14
willetti, Sula, p. 7
woodwardi, Morphnus, p. 14
yepomerae, Wasonaka, p. 10

so *]* J3

£zL

NUMBER 229
JUNE 12, 1972

A NEW SPECIES OF SWIFT OF
THE GENUS CYPSELOIDES FROM
NORTHEASTERN SOUTH AMERICA
(AYES: APODIDAE)

By Charles T. Collins

CONTRIBUTIONS IN SCI6NC6

Q

NATURAL HISTORY MUSEUM • LOS ANGELES COUNTY

CONTRIBUTIONS IN SCIENCE is a series of miscellaneous technical papers
in the fields of Biology, Geology and Anthropology, published at irregular intervals
by the Natural History Museum of Los Angeles County. Issues are numbered sep-
arately, and numbers run consecutively regardless of subject matter. Number 1 was
issued January 23, 1957. The series is available to scientific institutions and scien-
tists on an exchange basis. Copies may also be purchased at a nominal price. Inquiries
should be directed to Virginia D. Miller, Natural History Museum of Los Angeles
County, 900 Exposition Boulevard, Los Angeles, California 90007.

INSTRUCTIONS FOR AUTHORS

Manuscripts for CONTRIBUTIONS IN SCIENCE may be in any field of
Life or Earth Sciences. Acceptance of papers will be determined by the amount
and character of new information. Although priority will be given to manuscripts
by staff members, or to papers dealing largely with specimens in the collections of
the Museum, other technical papers will be considered. All manuscripts must be
recommended for consideration by the curator in charge of the proper section or
by the editorial board. Manuscripts must conform to those specifications listed
below and will be examined for suitability by an Editorial Committee including
review by competent specialists outside the Museum.

Authors proposing new taxa in a CONTRIBUTIONS IN SCIENCE must
indicate that the primary type has become the property of a scientific institution of
their choice and cited by name.

MANUSCRIPT FORM.— ( 1 ) The 1964 AIBS Style Manual for Biological
Journals is to be followed in preparation of copy. (2) Double space entire manu-
script. (3) Footnotes should be avoided if possible. Acknowledgments as footnotes
will not be accepted. (4) Place all tables on separate pages. (5) Figure legends and
unavoidable footnotes must be typed on separate sheets. Several of one kind may
be placed on a sheet. (6) An abstract must be included for all papers. This will be
published at the head of each paper. (7) A Spanish summary is required for all
manuscripts dealing with Latin American subjects. Summaries in other languages
are not required but are strongly recommended. Summaries will be published at
the end of the paper. (8) A diagnosis must accompany any newly proposed taxon.
(9) Submit two copies of manuscript.

ILLUSTRATIONS. — All illustrations, including maps and photographs, will
be referred to as figures. All illustrations should be of sufficient clarity and in the
proper proportions for reduction to CONTRIBUTIONS page size. Consult the 1964
AIBS Style Manual for Biological Journals in preparing illustration and legend copy
for style. Submit only illustrations made with permanent ink and glossy photo-
graphic prints of good contrast. Original illustrations and art work will be returned
after the manuscript has been published.

PROOF.— Authors will be sent galley proof which should be corrected and
returned promptly. Changes in the manuscript after galley proof will be billed to
the author. Unless otherwise requested, page proof also will be sent to the author.
One hundred copies of each paper will be given free to each author or divided
equally among multiple authors. Orders for additional copies must be sent to the
Editor at the time corrected galley proof is returned. Appropriate order forms will
be included with the galley proof.

Virginia D. Miller
Editor

A NEW SPECIES OF SWIFT OF THE GENUS CYPSELOIDES FROM
NORTHEASTERN SOUTH AMERICA (AVES: APODIDAE)1

By Charles T. Collins2

Abstract: A review of the Chestnut-collared Swift, Cypse-
loides rutilus, indicates that the labelled type specimen of
Hirundo rutila Vieillot, 1817 is from Trinidad, that Vieillot’s
original description agrees with the type, and that the distinctive
population from the Pantepui area of southern Venezuela and
neighboring Guyana and Brazil, long believed to the the same as
the Trinidad population and also called by that name by recent
authors, is deserving of separate species status. Although charac-
terized nearly 100 years ago, this species lacks a valid name, and
Cypseloides phelpsi, the Tepui Swift, is here proposed.

In the course of field studies of the Chestnut-collared Swift ( Cypseloides
rutilus) in Trinidad (Collins, 1968) I became increasingly aware that the
plumages of some individuals, particularly those of females and juveniles,
were sharply at odds with some published accounts. A subsequent review of
the molts and plumages of this species throughout its range (Collins, in
preparation) has also pointed out a particularly distinct population which,
although accurately characterized nearly 100 years ago, lacks a valid scientific
name. Correcting this situation entails first a review of the taxonomic history
of C. rutilus and this distinctive population.

The Chestnut-collared Swift ( Cypseloides rutilus) was first described by
Vieillot (1817) under the name of Hirunda rutila. The type specimen, stated
by Vieillot to be in the collections of the Museum National d’Histoire Natu-
relle, (Paris), is extant in the collections of that museum, mounted on a small
stand as was then the custom. No locality for the type was given by Vieillot
although “La Trinite” (i.e., Trinidad) is written on the underside of the base
of the stand. This omission is not surprising as it is well known that Vieillot
often described new species from mounted specimens he did not handle but
only observed in locked exhibit cases in the Paris museum. In any event
Vieillot’s description agrees with the specimen. Also appearing on the under-
side of the stand are the determinations “Chaetura rutila Vieillot” and
“ Hirundo robini Lesson,” the latter being a long accepted junior synonym
published in 1831 with type locality given as “Pile de la Trinite.” Later authors

Editorial Committee for this Contribution
Eugene Eisenmann
Herbert Friedmann
Kenneth E. Stager

2Research Associate in Ornithology, Natural History Museum of Los Angeles
County; and Department of Biology, California State College, Long Beach, Cali-
fornia 90801.

1

2

Contributions in Science

No. 229

(Sclater, 1855; Salvin and Sclater, 1860; Orton, 1871) state that this species
was collected by M. Robin in Trinidad and that his specimens form the types
of Vieillot’s and Lesson’s descriptions. I have been able to find only the single
specimen. There is no evidence that there were ever more and I suggest that
both descriptions were based on the same specimen, as indicated by the labels
on the stand. As indicated below there is no reason to doubt that the specimen
described by Vieillot is that marked as type in the Paris museum and labeled
as being from Trinidad, and this should be designated as the type locality for
Hirundo rutila Vieillot and H. robini Lesson in future studies.

In the next 100 years Chestnut-collared Swifts were collected in most
parts of their present known range: in mountainous country from Mexico to
Bolivia. Additional taxa were described from Colombia ( Chaetura brunni-
torques Lafresnaye, 1844 = Cypseloides rutilus brunnitorqu.es) ; and from
Mexico ( Cypselus brunneitorques (sic) griseifrons Nelson, 1900 = Cypse-
loides rutilus griseifrons ; Chaetura nubicola Brodkorb, 1938 = Cypseloides
rutilus nubicola ). A full review of these taxa and a yet to be described sub-
species from South America will be presented later (Collins, in preparation).

During this period authors have not been in full agreement as to whether
brunnitorques and rutilus were races of a single species, C. rutilus, or alter-
natively, separate species, with griseifrons and nubicola being races of
C. brunnitorques. Species limits will be discussed in detail below. Although
several fairly recent authors, mostly following Peters (1940), have included
these birds in the genus Chaetura, as was also done by some very early work-
ers, the current consensus favors inclusion in Cypseloides. This is based both
on various aspects of the reproductive biology (Lack, 1956; Snow, 1962;
Collins, 1968) but also on more traditional morphological characters (Zim-
mer, 1953; Eisenmann and Lehmann, 1962).

In the last century two specimens of an allied but strikingly different
swift, one of which is presently located in the collections of the British
Museum, were collected by H. Whitely in the Merume Mountains of British
Guiana (Guyana). As described by Salvin and Godman (1882:82), these
specimens differed from those of other areas in having a “brighter” chestnut
collar and a tail “much longer and distinctly forked.” Also, the chestnut of
the collar included the chin, an area normally brown in specimens from all
other populations. These are in fact some of the salient characteristics of this
population. However, the following statement made by Salvin and Godman
with regard to the correct name applicable to these specimens was evidently
made without examining Vieillot’s type and in disregard of a salient aspect of
his description: “There can be little doubt that the species described as
Hirundo rutila by Vieillot was the Guiana bird, though the origin of the
specimens was unknown.”

This view is contradicted by Vieillot’s original description of Hirundo
rutila in which, among other things, the tail is stated to be square (not forked),
and the chin is not chestnut colored. Salvin and Godman further suggested

1972

A New Species of Swift from South America

3

that the name Hirundo robini be applied to these Guianan birds as well and
that the island of Trinidad be included in the range. The erroneous assumption
was made that the two populations were the same. The designation of
“Guiana” as the type locality for H. rutila by Peters (1940) following Salvin
and Godman, is not supportable on the basis of the known facts. Peters did
correctly include Trinidad in the range of this form, not realizing that two
very different forms were included under one designation.

I have examined most of the available specimens of the Chestnut-collared
Swifts from all portions of their range and they are in agreement with Vieillot’s
original description of H. rutila and the type specimen, which I have also
examined. It is important to note that in all populations the tail is relatively
short, essentially square and unforked. In worn plumages the rectrices may
become abraded thus exposing the terminal portion of the shafts of some
feathers. This gives them the superficial appearance of having the bare ter-
minal “spines” typical of species of Chaetura. The specimen described by
Vieillot had these characteristics, for he stated “la queue carree; les deux
pennes intermediares terminees en pointe; les autres arrondies a leur extremite,”
which I translate as “the tail square; the two middle feathers ending in a point;
the others rounded at their ends.” Personal observations made on numerous
living birds netted in the field confirm that these characteristics are also true
of the Chestnut-collared Swifts presently breeding on the island of Trinidad.
Surprisingly there are but two museum specimens of these swifts from Trini-
dad, and only one is of an adult. I have examined both and they are of the
form described by Vieillot.

The correct view that Cypseloides rutilus (Vieillot) is applicable to the
birds inhabiting the island of Trinidad has been uniformly accepted by all
authors considering the avifauna of this island from Leotaud (1866) to the
present day. However, the erroneous conclusion of Salvin and Godman
(1882) that Guianan birds were the same was repeated by Salvin (1885), and
unfortunately was uncritically followed by Peters and nearly all later authors.
This gave rise to the view that the name Cypseloides rutilus was applicable
not only to the Trinidad form (which is square-tailed), but to the distinctive,
fork-tailed birds now known from many specimens from the tabletop moun-
tains (tepuis) south of the Orinoco River in Venezuela and the immediate
adjacent parts of Guyana and Brazil (“Pantepui Area” of Mayr and Phelps,
1967). This situation was abetted by a near absence of specimens of these
swifts from the mountains of northern Venezuela, thus giving the impression
that there existed a large discontinuity in the range of these swifts and that the
nearest continental population to Trinidad was in fact that inhabiting Pantepui.
With the collection of specimens of C. rutilus from various localities in
northern Venezuela (Phelps and Phelps, 1958), and the filling of this seeming
discontinuity in their range, it is now obvious that the zoogeographical affini-
ties of the population in Trinidad ( Cypseloides rutilus sensu stricto ) are with
northern Venezuela and Colombia (so-called brunnitorques) . There is much

4

Contributions in Science

No. 229

less morphological resemblance and less close relationship with the distinctive
population inhabiting Pantepui. In fact, rutilus and brunnitorques are so
similar as to be doubtfully distinct even as subspecies, and after further study
the latter may prove to be synonymous with rutilus which has priority. The
Pantepui swifts, characterized by Salvin and Godman (1882) form a dis-
tinctive allopatric population for which I now propose the name:

Cypseloides phelpsi, new species

TYPE: Adult male, AMNH 324213, original expedition number 1594;
collected 14 February 1938, on Cerro Auyan-tepui, Bolivar, Venezuela at an
elevation of 1100 meters by the Phelps Venezuela Expedition.

DIAGNOSIS: Adults of C. phelpsi are readily separable from those of all
populations of C. rutilus (whatever the subspecies) in having a longer “softer”
(less stiffened) and deeply forked tail lacking the stiffened, and sometimes
bare-tipped shafts and square tail of C. rutilus and in longer wings. Moreover,
in color they also differ from all populations of C. rutilus in 1) having the
plumage more nearly black rather than a blackish brown, 2) the collar a more
orange-chestnut tone rather than a deep red-brown or chestnut-brown, and
3 ) in having the coloration of the collar extend upward over all of the chin or
interramal area. The extent of this coloration is the same in both sexes
although the breast is a bit paler and mixed with brown in some females. The
white supraocular streak is present in nearly all individuals. In C. rutilus only
exceptional females have the full male coloration; most females have no
chestnut collar, or only a partial one confined to the nape and part of the
sides of the neck. In both C. phelpsi and C. rutilus there is a tendency for
males to be larger than females in most linear measurements, although even
the smallest females of C. phelpsi are generally larger than the largest males
of C. rutilus. Table 1 presents measurements of the available specimens of
C. phelpsi (both from Venezuela and Guyana) and, for comparison, a series
of C. rutilus from the mountainous areas of northern Venezuela in the states
of Tachira, Merida, Barinas, Yaracuy, Carabobo, Aragua, Distrito Federal,
Miranda and Sucre.

As mentioned earlier, specimens from all parts of the range of C. rutilus
have been examined in this study, although only measurements from this one
nearby part of the range are presented here. The darkness of the body and
flight feathers, the more orange color and extent of the collar, the length of
wing and tail, and depth of forking of the tail, individually as well as collec-
tively, serve to separate C. phelpsi from this or any other population of
C. rutilus throughout its range. The degree of whiteness of the supraocular
stripe in C. phelpsi is approached in one population of C. rutilus in Middle
America ( nubicola ). As also usually (but not invariably) true in C. rutilus,
the outermost (tenth) primary of C. phelpsi is shorter than the ninth (see tip
measurement, Table 1).

1972

A New Species of Swift from South America

5

Table 1

Measurements8 of Cypseloides phelpsi and
Cypseloides rutilus from Venezuela

phelpsi
Males
(N = 12)c

phelpsi
Females
(N= 18)

rutilus
Males
(N = 22)

rutilus
Females
(N = 20)

Wing

(Flattened)

136.92 ±0.61
(133-140.5)

133.92 ±0.64
(129.5-138)

122.50 ±0.89
(116-130.5)

119.15 ±0.71
(112-124.5)

Wing Tipb

5.15 ±0.42
(3.0-7. 5)

4.86 ±0.28
(2.5-7. 5)

4.98 ±0.53
(2.5-10.0)

4.31 ±0.18
(2. 5-5. 5)

Tail

61.31 ±0.81
(56.5-66)

58.87 ±0.36
(56.5-61.5)

44.84 ±0.58
(39.5—48.5)

42.68 ±0.49
(37.5-47.0)

Depth of
Tail Fork

9.61 ±0.45
(7.0-11.5)

9.71 ±0.48
(5.5-13.0)

2.79 ±0.62
(1. 0-3.0)

1.36 ±0.29
(0.0-4. 5)

Culmen

(from nostril)

4.18 ±0.06
(4.0-4.5)

4.21 ±0.06
(3.7— 4.5)

4.17 ±0.09
(3.7-4. 5)

4.16 ±0.05
(3. 7-4. 5)

Tarsus

13.76 ±0.13
(12.7-14.5)

13.60 ±0.11
(12.7-14.3)

12.33 ±0.11
(11.5-13.0)

12.03 ±0.11
(11.3-13.0)

a. All measurements in millimeters; presented are: Mean± standard error and
(range).

b. Difference in length of ninth and tenth primaries (ninth longest).

c. Does not include extralimital specimen from Aragua: wing, 139; wing tip, ?
(primary 10 not full length) ; tail, 59.2; depth of fork, 9.8; culmen, 4.0; tarsus,

14.0.

DESCRIPTION OF TYPE: Dark sooty black all over except for pro-
nounced orange-chestnut collar including nape, upper breast, throat, chin and
sides of head up to level of eyes; light white streak above eyes on edge of dark
crown. Tail deeply forked; shafts of rectrices not markedly stiffened nor pro-
jecting beyond vane. Soft parts (on label) : iris brown, bill black, feet purplish
gray. Wing (flattened) 136 mm, tail 61.5 mm; culmen (from nostril) 4 mm;
tarsus 14.5 mm; depth of tail fork 8.5 mm; gonads not fully enlarged; no
appreciable molt but not in fresh plumage.

RANGE: Pantepui area of southeastern Venezuela, northwestern
Guyana, and probably (no specimen) extreme northeastern portion of Terri-
torio Federal de Roraima, Brazil. A single extralimital specimen has been
taken at Rancho Grande, Aragua, in northern Venezuela.

SPECIFIC STATUS: C. phelpsi is unquestionably a distinctive popula-
tion, but, it may be argued, no more so than numerous insular or otherwise
isolated populations of other birds entitled to only subspecific rank. It should
be remembered, however, that swifts are exceedingly mobile animals and
that the geographic distances which restrict gene flow between populations of
many bird species may be encompassed by the daily foraging flights of swifts.
Thus it is unlikely that in itself the distance between Pantepui and the nearest

6

Contributions in Science

No. 229

breeding populations of C. rutilus in northern Venezuela (900 ± kms) is
enough of a barrier to gene flow to justify considering the striking differences
of C. phelpsi as simply those of a geographically isolated but potentially
interbreeding population. The Mexican and Bolivian populations of C. rutilus
are more like those of Trinidad and northern Venezuela than is the com-
paratively nearby Pantepui population of C. phelpsi. As has also been pointed
out by Orr (1963) and Brooke (1971), good species of swifts frequently
show little divergence in appearance so that seemingly minor morphological
difference may be of greater importance in delimiting species than in other
avian taxa. Thus the striking difference in wing and tail length, degree of
forking of the tail, and decreased sexual dimorphism in plumage of C. phelpsi
seem especially significant in appraising specific limits in this case. Two further
bits of evidence are available. Firstly, if the Pantepui area is as isolated for
swifts as it is for the other less mobile species, we should expect to find
similar degrees of difference in other swift species living there. Such is not
the case! Aeronautes montivagus and Chaetura chapmani show little or no
geographic variation over this part of their ranges. A second bit of evidence
that C. phelpsi is not sedentary is the existence of a single specimen collected
at the Rancho Grande Biological Station in Aragua on 16 February 1960.
This specimen, now housed in the collection at that station, is typical in every
way of the Pantepui specimens of C. phelpsi. Rancho Grande is well within
the breeding range of C. rutilus, which has also been collected there (Beebe,
1949; Collins, in preparation). This indicates that at least an occasional indi-
vidual of C. phelpsi may occur in the range of C. rutilus and that the appre-
ciable morphological differences between these birds are maintained despite
this possible sympatry and potential for genetic interchange. For these rea-
sons I feel that tentatively full specific status is warranted for Cypseloides
phelpsi. This is essentially a reversion, although with new nomenclature, to
the treatment prevailing before Peters (1940).

REMARKS: There is no information available on the ecology, feeding
habits, or body weight of C. phelpsi. It was observed flying in large flocks
around Cerro Auyan-tepui in the non-breeding season by Gilliard (1941).
Although Mayr and Phelps (1967:297) include the Tepui Swift in a list of
“cliff dwellers,” this, although probably true, is still a supposition, for its
nesting and roosting habits are presently unreported. In all likelihood, it will
show the same affinities for nest and roosting sites in damp, dark areas with
high relief, near or behind waterfalls, exhibited by other Cypseloides swifts
including C. rutilus (Snow, 1962; Collins, 1968). Nest sites of C. phelpsi
should be looked for in the vicinity of the numerous waterfalls coming off
the tepuis.

Two specimens showing early stages of molt of the wing feathers, typical
of the end of the breeding season, were taken on 26 July. This probably indi-
cates a late “spring”- early “summer” breeding season (in the northern hemi-
sphere sense) closely tied to the onset of the rainy season in this area. Only

1972

A New Species of Swift from South America

7

one of a large series collected in February showed even slightly enlarged
gonads (Gilliard, 1941 ). Two specimens show from three to six white feathers
in the central breast region at the lower border of the collar. These specimens,
both of female (AMNH 323327 and 324266), were collected on Cerro
Auyan-tepui on 14 February and 13 March 1938. Such cases of partial albin-
ism have been recorded for several other neotropical swifts including C.
rutilus (Eisenmann and Lehmann, 1962; Collins, 1967).

The name Cypseloides phelpsi, based on information provided by me,
has appeared as a nomen nudum in two recent faunal lists but without any
diagnosis or description (Brooke, 1970a, 1970b).

Since most of the possible vernacular names incorporating the color of
the collar have been used in reference to C. rutilus, Tepui Swift would seem
the most appropriate English name for Cypseloides phelpsi in recognition of
its range in Pantepui.

ETYMOLOGY: It is my pleasure to name this swift after William H.
Phelps, Jr., who, by so ably continuing the efforts devoted by his father, the
late William H. Phelps, to the study and preservation of the avifauna of
Venezuela and the Pantepui area in particular, has contributed so much to
our ornithological knowledge of these areas.

Specimens Examined

Cypseloides phelpsi

Venezuela, Bolivar, Mt. Auyan-tepue: 9 males, 14 females (AMNH, R.G.)
Cerro Duida: 1 female (AMNH)

GranSabana: 1 male, 1 female (P.)

Cerro Serrania: 1 male, 1 female (P.)

Territory Amazonas, Cerro Yapacana: 1 female (R.G.)

Aragua, Rancho Grande: 1 male (R.G.)

British Guiana (Guyana): Merume Mountains: 1 male (B.M.)

Cypseloides rutilus

Over 250 specimens from all parts of the range of this species have been
examined in this study including a sample of 44 from northern Venezuela
(localities listed in text). A complete analysis of this species will be presented
later (Collins, in preparation).

(AMNH = American Museum of Natural History, New York; P. = Phelps Orni-
thological Collection, Caracas; R.G. = Estacion Biologica de Rancho Grande,
Aragua; B.M. = British Museum, Tring.)

Acknowledgments

This study of Cypseloides rutilus and C. phelpsi, part of a wider study
of the biology of Neotropical swifts, has been generously supported by re-
search awards for field studies in Trinidad and Venezuela and a post-doctoral
fellowship from the Frank M. Chapman Memorial Fund of the American
Museum of Natural History. Without this support this work would not have
been possible. I am grateful to the curators of the many museum collections

8

Contributions in Science

No. 229

from which I borrowed specimens for this study, and patricularly the authori-
ties of the Museum National d’Histoire Naturelle for allowing me to examine
Vieillot’s type of Hirundo rutila. I am also most grateful to E. Eisenmann
and R. K. Brooke for their most helpful comments which improved an earlier
draft of this paper.

Resumen

Una revision del vencejo de collar castano, Cypseloides rutilus, ha indi-
cado que el especimen tipo ( Hirundo rutila Vieillot, 1817) es de Trinidad,
y que la poblacion distintiva del area de Pantepui del sur de Venezuela, la
vecina Guayana y Brasil, desde hace mucho asociada con este nombre,
merece ser separada en categoria de especie. Aunque caracterizada hace cerca
de 100 anos, esta especie carece de nombre valido y Cypseloides phelpsi es
propuesto aqui para el vencejo tepuiano.

Literature Cited

Beebe, W. 1949. The swifts of Rancho Grande, North-Central Venezuela, with
special reference to migration. Zoologica 34: 53-62.

Brodkorb, P. 1938. New birds from the district of Soconusco, Chiapas. Oc. Papers.
Mus. Zool., Univ. Mich. 369: 1-7.

Brooke, R. K. 1970a. Zoogeography of Swifts. Ostrich, Supplement 8: 47-54.

1970b. Taxonomic and evolutionary notes on the subfamilies, tribes,

genera and subgenera of the swifts. Durban Mus. Novitates 9: 13-24.

1971. Geographical variation in the Little Swift Apus affinis (Aves:

Apodidae). Durban Mus. Novitates 9: 93-103.

Collins, C. T. 1967. Partial albinism in the Chestnut-collared Swift in Trinidad.
Bull. Brit. Orn. Club 87: 122-123.

1968. The comparative biology of two species of swifts in Trinidad, West

Indies. Bull. Fla. St. Mus. 11: 257-320.

Eisenmann, E., and F. C. Lehmann V. 1962. A new species of swift of the genus
Cypseloides from Colombia. Amer. Mus. Novitates 2117: 1-16.

Gilliard, E. T. 1941. The birds of Mt. Auyan-tepui, Venezuela. Bull. Amer. Mus.
Nat. Hist. 77: 439-508.

Lack, D. 1956. A review of the genera and nesting habits of swifts. Auk 73: 1-32.
Lafresnaye, M. de 1844. Nouvelles especes d’oiseaux de Colombie. Revue Zoo-
logique 1844: 80-??

Leotaud, A. 1866. Oiseaux de Lisle de la Trinidad. Chronicle Publishing Office,
Port of Spain. 560 p.

Lesson, R. P. 1831. (type description of Hirundo robini ) Traite d’Ornithologie,
Paris.

Mayr, E., and W. H. Phelps, Jr. 1967. The origin of the bird fauna of the south
Venezuelan highlands. Bull. Am. Mus. Nat. Hist. 136: 269-328.

Nelson, E. W. 1900. Descriptions of thirty new North American birds in the
Biological Survey collection. Auk 17: 253-270.

Orr, R. T. 1963. Comments on the classification of swifts of the subfamily
Chaeturinae. Proc. XIII Internat. Ornith. Cong. 1: 126-134.

Orton, J. 1871. Notes on some birds in the Museum of Vassar College. Amer.
Nat. 4: 711-717.

Peters, J. L. 1940. Check-list of Birds of the World. Vol. 4. Harvard Univ. Press,
Cambridge.

1972

A New Species of Swift from South America

9

Phelps, W. H., and W. H. Phelps, Jr. 1958. Lista de las aves de Venezuela con
su distribucion. Tomo 2, Parte 1, No Passeriformes.

Salvin, O. 1885. A list of the birds obtained by Mr. Henry Whitely in British
Guiana. Ibis 3 (5th Ser.): 418-439.

Salvin, O., and F. D. Godman. 1882. Notes on birds from British Guiana. Ibis 6
(4th Ser.): 76-84.

Salvin, O., and P. L. Sclater. 1860. Contributions to the ornithology of Guate-
mala. Ibis 2: 28-45.

Sclater, P. L. 1855. On the birds received in collections from Santa Fe de Bogota.
Proc. Zool. Soc. Lond. 1855: 131-166.

Snow, D. W. 1962. Notes on the biology of some Trinidad swifts. Zoologica 47:
129-139.

Vieillot, L. J. P. 1817. Nouveau Dictionnaire D’Historie Naturelle, Vol. 14.

Zimmer, J. T. 1953. Studies of Peruvian birds. No. 64. The Swifts: Family
Apodidae. Am. Mus. Novitates 1609: 1-20.

Accepted for publication April 17, 1972

7, ys

C L

NUMBER 230
JUNE 23, 1972

HYPSOCEPHALUS ATLANTICUS, A NEW
GENUS AND SPECIES OF LUTJANID
FISH FROM MARINE EOCENE
LIMESTONES OF NORTHERN FLORIDA

By Camm Swift and Brooks Ell wood

CONTRIBUTIONS IN SCIENCE

iff zo

U;

Piiil

NATURAL HISTORY MUSEUM • LOS ANGELES COUNTY

CONTRIBUTIONS IN SCIENCE is a series of miscellaneous technical papers
in the fields of Biology, Geology and Anthropology, published at irregular intervals
by the Natural History Museum of Los Angeles County. Issues are numbered sep-
arately, and numbers run consecutively regardless of subject matter. Number 1 was
issued January 23, 1957. The series is available to scientific institutions and scien-
tists on an exchange basis. Copies may also be purchased at a nominal price. Inquiries
should be directed to Virginia D. Miller, Natural History Museum of Los Angeles
County, 900 Exposition Boulevard, Los Angeles, California 90007.

INSTRUCTIONS FOR AUTHORS

Manuscripts for CONTRIBUTIONS IN SCIENCE may be in any field of
Life or Earth Sciences. Acceptance of papers will be determined by the amount
and character of new information. Although priority will be given to manuscripts
by staff members, or to papers dealing largely with specimens in the collections of
the Museum, other technical papers will be considered. All manuscripts must be
recommended for consideration by the curator in charge of the proper section or
by the editorial board. Manuscripts must conform to those specifications listed
below and will be examined for suitability by an Editorial Committee including
review by competent specialists outside the Museum.

Authors proposing new taxa in a CONTRIBUTIONS IN SCIENCE must
indicate that the primary type has become the property of a scientific institution of
their choice and cited by name.

MANUSCRIPT FORM.— (1 ) The 1964 AIBS Style Manual for Biological
Journals is to be followed in preparation of copy. (2) Double space entire manu-
script. (3) Footnotes should be avoided if possible. Acknowledgments as footnotes
will not be accepted. (4) Place all tables on separate pages. (5) Figure legends and
unavoidable footnotes must be typed on separate sheets. Several of one kind may
be placed on a sheet. (6) An abstract must be included for all papers. This will be
published at the head of each paper. (7) A Spanish summary is required for all
manuscripts dealing with Latin American subjects. Summaries in other languages
are not required but are strongly recommended. Summaries will be published at
the end of the paper. (8) A diagnosis must accompany any newly proposed taxon.
(9) Submit two copies of manuscript.

ILLUSTRATIONS.- — All illustrations, including maps and photographs, will
be referred to as figures. All illustrations should be of sufficient clarity and in the
proper proportions for reduction to CONTRIBUTIONS page size. Consult the 1964
AIBS Style Manual for Biological Journals in preparing illustration and legend copy
for style. Submit only illustrations made with permanent ink and glossy photo-
graphic prints of good contrast. Original illustrations and art work will be returned
after the manuscript has been published.

PROOF.— Authors will be sent galley proof which should be corrected and
returned promptly. Changes in the manuscript after galley proof will be billed to
the author. Unless otherwise requested, page proof also will be sent to the author.
One hundred copies of each paper will be given free to each author or divided
equally among multiple authors. Orders for additional copies must be sent to the
Editor at the time corrected galley proof is returned. Appropriate order forms will
be included with the galley proof.

Virginia D. Miller
Editor

HYPSOCEPHALUS ATLANTICUS, A NEW GENUS AND
SPECIES OF LUTJANID FISH FROM
MARINE EOCENE LIMESTONES OF
NORTHERN FLORIDA1

By Camm Swift2 and Brooks Ellwood3

Abstract: A single neurocranium (and a few other bone
fragments) representing an undescribed genus and species of
hoplopagrine lutjanid was discovered in solution caverns in
Eocene limestones in Jackson County, Florida. It is quite dis-
tinct from the only living member of this subfamily, Hoplopa-
grus guntheri, known from southern Baja California to Panama
in the eastern Pacific Ocean. The Hoplopagrinae are distinct
from other lutjanids in possessing: 1) strong, blunt, conical
teeth on the premaxillary, dentary, vomer and palatine; 2) a
largely vertical posterior face on the basioccipital; 3) exoccipital
condylar surfaces which fail to meet in the midline; 4) relatively
small otic capsules; and 5) a strong, globular, ventral swelling
near the posterior end of the parasphenoid, apparently serving
as a brace for the upper pharyngeals.

In Eocene times when the sea was deeper and warmer, the
hoplopagrine lutjanids were present near the northern Gulf of
Mexico of today. For some reason the group disappeared on the
Atlantic side but persisted in the eastern Pacific Ocean. This
subfamily is not known outside the New World.

The snappers, family Lutjanidae, are common, worldwide fishes in
tropical and subtropical marine shore waters. A few species enter estuaries,
and several others are little known species occurring about hard substrate in
deep water. Despite this recent abundance, snappers are scarce in the fos-
sil record, and only two records could be found for fossil snappers in North
America (Gregory, 1930; Jordan and Gilbert, 1919). Elsewhere in the world
three genera ( Caesio , Lednevia, Lutjanus ) have been recorded from Eocene
and Miocene deposits of Europe, and Lutjanus has been noted from the
Miocene of Australia (Romer, 1966). Summary works on fossil fishes by
Smith-Woodward (1901), Casier (1966), Danil’chenko (1967), and Lehman
(1966) mention no lutjanid genera. Six other doubtful fossil records for the
family are based on otoliths (Weiler, 1968), five from the London Clay and
one of Lutjanus from Borneo.

1Editorial Committee for this Contribution
William A. Gosline
Robert J. Lavenberg
Stanley H. Weitzman

2Associate Curator of Fishes, Natural History Museum of Los Angeles County,
900 Exposition Boulevard, Los Angeles, California 90007.

3Graduate School of Oceanography, University of Rhode Island, Kingston, Rhode
Island 02881.

1

Contributions in Science No. 230

1972

New Genus and Species of Lutjanid Fish

3

The lutjanid fossils known from North America come from the Miocene
of California, Lutianus hagari Jordan and Gilbert, 1919, and the Oligocene
of Florida, Lutjanus avus Gregory, 1930. The first known fossil of a hop-
lopagrine snapper (described below) comes from within a few miles of the
site of collection of Lutjanus avus in northwest Florida (Figure 1). The Hop-
lopagrinae are otherwise known only by the sole living species, Hoplopagrus
guntheri Gill (1862a) which ranges from Abreojos (UCLA S-392) and Mag-
delena (LACM 32086-3) Bays on the west coast of Baja California and
Guaymas (CAS IU 7749) in the Gulf of California south to Panama (Wal-
ford, 1937). It occurs about reefs from shore to “deep, cold water near the
Pearl Islands” (Walford, 1937) in the Gulf of Panama.

Previous authors have compared the hoplopagrines with other lutjanids
and with the sparids when searching for the affinities of the subfamily (Gill,
1862a, 1862b; Jordan and Evermann, 1898; Regan, 1913). All agree that
the hoplopagrines resemble lutjanids more than sparids, and this seems to be
true based on our comparison of the skeletons of most of the North American
genera of both families. The new fossil has been compared primarily with
North American lutjanids, but osteological resemblances to sparids and pom-
adasyids have been recorded. References to characters of those families are
based on examination of materials listed below. Several old world genera
of both families have not been examined, and this should be considered in
assessing the comparisons.

Materials and Methods

The comparison below of the fossil with recent species of lutjanids,
pomadasyids and sparids are based on the following specimens. Neuro-
cranium length was measured from the anterior end of the vomer to the
posterior end of the basioccipital. Abbreviations are: California Academy of
Sciences (CAS), University of California at Los Angeles (UCLA), and
Natural History Museum of Los Angeles County (LACM).

Lutjanidae

Hoplopagrus guntheri Gill: CAS 14158, 460 mm SL, Mexico, Cerralvo
Is., 4 March 1945; LACM-VP-F422, Mexico, Gulf of California, Cabo Lobos,
14 Feb. 1970, S. P. Applegate; UCLA S-392, about 410 mm SL, Mexico,
Baja California, Abreojos Bay, May, 1954; LACM 31774-1, 208 mm SL,
Mexico, Gulf of California, Baja California, just S Pta. Arena, R/V Searcher
Sta. 44, 2 Feb. 1971.

Lutjanus apodus (Walbaum) : LACM 31732-1, SL unknown (neuro-
cranium 48.9 mm long), Bahamas, San Salvador Island, Graham Harbor, 10
June 1966, E. S. Wing.

Lutjanus campechanus (Poey): LACM 31737-1, 523 mm SL, Florida,
Gulf of Mexico, Middle Grounds, Sept. 1969, P. McCaffrey.

Lutjanus griseus (Linnaeus): LACM 31735-1, 393 mm SL, Florida,
Monroe Co., Tortugas, 17 or 18 Dec. 1968, C. Combs, H. Austin, H. Mattraw;

4

Contributions in Science

No. 230

LACM 31852-1, 460 mm SL, Broward Co., off Hollywood Beach, 18 March
1969, H. Yaffa; LACM 31736-1, 215 mm SL, Florida, Okaloosa Co., mouth
Choctawhatchee Bay, 12 Oct. 1968, R. Hastings.

Lutjanus synagris (Linnaeus) : LACM 31731-1, ca. 335 mm SL, Florida,
Volusia Co., vie. Daytona Beach, summer 1966, M. Gomon.

Ocyurus chrysurus (Bloch) : LACM 31853-1, 461 mm SL, Florida, Gulf
of Mexico, Middle Grounds. 28° 25' N, 84° 18' W, 10 Nov. 1969, J. Bishop,
H. Austin; LACM 31734-1, 245 mm SL, Panama, Atlantic Ocean, San Bias,
Holandes Cay, 27 Sept. 1970, J. E. McCosker.

Pristipomoides aquilonaris (Goode and Bean): LACM 31730-1, ca. 110
mm SL, Oregon Station 10892, Gulf of Mexico.

Rhomboplites ciurorubens (Cuvier): LACM 31738-1, 190 mm SL,
Florida, Okaloosa Co., vie. Destin, 2 April 1966, C. Swift; LACM 31733-2,
219 mm SL, and LACM 31733-1, 165 mm SL, Gulf of Mexico, 23 mi SSE
of Pensacola, 29 June 1969, S. Bortone.

Pomadasyidae

Anisotremus davidsoni (Steindachner) : LACM 32587-1, 319 mm SL,
California, Orange Co., Huntington Beach, 8 Aug. 1970, C. Swift, et al.;
LACM 32588-1, 298 mm SL, California, Orange Co., Huntington Beach,
5 July 1970, J. Fitch.

Brachydeutereus corvinaeformis Steidachner: LACM 32585-1, neuro-
cranium 38 mm long, Panama, Atlantic, Bahia Limon, 1970, J. E. McCosker.

Haemulon aurolineatum Cuvier: LACM 32584-2, 129 mm SL, Florida,
Franklin Co., 15 mi S Alligator Harbor, 10 May 1969, S. Bortone; LACM
31849-6, 151 mm SL, Florida, Gulf of Mexico, Middle Grounds, 12, 13 June
1969, C. Swift and party.

Haemulon plumieri (Lacepede) : LACM 32584-1, neurocranium 72 mm
long, Florida, Franklin Co., ca. 15 mi S Alligator Harbor, 10 May 1969,
S. Bortone; LACM 31849-5, 383 mm SL, Florida, Gulf of Mexico, Middle
Grounds, 12, 13 June 1969, C. Swift and party.

Haemulon sciurus (Shaw) : LACM 32586-1, neurocranium 52 mm long,
Florida, Monroe Co., Vaca Key, June, 1964, K. Ainslie and party.

Orthopristis chrysoptera (Linnaeus): LACM 31848-3, 177 mm SL,
LACM 31848-4, 170 mm SL, Florida, Franklin Co., mouth Alligator Harbor,
4 July 1968, C. Swift and party; LACM 32583-1, 188 mm SL, LACM
32583-2, 179 mm SL, Florida, Franklin Co., St. George Island, 11 June 1968,
F. W. Vockell.

Sparidae (all from Florida)

Archosargus probatocephalus (Walbaum): LACM 31591-3, 270 mm SL,
Okaloosa Co., mouth Choctawhatchee Bay, 11 Sept. 1969, C. Swift and
party; LACM 31850-1, SL unknown (neurocranium 32.9 mm long), Santa
Rosa Co., Santa Rosa Sound at Gulf Breeze, 15 July 1966, R. W. Hastings.

Calamus nodosus Randall and Caldwell: LACM 31849-2, 242 mm SL,

1972

New Genus and Species of Lutjanid Fish

5

Gulf of Mexico, Middle Grounds, 28° 3 O' N, 84° 15' W, 12, 13 June 1969,
C. Swift and party.

Calamus arctifrons Goode and Bean: LACM 31537-36, 137 mm SL,
and LACM 31537-38, 159 mm SL, Wakulla Co., 5.5 mi WSW St. Marks
Light, 6 Sept. 1969, C. Swift and party.

Diplodus holbrooki (Bean): LACM 31848-2, 123 mm SL, Franklin
Co., mouth Alligator Harbor, 4 July 1968, C. Swift and party; LACM
31851-1, 159 mm SL, Franklin Co., off mouth Ochlockonee River, 4 Oct.
1969, J. Wiese, R. Lazor.

Lagodon rhomboides (Linnaeus): LACM 31845-1, 143 mm SL, Bay
Co., St. Andrews Bay, 14 Oct. 1967, C. Swift and party, LACM 31848-1,
151 mm SL, Franklin Co., mouth Alligator Harbor, 4 July 1968, C. Swift
and party.

Pagrus sedecim Ginsburg: LACM 31738-2, head only (neurocranium
51.8 mm long), Okaloosa Co. ca. 15 mi S Destin, 2 April 1966, R. W. Yerger
and class; LACM 31849-1, 292 mm SL, Gulf of Mexico, Middle Grounds,
28° 3 O' N, 84° 15' W, 12, 13 June 1969, C. Swift and party.

Anatomical Abbreviations

: following abbreviations are used in

figures 2 to 5 :

bo

basioccipital

para

parasphenoid

bs

basisphenoid

pro

prootic

epo

epiotic

pto

pterotic

exo

exoccipital

seth

supraethmoid

fr

frontal

soc

supraoccipital

int

intercalar

spho

sphenotic

leth

lateral ethmoid

VO

vomer

pa

parietal

Hypsocephalus , new genus

Diagnosis: A hoplopagrine lutjanid distinguished from the living and
only other known genus of the subfamily, Hoplopagrus, by: 1) a skull which
is high and deep rather than elongate; 2) supra- and lateral temporal fossae
shallow and flattened rather than deeply excavated; 3) a supraoccipital crest
extending anterior to a vertical through the center of the bony orbit rather
than forward beyond the anterior edge of this orbit; 4) the globular swelling
at the posterior end of the parasphenoid excavated posteriorly rather than a
solid protuberance; 5) vomerine teeth in a roundish patch rather than a
transverse band; 6) molariform palatine teeth present rather than lacking
altogether; 7) two rather than one row of teeth for most of the length of the
dentary; 8) three rather than two rows of teeth for most of the length of the
premaxillary; and 9) a deeply excavated cavity in the basioccipital broadly
confluent with the myodome rather than only narrowly excavated and slightly
confluent. Type species Hypsocephalus atlanticus. The name Hypsocephalus
(vif/os, high or elevated, + cephalus, head) refers to the high, deep skull

6

Contributions in Science

No. 230

and the specific name atlanticus refers to the Atlantic Ocean, the general
locality of the fossil, in contrast to the eastern Pacific Ocean, the area where
the only living relative, Hoplopagrus, occurs.

Hypsocephalus atlanticus , new species
Figures 2-7

Holotype: LACM VP 27859, a single neurocranium, 49.4 mm long, with
the right posterolateral side broken off, thus the epiotic, exoccipital, pterotic,
and intercalar are absent from this side (see below). Other bones found in
definite association with, and certainly part of, this one fish are: a left cleith-
rum about three-fourths complete, the middle half of the right cleithrum, a
fragmentary anterior one-third of the right maxillary, the middle two-thirds
of the right premaxillary, the anterior one-fifth of the left premaxillary, the
middle three-fourths of the right dentary, a small fragment of the antero-
dorsal edge of the left dentary, the anterior half of the right articular, the
distal two-thirds and the anterior and proximal one-fourth of the right hyo-
mandibular, impressions of three anterior premaxillary teeth, about half of
the right exoccipital with the articular facet for the atlas vertebrae intact, one
complete neural arch and spine with the dorsal one-fifth of an anterior
abdominal vertebrae attached, several fragmentary branchiostegal rays
imbedded in a small piece of limestone, one dorsal spine pterygiophore, and
the posterolateral corner of the skull also imbedded in a limestone block.

Locality: LAV Loc. 7189 Florida, Jackson Co., T: 5N, R: 1 1W, Sec. 13,
2.7 airline miles NW of Marianna (Figure 1), collected by Brooks and
Suzanne Ellwood and Edward M. Renner on 2 April 1970. The skull was
taken from Milton’s Cave (Figure 1) in the lower member of the Crystal
River formation, the uppermost Eocene formation in this north Florida area.
The locality is a small, intricate cave, and the skull and associated bones were
found in a solution cavity of the cave 13.2 ± 0.5 meters below the surface of
the ground.

Diagnosis: As for the genus.

Description

Vomer: The ventral surface of the vomer is roughly circular and covered
with stout bluntly pointed teeth (Figure 4). The anteriormost tooth is longest
and largest in diameter. It is flanked posterolaterally on each side by a tooth
slightly smaller in diameter, and about half as high. Posterior to these three,
and partially between the posterior two, is a cluster of six small teeth. A
seventh small tooth was present as evidenced by a small empty socket just
posterior to the lateral robust tooth on the right side. Ventrally a low, rounded
keel on the vomer is continuous with that on the parasphenoid. Laterally a
low rounded ridge extends posterodorsally to, and is continuous with, that
of the lateral ethmoid. The vomer bears a broadly rounded bridge middorsally
as well, and with the ventral and lateral ones, the vomer is diamond shaped
in cross section just above the tooth patch. Posterodorsally the dorsal ridge

1972

New Genus and Species of Lutjanid Fish

7

bifurcates narrowly around the narrow rostral fenestra (of Starks, 1926) to
meet the supraethmoid.

Comparison: In Hoplopagrus the vomerine tooth patch is narrow antero-
posteriorly and wider laterally; an anterior transverse row of three or four
stout, almost molariform teeth is followed by a row of four to six much smaller
teeth similarly proportioned. All the living snapper genera known have villi-
form vomerine teeth (when they are present), and the vomerine tooth patch
assumes a variety of shapes (Gill, 1884; Regan, 1913; Norman, 1966; Ander-
son, 1967).

Parasphenoid: The anterior half of the parasphenoid bears a wide, thin
ventral keel, which is least developed anteriorly at the vomer-parasphenoid
articulation (Figure 2). The keel extends further ventrally to the posterior
and abruptly ends in 90° angle ventral to the ascending parasphenoid processes
articulating with the prootics. Just posterior to the keel, the parasphenoid
expands into a globular swelling that is concave posteriorly, and the para-
sphenoid continues posteriorly as a narrow, dorsoventrally flattened flange
ventral to, and articulating dorsally with the basioccipital.

Comparison: The parasphenoid of Hoplopagrus is similar to Hypsoce-
phalus except that the swollen brace for the upper pharynageals is not as
strongly developed. The parasphenoid of Lutjanus lacks this swelling, is rela-
tively longer, and is keeled ventrally. In Ocyurus a strong, rounded ridge
extends laterally and posterodorsally from the posterior end of the para-
sphenoid along the ventral and anterolateral edge of each otic bulla. In
Ocyurus the keel on the parasphenoid is very low. In all of the above except
Ocyurus the parasphenoid is straight, and the ventral edges of the vomerine
tooth patch, of the parasphenoid keel, and of the basioccipital lie along a
straight line. The ventral surface of the anterior half of the parasphenoid is
slightly concave in Pristipomoides and Rhomboplites. In Rhomboplites, Pris-
tipomoides, and Ocyurus the longitudinal profile of the parasphenoid is a shal-
low V. The ventralmost point is at the posterior end of the keel, just ventral
to the ascending processes.

Lateral Ethmoid: The lateral ethmoid is essentially rectangular antero-
posteriorly. The ventral edge broadly articulates with the parasphenoid, and
its anteroventral and posteroventral angles bear short robust pillars which
articulate with the palatine. The posterior pillar is shorter and its flat, longi-
tudinally oval facet faces ventrally. The flat oval surface of the anterior pillar
faces about equally anteriorly, laterally, and dorsally. The anterior pillar lies
on a ridge extending from the lateral corner of the vomer, through the facet,
and posterodorsally to the posteromedial portion of the lateral ethmoid where
it converges with a thick vertical ridge on the posterior edge of the lateral
ethmoid. This vertical ridge narrows ventrally, terminating in the posterior
facet. Dorsally it thickens, extends laterally, and its cancellous dorsal surface
articulates with the lateral edge of the frontal. A space separates the two
lateral ethmoids medially just under the frontals. The lateral ethmoids meet

8

Contributions in Science

No. 230

along the middle one-fourth of the vertical distance between the frontals
and the parasphenoid. Ventral to this midline contact the lateral ethmoids
are separated narrowly to their ventral articulation with the parasphenoid.
Ventral to and slightly lateral to the anterior opening of the supraorbital
lateral line canal in the frontal, the olfactory canal courses anteriorly from

v seth ieth

1cm

soc

Figure 2. Lateral view of the neurocrania. A, Hypsocephalus atlanticm (LACMVP
27859); B, Hoplopagrus guntheri (LACM 31774-1).

1972

New Genus and Species of Lutjanid Fish

9

the orbital cavity through the lateral ethmoid. This canal is a vertical oval in
cross section, 3 mm high and 1 mm wide.

Comparison: The two facets for articulation with the palatine are similar
to Hypsocephalus in Hoplopagrus , but in the latter both face slightly more
laterally. In Ocyurus and Lutjanus the anterior facet faces anteroventrally and
slightly laterally, and the posterior one faces ventrally and slightly anteriorly.
The facets are at the ends of ridgelike struts of bone in the above genera.
Rhomboplites and Pristipomoides bear these two facets in the same orientation
as in Hypsocephalus, but they are only slightly raised from the lateral bone
surface. The third facet which receives the medial side of the lachrimal lies
slightly more dorsal than the anterior palatine facet and is dorsal, lateral and
slightly posterior to the posterior palatine facet in Hypsocephalus, Hoplop-
agrus, Ocyurus , and Lutjanus . In Rhomboplites and Pristipomoides the
lachrimal facet is on the same level as the anterior palatine facet, and is above
and lateral to the posterior one. Hoplopagrus shares with Hypsocephalus the
strongly developed dorsal and dorsolateral portion of the lateral ethmoids. In
Lutjanus, Ocyurus, and Pristipomoides, this surface faces laterally and slightly
anteriorly, meeting about perpendicularly with the lateral edge of the frontals.
In Rhomboplites the lateral ethmoid faces more dorsally than in Lutjanus
and Ocyurus, but still meets the frontal with an abrupt angle rather than
through a continus surface as in the hoplopagrines. The olfactory nerve
foramen is large in Hoplopagrus (which has an exceptionally large nasal cap-
sule) and about the same relative size as in Hypsocephalus in the remaining
recent genera.

Supraethmoid: The dorsal surface of the supraethmoid is shaped like a
posteriorly directed arrowhead, its posterolateral surfaces bounded by the
frontals (Figure 3). The supraethmoid is widest just anterior to the frontals;
immediately anterior to this widest point it narrows in width by about
one-third. Here the anterolateral transverse edges each bear a short pointed
process. A thick median ridge extends anteriorly a short distance and then
bends perpendicularly and continues ventrally. The ridge narrows ventrally
and the anterior edge flattens out anteroventrally around the rostral fenestra.
This thick anterodorsal ridge of the supraethmoid narrows slightly postero-
ventrally before the lateral surfaces of the bone flare laterally to meet the
lateral ethmoids.

Comparison: The dorsal surface of the supraethmoid is similar to Hyp-
socephalus in Pristipomoides and is reduced to a small square in Rhomboplites
and Ocyurus. In Hoplopagrus it is oblong, gently rounded anteriorly and
directed posteriorly between the frontals. It is long and slender in Lutjanus,
where it is widest between the anterior tips of the frontals and gradually nar-
rows posteriorly. In the recent genera the anterior end of the dorsal surface
of the supraethmoid meets at right angles with the vertical, mid-longitudinal
ridge of the anterior surface which slopes ventrally and anteriorly. The rostral
fenestra is large in Hoplopagrus and Pristipomoides (as in Hyposocephalus) ,

10

Contributions in Science

No. 230

small in Ocyurus and L. synagris, and absent in Rhomboplites, L. campe-
chanus, and L. griseus. In Hoplopagrus a low ridge is present parallel and
lateral to the median ridge. It is absent in Hypsocephalus and is absent or
only present as a slight suggestion of a raised area in the other recent genera.

Frontal: The frontals are thin and dip medially to produce a shallowly
concave interorbital region. The articulation between them extends posteriorly
and slightly dorsally from the posterior apex of the supraethmoid to a vertical
through the middle of the bony orbit and the beginning of the supraoccipital
crest. Each frontal continues posteriorly and laterally of the supraoccipital
crest to a vertical between the two facets for articulation of the hyomandibular.
Anteriorly, laterally, and ventrally the frontals firmly articulate with the lateral
ethmoids, and continue posteriorly to rim the orbit. Along the posterodorsal
edge of each orbit they meet the sphenotics, the articulation itself continues

Figure 3. Dorsal view of the neurocrania. A, Hypsocephalus atlanticus (LACMVP
27859); B, Hoplopagrus guntheri (LACM 31774-1).

1972

New Genus and Species of Lutjanid Fish

11

medially and posteromedially of the dorsal edge of the orbit as a shelf forming
the anterolateral edge of the supratemporal fossae. This shelf continues pos-
teriorly and laterally on the pterotic bone.

Comparison: In all the recent lutjanid genera examined except Pristipo-
moides the frontals contribute to the anterior portion of the supraoccipital
crest, slightly in Rhomboplites and considerably in the remaining genera. In
Pristipomoides and Lutjanus a pod us the crest extends anterior to reach the
posterior third of the orbit. In Rhomboplites and Lutjanus griseus the crest
reaches only to a vertical line through the anterior third of the orbit diameter.
The longitudinal ridge which extends ventrally to meet the pterosphenoid
from the medial portion of each frontal is short in Rhomboplites and Ocyurus.
This ridge extends more ventrally in Hoplopagrus and Lutjanus to form about
the dorsal third of the wall separating the braincase from the orbit. The
anterior supraorbital canal foramen in the frontal opens over the anterior
third of the orbit in Rhomboplites and Lutjanus griseus, dorsal and slightly
medial to the anterior edge of the orbit in L. synagris and L. apodus, and
slightly anterior to the front edge of the orbit in Pristipomoides, Ocyurus ,
Hoplopagrus, and Hypsocephaus.

Parietal: Each parietal is largely a flat shelf which extends laterally and
slightly dorsally from near the posterodorsal margin of the orbit to the epiotic
posteriorly. The shelf forms the ventral and lateral surface of the shallow
supratemporal fossa. From the lateral edge of this shelf, the parietal extends
ventrally toward the pterotic about as far as it does medially toward the
supraoccipital. The vertical, laterally facing portion forms (with the frontal
anteriorly and the epiotic posteriorly) the medial boundary of the lateral
temporal fossa. This ridge diverges laterally toward the posterior region of
the skull. Dorsally the parietal bears a shallow longitudinal trough.

Comparison: Only the parietal of Rhomboplites resembles that of Hyp-
socephalus, largely covering the floor of the supratemporal fossae and extend-
ing laterally and slightly dorsally into the low ridges extending from the
epiotics to the frontals. In Hoplopagrus, Ocyurus, and Lutjanus these ridges
resemble each other and are oriented vertically and slightly laterally. In
Pristipomoides this ridge is largely restricted to the parietal bone, with a
slight contribution from the frontals anteriorly and no involvement of the
epiotics posteriorly. The ridges are parallel to the supraoccipital crest in
Rhomboplites, Ocyurus, Lutjanus griseus, L. synagris, and L. campechanus.
They diverge slightly laterally to the posterior in Hoplopagrus, Pristipomoides,
and L. apodus.

Epiotic: The dorsal surface of the epiotics inclines slightly posteriorly
and slightly laterally. The lateral half of the dorsal surface is shallowly
excavated for articulation with the dorsal limb of the posttemporal. The
medial half of the dorsal surface is slightly depressed, and there is no trace
of a posteriorly directed spine. The epiotic articulates ventrally with the
exoccipital via a strong columnar strut directed ventrally and slightly medially

12

Contributions in Science

No. 230

and posteriorly. This strut along with that of the exoccipital, forms the
posterolateral corner of the temporal region.

Comparison: Just ventromedial to the epiotic facet for receiving the
upper limb of the posttemporal, Hoplopagrus, Ocyurus, Rhomboplites, and
Lutjanus bear a posteriorly directed process. This spine is lacking in Hypso-
cephalus and Pristipomoides. The facet for the posttemporal faces posteriorly
and dorsolaterally in Ocyurus, Hoplopagrus, and Lutjanus and almost directly
dorsally in Rhomboplites and Pristipomoides.

Prootic: The prootic is gently inflated laterally and dorsally to accommo-
date the anterior end of the otolith. Posteriorly the prootic articulates with
the basioccipital below and the exoccipital above. It joins broadly with the
parasphenoid ventrally. A narrow portion directed dorsally and slightly
laterally occupies the ventral half of the anterior hyomandibular facet and
articulates with the sphenotic dorsally. A short shelf extends laterally from
and borders the posteroventral aspect of this facet. Medially the prootics
meet as a flat shelf forming the floor of the anterior one-third of the brain-
case and abut against the basisphenoid anteriorly. Each posterior and vertical
edge of this shelf extends anterolaterally from the anterior end of a thin
medial process of the basioccipital to form the anteromedial wall of the
chamber for the saccular otolith. Anteriorly and dorsally the prootic forms a
transverse squarish plate articulating with the basisphenoid ventromedially,
the pterosphenoid dorsomedially, and the sphenotic dorsally. Just dorsolateral
to the tripartite juncture with the basisphenoid and pterosphenoid the prootic
bears a large foramen. The pars jugularis, with its two large foramina, is
overlain by a narrow arch which extends ventrally, anteriorly, and medially
from the anteroventral corner of the anterior hyomandibular facet, and
broadly inserts on the anterolateral edge of the prootic. The shelf bordering
the posteroventral edge of the anterior hyomandibular facet bears a thin blade
which extends ventrally parallel and posterolateral to this arch. This blade
terminates in a free end about half way along the arch. The prootic bears
another free ending blade which originates on the medial side of the antero-
lateral edge of the prootic. It extends dorsally, parallel, and anteromedial to
this arch.

Comparison: The arch over the pars jugularis consists of a single pillar
of bone in Rhomboplites, Pristipomoides, Ocyurus, and Lutjanus. In Hop-
lopagrus the two incomplete arches parallel to the main one (as in Hypso-
cephalus) are present but not as extensively developed. In Hoplopagrus the
pars jugularis is partly bridged over anteriorly by three narrow flat shelves
of bone: a ventrolateral projection of the pterosphenoid, a ventromedial
extension of the sphenotic and a medial extension from the prootic where
it abuts the pterotic anterior to the hyomandibular facet. This arrangement
is only present in large Hoplopagrus. The posteroventral portion of the
prootic is inflated to accommodate the large otolith in Rhomboplites, Ocyu-
rus, Pristipomoides, Lutjanus synagris, and L. campechanus. In L. griseus

1972

New Genus and Species of Lutjanid Fish

13

and L. apodus it is only slightly expanded, but more so than in Hoplopagrus
which has a small otolith, as apparently did Hypsocephalus.

Sphenotic: The sphenotic is a flat bone which occupies the middle one-
third of the posterior face of the orbit and the dorsal one-half of the anterior
hyomandibular facet (Figure 2). The large eye and foreshortened skull leave
the sphenotic (and pterotic) with greatest dimensions in a vertical rather than
longitudinal direction. The sphenotic articulates with the pterotic posteriorly
via a suture proceeding vertically and then anteriorly from the posterior mar-
gin of the anterior hyomandibular facet. The sphenotic articulates with the
prootic ventrally via the anterior hyomandibular facet and via an articulation
extending medially from this facet to the articulation with the pterosphenoid.
Dorsally the sphenotic articulates with the frontal.

Comparison: In Lutjanus griseus, L. campechanus, Ocyurus, and Rhom-
boplites the sphenotic lies more dorsal and more anterior in the posterodorsal
quarter of the orbit. In Lutjanus synagris, L. apodus, Pristipomoides, and
Hoplopagrus it is placed only slightly more dorsal than in Hypsocephalus.
In all of the recent lutjanid genera examined the sphenotic bears a laterally
directed spine which originates on the posterior surface at the anterodorsal
corner of the anterior hyomandibular facet. This region is broken on both
sides in Hypsocephalus.

Pterotic: Like the sphenotic, the pterotic is oriented largely vertically.
Its dorsal two-thirds directly posterior to the sphenotic, and the ventral one-
third is posterior to as well as slightly ventral to the anterior hyomandibular
facet. The pterotic bears the entire posterior hyomanidbular facet. This facet
is horizontally elongate, and slightly wider posteriorly. It lies directly posterior
to the anterior facet, from which it is slightly separated. The short pointed
process of the pterotic just below and lateral to the articular surface for the
ventral limb of the posttemporal (on the intercalar) is directed ventrally and
only slightly laterally and posteriorly. Just ventral and medial to the posterior
hyomandibular facet, a shelf of the pterotic extends medially and slightly
ventrally to articulate with the intercalar and exoccipital posteriorly, and
the prootic anteriorly. The pterotic occupies the posterolateral edge of the skull
and has a ridge which bears the temporal lateral line canal. From this ridge
the pterotic dips ventromedially to meet (anterior to posterior) the frontal,
parietal, epiotic, and exoccipital, and forms the lateral border and floor of the
lateral temporal fossa.

Comparison: The pterotic in Rhomboplites, Ocyurus, Pristipomoides,
and Lutjanus extends almost directly posterior from the dorsal half of the
orbit. In these genera the process on the posterior lateral border of pterotic
extends largely posteriorly and only slightly laterally and ventrally. In Hoplo-
pagrus this spine projects much more ventrally as it does in Hypsocephalus,
and the pterotic is posterior to the middle half of the orbit in both of these
genera as well.

Intercalar: The intercalar is flat and occupies an almost horizontal, ven-

14

Contributions in Science

No. 230

trally facing surface (Figure 4). It is narrowly pointed laterally just ventral
and posterior to the posterolateral spine of the pterotic. The intercalar rapidly
widens medially and slightly ventrally to articulate broadly with the prootic
anteriorly and exoccipital posteriorly. This articulation lies just dorsolateral
to a shallow longitudinal groove marking the dorsal edge of the otic bulla.
The transversely oval, concave facet which receives the lower limb of the
postemporal faces directly posterior and about three-fourths of the facet lies
above the spine of the pterotic just lateral to the facet.

v leth 1cm fr spho pro pto

Figure 4. Ventral view of the neurocrania. A, Hypsocephalus atlanticus (LACMVP
27859); B, Hoplopagrus guntheri (LACM 31774-1).

1972

New Genus and Species of Lutjanid Fish

15

Comparison : The intercalar is roughly trapezoidal in Lutjanus, Ocyurus,
and Rhomboplites, the anterior and posterior edges roughly parallelling each
other, and the medial and lateral ones converging anteriorly. The intercalar
is quite similar in orientation and shape in Hypsocephalus and Hoplopagrus.
The facet for reception of the postemporal lies entirely above the lateral
pterotic process in Hoplopagrus. In the other genera examined the facet lies
directly medial to the process, and the facet is directed laterally as well as
posteriorly rather than just to the posterior. The facet is immediately medial
to the process of the pterotic in Lutjanus, Hoplopagrus, and Hypsocephalus ,
but is separated by a short, thin, horizontal ridge in Rhomboplites and Ocyu-
rus. In Pristipomoides the facet is posterior as well as medial to, and widely
separated from the pterotic process.

Basisphenoid: Only the dorsal half of the basisphenoid is present, assum-
ing the ventral, basal portion was originally possessed. A thin, compressed
piece of this basal limb extends anteroventrally. Posterodorsally it is narrowly
confluent with the two, dorsolaterally extending wings. From this attachment
these two wings spread a short distance, transversely and fanlike, to form a
small part of the central anteroventral wall of the braincase. They articulate
broadly with the pterosphenoids laterally. Just posterior to the basisphenoid
is the large hypophyseal foramen which the ventral and posteroventral tips of
each fan virtually encircle before articulating with the two prootics posteriorly.
The dorsal tips of the basisphenoid rise only slightly to articulate with the
pterosphenoids, and form the ventral border of a large, vertically elongate
opening which extends dorsally to the underside of the frontals and lies be-
tween the brain and orbital region.

Comparison: The dorsal edge of the basisphenoid in Hoplopagrus, as in
Hypsocephalus, curves dorsally only slightly towards the pterosphenoids. In
Lutjanus, Ocyurus, Pristipomoides, and Rhomboplites the dorsolateral edges
of the basisphenoid curve dorsally, entering the ventrolateral as well as the
ventral edge of the cavity connecting brain and orbit. The hypophyseal fora-
men is larger ( Lutjanus apodus ), about the same size ( Pristipomoides , Rhom-
boplites), or smaller ( Ocyurus , Hoplopagrus, Lutjanus synagris, L. campe-
charus, L. griseus) than in Hypsocephalus. The foramen is rounded in
Hypsocephalus and Hoplopagrus, and is transversely oval in the other snapper
genera.

A slight projection extends anterodorsally into the orbit from the basal
portion of the basisphenoid in Hoplopagrus. There is no such projection on
the basal portion in Rhomboplites and Pristipomoides, but a broad flat pro-
jection is present in Lutjanus and Ocyurus.

Pterosphenoid: The two pterosphenoids form much of the anterior wall
of the braincase and, much of the lateral border of the large foramen con-
necting the braincase with the orbital region. About the middle of the medial
margin of each bears a short, medially directed point of bone. On the left side
a small foramen occurs ventral and slightly lateral to this projection, and

16

Contributions in Science

No. 230

another foramen is present dorsolaterally about half way between the tip of
the projection and the surface of articulation with the frontal. On the right
side the corresponding ventrolateral foramen is lacking, and in the position
of the dorsolateral one are two smaller foramina. The pterosphenoid articu-
lates, in a broad arc laterally (dorsal to ventral) with the frontal, sphenotic,
prootic, and basisphenoid, respectively.

Comparison: In Lutjanus, Rhomboplites, and Ocyurus the pterosphenoid
occupies about the same relative position as in Hypsocephalus, and bears
slight ( Lutjanus , Rhomboplites ) to prominent, narrowly pointed medial
projections {Ocyurus). The pterosphenoids are straight edged medially in
Hoplopagrus, and lack any medial projections. The pterosphenoids are less
extensive in Hoplopagrus, where the frontals extend ventrally to occupy the
dorsolateral walls of the cavity connecting the brain with the orbit. This cavity
is a wide vertical oval in Hypsocephalus, only slightly elongated dorsoven-
trally in Hoplopagrus, much elongated dorsoventrally in Lutjanus and Ocyu-
rus, and narrowly constricted in Rhomboplites with a roughly circular opening
dorsal and ventral to a narrow interspace. In Pristipomoides the pterosphe-
noids firmly articulate medially for the middle third of the vertical distance
between the dorsal edge of the basisphenoid and the underside of the frontals.
Thus the orbit and brain cavity are connected by two subequal circular open-
ings occupying the dorsal and ventral one-third of this distance.

Basioccipital: The posterior basioccipital facet is vertical, facing directly
posterior. Anterior to this facet the bone is almost a vertical rectangle in cross
section, compressed to about two-thirds the facet width. Ventrally it is deeply
excavated and this cavitation extends anteriorly above the transverse posterior
end of the parasphenoid. Further dorsally and anteriorly this cavity opens
widely into the posterior myodome. Laterally the basioccipital forms the
ventrolateral wall of the posterior portion of the chamber for the saccular
otolith. A thin, compressed, medial extension runs anteriorly and slightly dor-
sally to articulate with the thick median juncture of the prootics.

Comparison: The whole posterior facet of the basioccipital lies in one
plane and faces somewhat posterodorsally in Lutjanus, Ocyurus, and Rhom-
boplites. The posterior facet faces posteriorly in Hoplopagrus. The antero-
dorsal portions which enter the otic bulla are somewhat more expanded in
Lutjanus synagris and Ocyurus, but are compressed in Lutjanus griseus and
Hoplopagrus. Rhomboplites and Pristipomoides have large otoliths and a
widely expanded basioccipital. The basioccipital is only narrowly excavated
ventrally in Lutjanus and the cavity is not confluent with the posterior myo-
dome. The cavity is small and narrowly confluent in Hoplopagrus, and large
and confluent in Ocyurus, Pristipomoides, and Rhomboplites. In Rhombo-
plites the basioccipital is wider (almost square) just anterior to the posterior
facet rather than being narrowly compressed and rectangular.

Supraoccipital: The supraoccipital penetrates anteriorly between the
frontals to a position slightly behind a vertical through the middle of the

1972

New Genus and Species of Lutjanid Fish

17

orbit. It widens posteriorly and along the posterior edge of the skull occupies
the medial half of each supratemporal fossa. The lateral surfaces slope slightly
ventrolaterally to the parietal and epiotic. The supraoccipital crest is broken
dorsally and is described from photographs taken before it was collected when
the crest was more complete. The supraoccipital crest is low, only slightly
higher than the dorsal surface of the skull anteriorly. A slight ridge extends
posterodorsally from the posterodorsal border of the skull on each side of the
crest. About 5 mm along this ridge, another ridge extends from it posteriorly
and slightly ventrally.

Comparison: The supraoccipital (in the supraoccipital crest) extends
anteriorly to a vertical through the center of the orbit in Hoplopagrus, Rhorn-
boplites, and Ocyurus. In Lutjanus and Pristipomoides it extends only to a
vertical through the posterior one quarter of the horizontal orbit diameter.
The ridge extending posterodorsally from the posterodorsal corner of the
skull bears a ventral branch distally in Hoplopagrus, Rhomboplites, and Ocyu-
rus, but the ventral branch is lacking in Lutjanus. In Rhomboplites, Ocyurus,
and Hypsocephalus the main branch extends to the posterodorsal apex of the
supraoccipital crest, but in Lutjanus and Pristipomoides this ridge reaches to
a point a little below the apex along the vertical posterior edge of the crest.

Exoccipital: Although the right exoccipital is fragmentary, the left is com-
plete. Clearly both met in the midline over the basioccipital, and the ventral,
lateral and at least dorsolateral walls of the foramen magnum were bounded
by the exoccipitals (Figure 5). Each flat facet receiving the dorsal portion of
the atlas vertebra is a regular transverse oval facing ventromedially. The medial
edge terminates a millimeter or two short of the midline and the exoccipital
facets did not form a continuous articular surface. From the facet a strong
pillar of bone extends anterodorsolaterally and forms the posterolateral corner

soc

Figure 5. Posterior view of the neurocrania. A, Hypsocephalus atlanticus
(LACMVP 27859); B, Hoplopagrus guntheri (LACM 31774-1).

18

Contributions in Science

No. 230

of the skull. The ridges soon bifurcate, one branch extending dorsally and
slightly anteriorly to the epiotic, and the other laterally and slightly anterior
to the intercalar and pterotic. A foramen is present just dorsolateral to the
articular facet for the atlas vertebra, and another larger one pierces the
exoccipital just anterior to the pillar of bone which extends laterally towards
the pterotic.

Comparison: Medial extensions from, and about half as wide as, the exoc-
cipital facets form a continuous articular surface across the midline in Ocyu-
rus, Rhomboplites, Pristipomoides, and Lutjanus. These facets have narrowed
medial extensions in Hoplopagrus but do not meet in the midline. The exoc-
cipitals have greater antero-posterior extent in Lutjanus and Ocyurus than in
Rhomboplites , Hoplopagrus, and Hypsocephalus.

Dentary: About three-fourths of the right dentary is present, and its
medial side is covered with limestone (Figure 6). Only about one-eighth of
the left dentary remains. The dentary is robust and bears a lateral row of
robust, bluntly pointed teeth about the size of the larger vomerine teeth. These
lateral teeth diminish slightly in size posteriorly. An inner row of robust teeth
about half the size of the outer teeth is visible on the fragmentary left dentary.
This inner row extends backward to at least half the length of the dorsal limb
of the dentary. The symphysis is lacking, but the dentaries seem to be oriented
as they were in life, and appear to have met via a deep, strong articulation.
The dentary rises sharply posteriorly and it appears that the length of the
intact dentary is subequal to the vertical distance between the posterior ends.

Comparison: Hypsocephalus and Hoplopagrus have an outer row of
bluntly pointed robust teeth. In Hypsocephalus the nature of the dentition on
the anterior ends of the dentaries is unknown. In Hoplopagrus two or three
larger blunt canines are developed anteriorly in each dentary, and the inner
row is restricted to two or three smaller teeth present just behind these canines.
In the remaining genera a single row of slender to robust canines is followed
by a small number of fine inner teeth restricted to the anterior one-third or
less of each dentary.

Articular: Only the anterior one-third of the right articular is present,
and it essentially occupies its normal position between the posterior limbs of
the dentary (Figure 6). The articular is deep and robust like the dentary.

Comparison: The most that can be seen from the fragmentary articular
is that it is relatively deep, at least anteriorly, as are the articulars in the other
genera.

Cleithrum: Only a fragment of the left cleithrum is present (Figure 6).
More complete and relatively intact cleithra were present in the cave before
the skull was collected, and they are described from photographs of two views.
Three-fourths of the left cleithrum and the middle third of the right one were
present. The upper limb was pointed on the anterodorsal edge. A short dis-
tance below it widens perpendicularly backward, so the flat plate of the dorsal
limb has a largely horizontal dorsal surface. The posterior edge was about

1972

New Genus and Species of Lutjanid Fish

19

anterior portion of the right articular. B, Lateral view of distal two-thirds of the
right hyomandibular. C, Posterior view of B. D, Lateral and slightly anterior view
of the fragmentary left cleithrum.

20

Contributions in Science

No. 230

vertical and the dorsal limb is roughly uniform in width below the narrow
anterodorsal point. The posteroventral angle of the dorsal limb bears a short,
rounded, flat projection just above the position of the coracoid. The antero-
ventral limb possesses a wide flange laterally. Both ventral limb flanges lack
the anterior one-fourth of their extent. A shallow groove is present on the
lateral edge and it extends dorsally and anteroventrally about one-half the
distance of each limb. The dorsal and posterior edge of the upper limb are
broken so the exact shape is not known.

Comparison: The cleithra of recent genera show only minor differences
which are not discernible on the fossil.

Hyomandibular: Most of the ventral limb and the anterior third of the
proximal articular portion of the right hyomandibular are present. Only the
anterior facet which articulates with the skull is present of the three proximal
articular surfaces of the hyomandibular. The facet is flat, slightly oval dorso-
ventrally, and bears a small notch ventrally. From this facet a ridge extends
posteriorly and laterally. It runs into a strong, dorsoventral ridge which is
directed anterolaterally, and is broken dorsally and ventrally. The anterior flat
blade of the hyomandibular extends ventrally a short distance to a transverse
break in the bone. The medial ridge leading from the anterior to the posterior
articular facet is broken just behind the anterior facet. The thickened pos-
terior edge of the hyomandibular is longitudinally oval in cross section, and
is hollow in the distal one-third. The proximal two-thirds is hollow also, and
opens out posterolaterally via an elongate oval foramen. Along the posterior
edge of the shaft a groove originates near the distal end of this foramen. The
groove widens and deepens proximally to the broken end which lies about two-
thirds of the estimated total length of the intact bone from the distal end
(Figure 6).

Maxillary : Only the ventral half of the anterior third of the maxillary is
present (Figure 7 C, D, H). The anterior excavation which accommodates
the premaxillary was high and narrow. The ventral edge of the medial side
bears a low rounded swelling which articulates laterally with the ascending
process of the premaxillary. Dorsal to this swelling extends a low, rounded,
vertical ridge, which is about as long as the thickness of the shaft of the
maxillary. The shelf extending anterolaterally from the head of the maxillary
was thin and does not appear to have been expanded distally.

Comparison: The medial articular surface and ridge dorsal to it are rela-
tively smaller in the fossil than in Hoplopagrus. The ridge is sharp rather than
rounded in Hoplopagrus as well. The ridge is rather sharp edged and much
higher and longer in all the other recent snapper genera. The ridge is straight
edged in the hoplopagrines and Rhomboplites and is a raised semicircle in the
other recent genera.

Premaxillary: The premaxillary is dorsoventrally flattened and bears an
outer row of enlarged teeth about the same size as those on the dentary and
an inner double row of molariform teeth about one-third the size of the outer

1972

New Genus and Species of Lutjanid Fish

21

Figure 7. Bones of the jaws and palate of Hypsocephalus atlanticus. A, Anterior
view of anterior two-thirds of right premaxillary. B, Ventral view of A. C, Dorsal
view of the head of the right maxillary. D, Ventral view of C. E, Medial view of
anterior end of left premaxillary. F, Ventral view of E. G, Ventral view of posterior
portion of left palatine. H, Medial view of the head of the right maxillary.

22

Contributions in Science

No. 230

ones (Figure 7 A, B, E, F). The second tooth from the medial end of the
bone in the outer row is about twice as large (at least in diameter) as the
others. The bases of the two medial teeth are present on the fragmentary
anteromedial end of the left premaxillary and most of the shaft is known from
the right one. One or two more enlarged teeth may have been present at the
anterior end of the premaxillary, but a total of more than three is unlikely.

Comparison: Hoplopagrus consistently has only a single row of smaller
inner teeth, and usually has two enlarged canines anteriormost in the outer
row of each premaxillary. The other recent snapper genera have a single
outer row of canines with the anterior one to five teeth enlarged. The inner
ones consist of two to five rows of villiform teeth, usually with more rows
anterior and medial and less posterior and lateral.

Palatine : The small posterior fragment of the left palatine bears 12 small
molariform teeth (Figure 7 G).

Comparison: This piece could have been from a palatine bone shaped
like that of Lutjanus, Ocyurus, Rhomboplites, or Pristipomoides, but it is too
fragmentary to discern the original shape. Hoplopagrus lacks palatine teeth
and the bone is a narrow shaft without the wide flattened area which bears
palatine teeth in other lutjanids.

Dorsal Spine Pterygiophore: The left side of the first dorsal spine ptery-
giophore and its two dorsal spines are exposed on a small chunk of limestone.
The basal two-thirds of the moderately robust spines are present, and articu-
lated with the pterygiophore. The pterygiophore is flat with a low, flat, straight
ridge running ventrally from the articulation of the second spine. This ridge
lies slightly posterior to a line vertically bisecting the lateral surface of the
bone, and is slightly enlarged and rounded just ventral to the second spine
base. Just ventral to the first spine, the pterygiophore bears a low, rounded
protuberance which extends anterodorsolaterally. The length of the dorsal
surface of the pterygiophore is about half the height of this bone; the ventral
tip is broken and the height cannot be precisely determined.

Comparison: The lateral ridge below the second dorsal spine base, and
the low protuberance below the first spine are of similar configuration in all
the other snappers, except possibly that of Pristipomoides which was not
examined. In the recent snaper genera the depth is two and one-half to four
times the length of the dorsal edge, rather than about twice as in Hypsocepha-
lus. The first two dorsal spines are slightly compressed, long and slender, and
the first spine is just about half the length of the second in the recent snapper
genera, and this appears to have been true for the fossil although only the
proximal two-thirds to three-fourths of each spine is present.

Discussion

The fossil resembles the living Hoplopagrus guntheri more than any
other percoid examined. The living and fossil species together appear closest
to lutjanids, although they are distinctive in their own right, and also bear

1972

New Genus and Species of Lutjanid Fish

23

some resemblance to sparids and pomadasyids. This conclusion is largely (and
necessarily) based on characters in the neurocranium and jaws, the only
elements available in the fossil. The nature of the teeth on the premaxillaries,
dentaries, vomer, and palatines has been stressed since they are often the only
osteological features described for Indo-Pacific percoids (Weber and de Beau-
fort, 1931, 1936; Gosline and Brock, 1960; Smith, 1961; Marshall, 1964).
Detailed search for relationships has been restricted to three families Lutjani-
dae, Pomadasyidae, and Sparidae. Comparative materials has been listed pre-
viously and the following works have also been useful: Gregory (1933);
Patterson (1964); and Leccia (1961).

Characters taken together which distinguish the hoplopagrines from
other percoids are: 1) robust, conical teeth on the premaxillaries, dentaries,
and vomer; 2) a vertical and transverse posterior facet on the basioccipital;
3) articular surfaces of exoccipitals for the atlas vertebrae not continuous
across the midline; 4) a ventral, globular swelling on the posterior end of the
parasphenoid; 5) a narrow, compressed otic region; 6) a strong, compact
dorsal surface of the lateral ethmoids lateral to the anterior ends of the
frontals; 7) lateral ethmoid facets for the palatine oriented as in lutjanids
(see below); and 8) supraethmoid (not vomer!) convex in profile. Many of
these characters are found elsewhere in percoids.

The robust conical teeth occur also among the lutjanids and sparids but
show more variation in size in these families. In lutjanids the outer robust
teeth are followed by minute villiform teeth, and the vomerine teeth are
usually present and villiform. The only exceptions are some species of the
lutjanid genus Lethrinus which have canines anteriorly and molariform teeth
posteriorly (Weber and de Beaufort, 1936). Pomadasyids resemble lutjanids
in having strong canines followed by fine villiform teeth, or having all jaw
teeth villiform. Sparids all show considerable range of tooth shapes, with
canines or incisors anteriorly and conical or molariform teeth posteriorly.
The relatively uniform shape, lack of great dimorphism in size, the teeth
diminishing in size posteriorly, and the presence of teeth on the vomer in
hoplopagrines makes them similar to lutjanids.

The posterior facet of the basioccipital faces posterodorsally in poma-
dasyids and lutjanids. This facet is transverse and vertical in sparids as it is in
hoplopagrines. In sparids the articular surfaces of the exoccipital facets accom-
modating the atlas vertebrae vary. Those of Calamus and Archosargus do not
meet in the midline like those of hoplopagrines. They meet narrowly in Lago -
don, and form a wide continuous surface in Diplodus. These facets meet
through a continuous surface in lutjanids and the pomadasyids Brachydeu-
tereus, Orthopristis, Anisotremus, Haemulon aurolineatum, and H. sciurus.
They fail to meet middorsally in Haemulon plumieri. The hoplopagrines most
resemble some sparids and some pomadasyids in the relations of the facets
for the atlas vertebrae.

No development of a globular swelling at the posterior end of the para-

24

Contributions in Science

No. 230

sphenoid is apparent in the lutjanids examined or in the pomadasyid Aniso-
tremus. The remaining pomadasyids and all the sparids examined have such
a swelling moderately to well developed. In contrast to the solid rounded
protuberance of the hoplopagrines, the swelling in sparids and pomadasyids
is bilateral with a midventral longitudinal groove partially dividing it. The
hoplopagrines seem to resemble sparids and pomadasyids rather than lutjanids
in possessing this swelling, but since it is differently formed in hoplopagrines
it may be independently developed and not indicative of relationship.

All the lutjanids and pomadasyids examined have moderately to greatly
inflated otic regions, but the hoplopagrines resemble the sparids examined in
having a compressed otic region.

The size and arrangement of the cephalic lateral line system pores in
hoplopagrines is within the range of variation seen in the lutjanids and sparids
examined. The pomadasyids have distinctive large cephalic canals quite dif-
ferent from those in hoplopagrines.

The lateral ethmoids have a well-developed dorsal surface lateral to the
anterior ends of the frontals in sparids and pomadasyids, and this surface is
deeply excavated in all of the genera examined in these two families except
in Brachydeutereus where the upper surface is only a shallow depression. In
hoplopagrines this surface is rugose and flat or rounded as it is in lutjanids,
although the surface faces largely laterally and slightly anteriorly in lutjanids
rather than dorsally.

The orientation and position of the palatine facets on the lateral ethmoids
of the hoplopagrines resemble those of all the lutjanids examined, namely one
some distance behind the other with the anterior one slightly more dorsal and
slightly more lateral than the posterior one. In pomadasyids the vomer and
lateral ethmoid are longer and the facets are much closer together. The ante-
rior one is directly anterolateral and slightly dorsal to the posterior one. The
anterior facet faces much more laterally than in lutjanids as well. In sparids
the anterior facet is strongly developed and faces anteriorly and slightly lat-
erally, and the posterior facet is obsolescent. The supraethmoid is similar in
size and shape in hoplopagrines and lutjanids, namely with a flat dorsal sur-
face between the anterior ends of the frontals with a midventral keel anterior
to this. The bone is convex dorsally in profile. In the sparids and pomadasyids
examined the supraethmoid is concave in profile and is flat or excavated along
the middorsal line, apparently to accommodate the long ascending processes
of the premaxillaries.

In most of the characters shared by Hoplopagrus and Hypsocephalus
and just discussed, the hoplopagrines resemble the lutjanids. A few characters
like the swelling at the posterior end of the parasphenoid and the orientation
of the posterior exoccipital facets resemble some sparids and some poma-
dasyids, but are differently developed or variably developed enough so that
they do not seem to be strong indicators of relationship.

The characters in the dentition and ethmoid region of hoplopagrines are

1972

New Genus and Species of Lutjanid Fish

25

probably the strongest evidence of a relationship with the lutjanids. The eth-
moid region, maxillaries, and premaxillaries are basically similar and reflect
the capability to expand the oral cavity both ventrally and laterally. The
lutjanids, including hoplopagrines, are predaceous and have moderately pro-
trusible mouths which also expand laterally, producing a large enough opening
to utilize vomerine and palatine teeth which are present in most lutjanids.
In both sparids and pomadasyids the mouth is restricted laterally, and the
upper jaw is much more protrusible. These fish are largely nibblers and grazers
(Randall, 1967), and the restricted lateral movement of smaller mouths has
eliminated the need for vomerine or palatine teeth which are uniformly lack-
ing in these two families. Thus, the hoplopagrines are interpreted as lutjanids
which have retained the typical larger, expansive mouth, but have specialized
to feed on resistant prey by developing strong, robust teeth resembling those
of some sparids.

Hypsocephalus is distinctive among lutjanids in possessing two flanges
on the prootic, one anterolateral and another posterolateral to the main col-
umnar arch forming the anterolateral wall of the prootic. Sparids typically
have two complete arches (Patterson, 1964) and I found this in all the sparids
examined except Lagodon in which the posterior one is incomplete, resem-
bling the posterior flange of Hypsocephalus. All the pomadasyids examined
have a single complete arch with an additional free ending posterolateral
flange from the shelf under the anterior prootic facet for the hyomandibular.
More variation probably exists than has been suspected and this character
should be investigated in as many acanthopterygians as possible.

Hypsocephalus is also unique among the lutjanids and sparids examined
in lacking a pointed process on the epiotic just medial to the facet for the
upper limb of the posttemporal. Among the percoids examined this process
is lacking in all the pomadasyids, and the significance of this absence is not
known.

The living Hoplopagrus guntheri has tubular anterior nostrils, a well-
developed knob on the upper interopercle, thick, enlarged, and conical canines
on the anterior ends of the dentaries and premaxillaries, conditions which can-
not be determined in the fossil. Lutjanids generally possess the interopercular
knob, but it is usually less well developed, and the sparids examined lack it.
The sparids lack palatine teeth also, and the large canines and tubular nostrils
are unique for Hoplopagrus among lutjanids and sparids. The lateral ethmoid
canal for the olfactory nerve in Hypsocephalus atlanticus is about the same
size as in other lutjanids, and it apparently did not have a particularly large
nasal capsule, as does Hoplopagrus (Pfeiffer, 1964).

Lutjanids in general feed largely on crustaceans and fishes, with fishes
forming a greater proportion of the diet in larger individuals (Randall, 1967).
The strong molariform teeth of Hoplopagrus indicates that it eats resistant
prey of some kind, as presumably did Hypsocephalus atlanticus. Edmund
Hobson (personal communication) found Hoplopagrus to be nocturnal, and

26

Contributions in Science

No. 230

believes that its feeding habits may be similar to a nocturnal Hawaiian sparid,
namely Monotaxis grandoculis (Forskal). He finds this Hawaiian sparid to
feed largely on hermit crabs and sea urchins which move out into the open
more at night. Possibly the feeding habits of Hoplopagrus are similar, as
perhaps were those of the fossil.

The description of the holotype of Lutjanus avus W. K. Gregory, 1930
shows that it consisted of a majority of the cranial bones, and they seem to
be typical of the genus Lutjanus. Lutjanus avus had strong outer teeth and
fine inner ones on the dentary and premaxillaries, and villiform vomerine
teeth. Unfortunately the holotype of L. avus could not be found during a
thorough search of the Florida Geological Survey Collections in 1957, (Stan-
ley J. Olsen, personal communication), and thus it has not been re-examined.

The holotype of Lutianus hagari Jordan and Gilbert, 1919, recently
transferred from Stanford University to the California Academy of Sciences,
and the counterpart (LACM 1329), were examined and they do not repre-
sent a lutjanid as Jordan and Gilbert (1919) believed. The first four or five
dorsal spines (11 in all) are longest. The next to the last four are about half
the height of the anterior ones. The last spine is a little longer than these
four and is very close to the much longer first soft ray, a condition found in
percichthids, scorpaenids, percids, and some sciaenids, but not in lutjanids.
The anal fin almost certainly had three anal spines, although the anteriormost
small spine is difficult to distinguish. The posterior two spines are clearly
marked, and the second is about one-third the diameter and about two-thirds
the length of the third. A good number of cycloid body scales is present. The
scale focus is placed posteriorly and six to nine radii occupy the anterior fields.
The skull is badly crushed but at least the dentaries (and probably the pre-
maxillaries) appear to have borne villiform teeth along with small canines.

The arrangement of dorsal spines and the cycloid scales definitely exclude
the fossil from the family Lutjanidae. Percichthids (except Stereolepis ) and
percids are extremely unlikely in deep water Miocene deposits from Cali-
fornia. All of the characters of the fossil noted above are found in Stereolepis
and many scorpaenids, and upon thorough study the fossil of Lutianus hagari
may prove to be one of these.

Geology and Paleoecology

About fourteen and one-half meters of limestones ranging from Oligo-
cene to Eocene in age occur within the measured stratigraphic section at
Milton’s Cave. The highest beds which outcrop at the surface are the marine
Oligocene Marianna limestones about 3.5 meters thick. Under these are upper
Eocene limestones which have been extensivly studied (Puri, 1957; Cheetham,
1963). The Ocala group is the uppermost late Eocene bed, and the top of the
Ocala group is represented by the Crystal River Formation. The Crystal River
Formation is divided into an upper Bumpnose Member and a lower member.
The skull of Hypsocephalus atlanticus was discovered in this lower member.

1972

New Genus and Species of Lutjanid Fish

27

The lower member is also comprised of an upper and lower zone. The skull
came from the upper zone which is a white to light brown, creamy, generally
soft, granular relatively permeable and pure limestone. This zone is quite
porous, has been carried into solution over large areas of the cave, and is
called the Operculinoides ocalanus-Asterocyclina Zone by Puri and Vernon
(1964). The two zones of the lower member are hard to distinguish and
locally grade into each other.

The Operculinoides-Asterocyclina Zone indicates a depositional environ-
ment of a continental shelf region between 33 and 66 meters, with salinities
from 32 °/00 to 37 °/00 water temperatures of 20°C or more, moderate
agitation, and no evidence of reef formation (Cheetham, 1963; Puri and Ver-
non, 1964). Conditions found today between the continental shelf margin of
Florida and the Bahamas Bank seem to be analogous with those which existed
in the late Eocene, namely a gently sloping continental shelf bounded on the
outside by a depression (Suwanee Straits of Eocene times) beyond which
existed a bank (Ocala Bank of Eocene time). In the late Eocene, the main-
land was in southern Alabama and Georgia and the highlands of central
Florida were occupied by the Ocala Banks.

The specimen of Hypsocephalus atlanticus died and was deposited at
moderate depths on a mainland shelf. The excellent three dimensional preser-
vation indicated a relatively undisturbed bottom. A fish entombed in a sedi-
ment consisting of these fine foraminiferal particles may have been well
preserved due to anaerobic bacterial action (Dunkle and Olsen, 1959). Dur-
ing late Eocene time the north Florida area was tropical or subtropical and
the sea level was gradually falling (Cheetham, 1963). There was a progressive
extinction of endemic forms among the cheilostome bryozoa (Cheetham,
1963), and the line of hoplopagrine snappers may have become extinct in the
western Atlantic in this period as well. However, the Miocene, Pliocene, and
Pleistocene also saw substantial sea level falls ( although at progressively lower
levels than the Eocene deposits) which were accompanied by cooling (Tanner,
1968), and the extinction of the hoplopagrines may have taken place at one
of these later times. A tropical and subtropical shallow water reef shark genus,
Heterodontus, was present through Miocene times in the western north
Atlantic, but is known today only from the eastern Pacific, Indo-Pacific,
eastern Atlantic and the Indian oceans. Both Heterodontus and Hypsocepha-
lus may have been eliminated at the same time, when conditions in the Carib-
bean area apparently became unfavorable for warm water forms living about
hard substrates at shallow and moderate depths.

Acknowledgments

We wish to thank the Florida State Cave Club of Florida State Univer-
sity, Tallahassee, particularly members Robert Royal, Eugene Neel, and
Edward Renner for assisting the junior author in discovery and collection of
the fossil. Their interest and concern for making the specimen available to

28

Contributions in Science

No. 230

science is highly commendable. Elizabeth Wing, Robert Christensen, David
Cupka, and .Tack Rudloe all helped us obtain skeletal material of recent
snappers, grunts, and porgies used in this study. An opportunity to collect
comparative material of Hoplopagrus guntheri in Baja California was afforded
by the generosity of the Janss Foundation, Thousand Oaks, California. Wil-
liam Eschmeyer, Lillian Dempster, and Pearl Sonada aided in sorting the old
Stanford University collection of fossil fishes now in the California Academy
of Sciences. The map of Milton’s Cave was drafted by Edward Renner, James
Leaird, and the junior author from a Bruntoe and Tape survey by the Florida
State Cave Club made on 5 July 1970. The photographs of the fossil in the
cave were made by the junior author. Various editions of the manuscript were
typed by Terri Kato, Barbara Savino, and Janet Dock. The excellent drawings
of the neurocrania of the fossil and of Hoplopagrus were done by Mary Butler,
and the remaining illustrations were made by the senior author.

Literature Cited

Anderson, W. D. 1967. Field guide to the snappers (Lutjanidae) of the western
Atlantic. U.S. Fish Wild!. Ser., Bur. Comm. Fish. Circ. 252, 14 p.

Casier, E. 1966, Faune Ichthyologique du London Clay. Brit. Mas. Nat. Hist.,
London, xiv + 496 p., plus a separate Atlas of 68 pi.

Cheetham, A. H. 1963. Late Eocene zoogeography of the eastern Gulf Coast
region. Geol. Soc. Amer. Mem. 91: xii + 113 p.

Danil’chenko, P. G. 1967. Superorder Teleostei, In: Fundamentals of Paleon-
tology, Vol. XI, Agnatha, Pisces. D. V. Obruchev (Ed), p. 603-712, Israel
Program for Scientific Translations, Jerusalem.

Dunkle, D. H., and S. J. Olsen. 1959. Description of a beryciform fish from the
Oligocene of Florida. Fla. Geol. Surv. Contrib. Fla. Vert. Paleontol. Pap.
No. 3, 20 p.

Gill, T. N. 1862a. On new generic types of fishes. Proc. Acad. Nat. Sci. Phila.
13: 77-78.

1862b. Catalogue of the fishes of lower California in the Smithsonian
Institution, collected by J. Xantus. Proc. Acad. Nat. Sci. Phila. 14: 249-262.

_ 1884. The osteological characters of the Lutjaninae. Proc. U.S. Nat.

Mus. 7(23): 351-355.

Gqsline, W. A. and V. E. Brock. 1960. Handbook of Hawaiian Fishes. Univ.
Hawaii Press, Honolulu. 372 p.

Gregory, W. K. 1930. A fossil teleost fish of the snapper family (Lutianidae) from
the lower Oligocene of Florida. Fla. St. Geol. Surv. Bull. 5:7-17.

--- 1933. Fish Skulls. A study of the evolution of Natural Mechanisms.

Trans. Amer. Phil. Soc. 23(2) : vii + 481 p.

Jordan, D. S., and B. W. Evermann. 1898. The fishes of North and Middle Amer-
ica: A descriptive catalog of the species of fish-like vertebrates found in the
waters of North America, north of the isthmus of Panama. Bull. U.S. Nat.
Mus. 47 (Pt. II), i-xxx, 1241-1936.

Jordan, D. S., and J. Z. Gilbert. 1919. Fossil fishes of southern California. I. el and
Stanford Jr. Univ. Pub. (Univ. Ser.) 38: 1-64.

Leccia, F. M. 1961. Osteologia comparada en echo especies de Fomadasyidae
(Pisces-Perciformes) del Golfo de Cariaco, Venezuela, y areas adyacentes.
Bol. Inst. Ocean. Univ. Oriente. 1(2): 396-473.

1972

New Genus and Species of Lutjanid Fish

29

Lehman, J. -P. 1966. Actinopterygii. In: Traite de Paleontologie. Tome IV, vol. 3,
p. 1-242, Jean Piveteau (Ed.) Masson et Cie, Paris, France.

Marshall, T. C. 1964. Fishes of the Great Barrier Reef and coastal waters of
Queensland. Angus and Robertson, Sydney, xiv + 566 p.

Norman, J. R. 1966. A draft synopsis of the Orders, Families, and genera of recent
fishes and fish-like vertebrates (excluding Ostariophysi, Scleroparei, Ammody-
tidae and a few other families, notably Centrarchidae, Percidae and Cichlidae)
covering literature up to 1938, and as far as it was available to the author,
from 1938 to 1944. Brit. Mus. Nat. Hist., London. 649 p.

Patterson, C. 1964. A review of Mesozoic Acanthopterygian fishes, with special
reference to those of the English Chalk. Phil. Trans. Roy. Soc. Lond. Ser. B,
247 (739): 213-482.

Pfeiffer, W. 1964. The morphology of the olfactory organ of Hoplopagrus guen-
theri, Gill, 1862. Can. J. Zool. 42(2) :235-237.

Puri, H. S. 1957. Stratigraphy and zonation of the Ocala group. Fla. Geol. Surv.
Bull. 38. 248 p.

Puri, H. S., and R. O. Vernon. 1964. Summary of the Geology of Florida and a
guidebook to the classic exposures. Fla. Geol. Surv. Spec. Publ. No. 5. ix +
312 p.

Randall, J. E. 1967. Food habits of reef fishes of the West Indies. Stud. Trop.
Oceanogr. Miami 5: 665-847.

Regan, C. T. 1913. The classification of the Percoid fishes. Ann. Mag. Nat. Hist.
8th Ser., 12(17): 111-145

Romer, A. S. 1966. Vertebrate Paleontology. Univ. Chicago Press, Chicago, 111.
ix + 468 p.

Smith, J. L. B. 1961. The sea fishes of southern Africa. 4th Ed. Central News
Agency, South Africa, xvi + 580 p.

Starks, E. C. 1926. Bones of the ethmoid region of the fish skull. Stanford Univ.
Pub., Univ. Ser., Biol. Sci. 4(3) : 139-338.

Tanner, W. F. (Ed.) 1968. Tertiary sea level fluctuations. Paleogeography, Paleo-
climatol., Paleoecol. 5: 5-171.

Walford, L. 1937. Marine game fishes of the Pacific Coast from Alaska to the
Equator. Univ. Calif. Press, Berkeley, xxix + 205 p., 69 pi.

Weber, M., and L. F. de Beaufort. 1931. The fishes of the Indo-Australian Archi-
pelago. Vol. VI, Perciformes (continued). E. J. Brill, Leiden, xii + 488 p.

— 1936. The fishes of the Indo-Australian Archipelago. Vol. VII, Perci-

formes (continued). E. J. Brill, Leiden, xvi + 607 p.

Weiler, W. 1968. Otolithi Piscium. Fossilium Catalogue 1: Animalia, Pars 117.
196 p.

Woodward, A. S. 1901. Catalogue of the fossil fishes in the British Museum Nat-
ural History. Pt. IV. Containing the actinopterygian Teleostomi of the sub-
orders Isospondyli (in part), Ostariophysi, Apodes, Percesoces, Hemibranchii,
Acanthopterygii and Anacanthini. London, xxxviii + 636 p.

Accepted for publication April 4, 1972

NUMBER 231
JUNE 23, 1972

^ ^ 7f 7g

CaLrcf

THE STATUS OF LEPTODACTYLUS PUMILIO
BOULENGER (AMPHIBIA, LEPTODACTYLIDAE)
AND THE DESCRIPTION OF A NEW SPECIES
OF LEPTODACTYLUS FROM ECUADOR

By W. Ronald Heyer

CONTRIBUTIONS IN SCICNCC

©

NATURAL HISTORY MUSEUM • LOS ANGELES COUNTY

CONTRIBUTIONS IN SCIENCE is a series of miscellaneous technical papers
in the fields of Biology, Geology and Anthropology, published at irregular intervals
by the Natural History Museum of Los Angeles County. Issues are numbered sep-
arately, and numbers run consecutively regardless of subject matter. Number 1 was
issued January 23, 1957. The series is available to scientific institutions and scien-
tists on an exchange basis. Copies may also be purchased at a nominal price. Inquiries
should be directed to Virginia D. Miller, Natural History Museum of Los Angeles
County, 900 Exposition Boulevard, Los Angeles, California 90007.

INSTRUCTIONS FOR AUTHORS

Manuscripts for CONTRIBUTIONS IN SCIENCE may be in any field of
Life or Earth Sciences. Acceptance of papers will be determined by the amount
and character of new information. Although priority will be given to manuscripts
by staff members, or to papers dealing largely with specimens in the collections of
the Museum, other technical papers will be considered. All manuscripts must be
recommended for consideration by the curator in charge of the proper section or
by the editorial board. Manuscripts must conform to those specifications listed
below and will be examined for suitability by an Editorial Committee including
review by competent specialists outside the Museum.

Authors proposing new taxa in a CONTRIBUTIONS IN SCIENCE must
indicate that the primary type has become the property of a scientific institution of
their choice and cited by name.

MANUSCRIPT FORM.— ( 1 ) The 1964 AIBS Style Manual for Biological
Journals is to be followed in preparation of copy. (2) Double space entire manu-
script. (3) Footnotes should be avoided if possible. Acknowledgments as footnotes
will not be accepted. (4) Place all tables on separate pages. (5) Figure legends and
unavoidable footnotes must be typed on separate sheets. Several of one kind may
be placed on a sheet. (6) An abstract must be included for all papers. This will be
published at the head of each paper. (7) A Spanish summary is required for all
manuscripts dealing with Latin American subjects. Summaries in other languages
are not required but are strongly recommended. Summaries will be published at
the end of the paper. (8) A diagnosis must accompany any newly proposed taxon.
(9) Submit two copies of manuscript.

ILLUSTRATIONS.— All illustrations, including maps and photographs, will
be referred to as figures. All illustrations should be of sufficient clarity and in the
proper proportions for reduction to CONTRIBUTIONS page size. Consult the 1964
AIBS Style Manual for Biological Journals in preparing illustration and legend copy
for style. Submit only illustrations made with permanent ink and glossy photo-
graphic prints of good contrast. Original illustrations and art work will be returned
after the manuscript has been published.

PROOF.— Authors will be sent galley proof which should be corrected and
returned promptly. Changes in the manuscript after galley proof will be billed to
the author. Unless otherwise requested, page proof also will be sent to the author.
One hundred copies of each paper will be given free to each author or divided
equally among multiple authors. Orders for additional copies must be sent to the
Editor at the time corrected galley proof is returned. Appropriate order forms will
be included with the galley proof.

Virginia D. Miller

Editor

THE STATUS OF LEPTODACTYLUS PUMILIO BOULENGER
(AMPHIBIA, LEPTOD ACTYLID AE ) AND THE DESCRIPTION OF A
NEW SPECIES OF LEPTODACTYLUS FROM ECUADOR1

By W. Ronald Heyer2

Abstract: Leptodactylus pumilio Boulenger, 1920, is
shown to be a junior synonym of Eleutherodactylus parvus
(Girard). The Pentadactylus species group of Leptodactylus is
redefined and a new species of this group is described from Ama-
zonian Ecuador. The presence of dorsolateral folds combined
with the uniformly black coloration of the posterior surface of
the thigh distinguish the new species from the other members of
the group. The karyotype of the new species has a diploid num-
ber of 22 bi-armed chromosomes with no secondary constric-
tions. A key to the species of the Pentadactylus group is provided.

Introduction

A preliminary analysis of a cross sectional representation of the genus
Leptodactylus indicated that the species could be grouped into five species
assemblages (Heyer, 1968). I am presently analyzing each of these groups
in detail (e.g.. Heyer, 1970). As in all long-term projects, data are gathered
continuously on all groups. The purpose of this paper is to report two findings
that are outside of my current main project. First, examination of the holotype
of Leptodactylus pumilio indicates a nomenclatural change is necessary. Sec-
ond, a new species of the Pentadactylus group is described from specimens
recently collected in Amazonian Ecuador.

Acknowledgments

Several people have helped in the research and preparation of this report.
Alice G. C. Grandison was a gracious hostess during my brief visit to the
British Museum (Natural History) (BMNH). Philip A. Silverstone, Natural
History Museum of Los Angeles County (LACM), kindly photographed the
type of Leptodactylus pumilio. Keith A. Berven, Pacific Lutheran University,
helped with the field work in Ecuador. Don Johnson, Director of the Summer
Institute of Linguistics in Ecuador, allowed us to undertake field work at their
institute base camp of Limoncocha during the summer of 1971. John W.

1 Editorial Committee for This Contribution

Robert L. Bezy

Roy W. McDiarmid

Ian R. Straughan

2 Research Associate, Section of Herpetology, Natural History Museum of Los
Angeles County; and Biology Department, Pacific Lutheran University, Tacoma,
Washington 98447.

1

2

Contributions in Science

No. 231

Wright, LACM, aided in the chromosome analysis and reviewed the manu-
script. Research support from NSF grant GB-27280 is gratefully acknowl-
edged.

Leptodactylus pumilio
Figure 1

In February of 1969, I had the opportunity to examine the type of
Leptodactylus pumilio at the British Museum (Natural History). The speci-
men was originally catalogued as 1914.3.20.7 but has been recatalogued as
1947.2.17.35. The salient features of the type (Fig. 1) are: 1) The sternum
has a cartilaginous plate; 2) Fingers III and IV have small disks, the toes
have large disks; 3) The finger and toe disks have peripheral grooves, the
upper surfaces are undivided; 4) The tympanum is not visible on the left,
barely visible on the right; 5) The tarsus is smooth; 6) There is a dark
triangular patch under the vent. Members of the genus Leptodactylus are
characterized in part by having a bony style in the sternum, disks (if present)
without peripheral grooves, and (usually) a tarsal fold. The holotype clearly

Figure 1. Dorsal {left) and ventral {right) views of holotype of Leptodactylus
pumilio (= Eleutherodactylus parvus), BMNH 1947.2.17.35, from Teresopolis,
Brasil.

1972

The Status of LEPTODACTYLUS PUMILIO

3

is not a member of the genus Leptodactylus, but of the genus Eleutherodac-
tylus. The holotype was collected in Teresopolis, Brasil, where fortunately,
few species of EJeutherodactylus occur. The dark seat patch is characteristic
of EJeutherodactylus parvus (Girard, 1853) and a comparison of the holotype
of L. pumilio with specimens of E. parvus in the collections of the British
Museum convinced me that they are conspecific. Leptodactylus pumilio
Boulenger is thus a junior synonym of EJeutherodactylus parvus (Girard).

The New Ecuadorian Species

During two months of field work in the upper Amazon basin, a series
of juvenile frogs of a new species of the genus Leptodactylus were collected.
With the exception of Leptodactylus laticeps, they are the most distinctively
colored species of Leptodactylus in life. As the species is so distinctive and
apparently has not been collected previously, I prefer to describe the new
species based on the juvenile specimens rather than await collection of adults.

The new species belongs to the Pentadactylus species group as provision-
ally defined earlier (Heyer, 1968). The group is in need of thorough revision
to determine the status of the L.. pentadactylus and L. pentadactylus-Yike
populations. In addition to the new species described below, the species group
consists of: L. laticeps Boulenger, 1918; L. pentadactylus (Laurenti) 1768
(probably a composite); L. rhodomystax Boulenger, 1883; L. rhodonotus
(Gunther), 1868;L. rugosus Noble, 1923; L. syphax Bokermann, 1969. Mem-
bers of this group have noticeable fringes on the toes as juveniles, but the
fringes are absent in adults. The adult character state of free toes separates
members of the Pentadactylus group from members of the Melanonotus and
Ocellatus groups which have extensive toe fringes as adults. Species of the
Marmoratus group are small, never exceeding 29 mm SV; species of the
Pentadactylus group are large, greater than 60 mm SV. The most distinctive
characteristic that separates members of the Pentadactylus group from the
Fuscus group is the presence of thumb spines and chest spines (usually) in
males of members of the Pentadactylus group. Male members of the Fuscus
group lack thumb and chest spines. Members of the Fuscus group are mod-
erate sized, only one species reaching 65 mm SV. Members of the Pentadac-
tylus group have broad, rounded snouts from above, members of the Fuscus
group have more pointed snouts.

For the new species I propose the name:

Leptodactylus knudseni, new species
Figure 2

Holotype— LACM 721 17, a juvenile female from Limoncocha, 0° 24'S,
76°37/W, Provincia de Napo, Ecuador. The specimen was collected in a
pasture, in a decaying log (15 cm diameter) at 14:38 hrs on 3 August 1970
by Keith A. Berven and W. Ronald Heyer. Elevation 260 m.

4

Contributions in Science

No. 231

Figure 2. Dorsal (left) and ventral (right) views of paratype of Leptodactylus
knudseni, LACM 72133, from Limoncocha, Provinica de Napo, Ecuador. Specimen
is 62.5 mm SV.

Topoparatypes — LACM 72118-149 (32 specimens), collected by Keith
A. Berven and W. Ronald Heyer between 7 June and 4 August 1971.

Diagnosis— In life, Leptodactylus knudseni is the only member of the
Pentadactylus group with prominent chartreuse markings on a black back-
ground. In preservative, L. knudseni can be recognized by the presence of a
pair of dorsolateral folds which differentiates it from L. laticeps, L. rugosus,
and L. syphax all of which lack dorsolateral folds. The posterior surface of
the thigh is uniformly black in L. knudseni, marbled in L. pentadactylus and
rhodonotus, and distinctly light spotted on a dark background in L. rhodomy-
stax.

Description of Holotype— Snout ovoid from above, rounded in profile;
canthus rostralis distinct; loreal concave; tympanum distinct, greatest diam-
eter 5/6 eye diameter; vomerine teeth in two arched series extending posterior
to choanae; finger lengths in order of decreasing size III > I > II = III, first
finger much longer than second; inner metacarpal tubercle large, ovoid,
smaller than heart-shaped outer metacarpal tubercle; dorsal surfaces sha-
greened, upper surface of tibia scattered with white tipped tubercles; one pair
of weak dorsolateral folds extending from eye to sacrum, another pair of folds
extending from posterior angle of eye over tympanum to angle of jaw, diffuse
gland at angle of jaw; ventral surfaces smooth, belly disk fold distinct; toe
tips not expanded; sides of toes with visible fringe, not extensively developed;
subarticular tubercles moderately developed; outer metatarsal tubercle dis-
tinct, rounded, about two-thirds length of elongate inner metatarsal tubercle,
tarsal fold distinct, extending 5/6 length of tarsus; no metatarsal fold; lower

1972

The Status of LEPTODACTYLUS PUM1LIO

5

surface of tarsus scattered with white tipped tubercles; sole of foot smooth
except for three or four white tipped tubercles on outermost edge of sole.

Measurements (in mm).— Snout-vent (SV), 63.2; head length, 22.9; head
width, 22.8; interorbital distance. 5.0; greatest diameter of typmanum, 4.8;
diameter of eye, 6.1; eye-nostril distance, 5.0; femur, 24.6; tibia, 27.4; foot,
30.8.

Coloration in preservative— Dorsal surfaces black with light gray pat-
terns, side of head light gray with dark triangles on upper lip, the dark triangle
under the eye extending to the eye; the light gray of the side of the head
bordering the lower half of the typmanum; tip of snout with light gray stripe
bifurcating at nostrils, extending along canthus rostralis, continuous with light
stripe on outer edge of eyelid and light interorbital bar; dorsum with light cross
bars, breaking down posteriorly; dorsolateral fold dark; upper arm with light
cross bars; upper femur and tarsus with irregular light cross bands; upper
tibia with light pattern surrounding dark central area; chin bordered with
alternating dark and light blotches; venter profused with melanophores scat-
tered with small light dots (visible under magnification, melanophores con-
tracted); bottom of tarsus and sole of foot black; posterior surface of thigh
uniform black.

Variation— The paratypes range in size from 32.8 to 62.5 mm. The
variation (minimum-mean-maximum ± 1 standard error) in measurement
ratios (expressed as per cent) among the type series is: head length/ snout-vent,
36-38.7-40 ± 1.0; head width/ snout-vent, 35-38.2-40 ± 1.4; femur/ snout-
vent, 40-43.0-46 ± 1.4; tibia/ snout-vent, 39-43.3-46 ± 1.6; foot/ snout-vent.
47-50.3-55 ± 2.0. The color pattern is similar among all the paratypes, the
greatest variation occurring in the degree of light marking on the dorsum in
the sacral region. In several specimens the melanophores are expanded on the
belly, producing a black belly with small light dots.

The color in life of specimen LACM 72118 was typical of other speci-
mens in the type series: posterior surface of thigh jet black; upper surfaces of
legs with barely discernible yellowish green cross bands; belly gray with lighter
punctations; chin with yellow marks along edge; dorsum with greenish yellow
bands enclosing brownish green areas which are black bordered; iris gold-
yellow above, rusty gold below; head mostly yellowish green.

Karyotype.— Twenty-four cells were examined from marrow and spleen
tissue of specimens 72145, 72147, and 72148. The slides will be deposited in
LACM. The terminology used is that defined by Patton (1967). Several
chromosomes are borderline in their classification and vary according to their
state of contraction. Three pairs of metacentrics (Fig. 3, chromosome pair
numbers 1, 4, 9), 4 pairs of submetacentrics (Fig. 3, numbers 2, 5, 10, 11),
and 4 pairs of subtelocentrics (Fig. 3, numbers 3, 6, 7, 8) are common. The
Fundamental Number is 44; there are no secondary constrictions. An analysis
of the karyotypic variation found within the genus is in progress and will be
reported on separately. Preliminary results indicate that the karyotype of

6

Contributions in Science

No. 231

L. knudseni is similar to the karyotypes of other members of the Pentadactylus
and Ocellatus groups.

Ecology — Two individuals were taken from a selectively logged secon-
dary forest. The primary forest at Limoncocha is Tropical Moist Forest
according to Holdridge’s classification (1964). The other specimens were
collected in a pasture (Fig. 4). All specimens were taken from under cover
during the day: one from bark, five from under boards, 21 from under logs
ranging in diameter from 15 to 70 cm, five from within rotten logs ranging in
diameter from 15 to 30 cm. Other species of Leptodactylus collected in sym-
patry with L. knudseni at Limoncocha were L. discodactylus, mystaceus,
pentadactylus, and wagneri. Further ecological aspects of the five sympatric
Leptodactylus will be reported in a later paper by Heyer and Beilin.

Etymology The new species is named for Dr. Jens W. Knudsen, who
was the most important influence in my decision to be a professional biologist,
and who continues to encourage my research efforts.

Remarks —Leptodactylus knudseni raises the number of recognized
species from Ecuador to 10. The other nine species as summarized by Heyer
and Peters (1971) are: Leptodactylus discodactylus, hylaedactylus, labrosus,

ft* A|

5 6

II M

3 4

K 4*** m >.

7 8

■ft 4L

m “

9 10

m j§

Vljp PlfP

II

Figure 3. Karyotype of Leptodactylus knudseni. Marrow and spleen preparation
from LACM 72147.

1972

The Status of LEPTODACTYLUS PUMILIO

1

Figure 4. Pasture habitat at Limoncocha where most specimens of Leptodactylus
knudseni were collected. Note selectively logged secondary forest in background.

melanonotus, mystaceus, pentadactylus, rhodomystax, ventrimaculatus, and
wagneri.

Specimens of Leptodactylus knudseni will key out to couplet 5 in Heyer
and Peters (1971 : 169). The uniformly colored posterior surface of the thigh
of L. knudseni will separate it from the variously patterned posterior thigh
surfaces of L. mystaceus, hylaedactylus, and ventrimaculatus.

A Preliminary Key to the Species of the Pentadactylus Group

IA. Dorsal pattern of large discrete dark spots on a

lighter background (Argentina) L. laticeps

IB. Dorsal pattern variable, never with distinct spots 2

2A. Dorsolateral folds lacking 3

2B. A pair of dorsolateral folds 4

3 A. Dorsum very rugose; males with a single thumb

spine (Guayana shield) L. rugosus

3B. Dorsum warty, not rugose; males with two thumb

spines (Brasil, Mato Grosso) L. sryphax

8

Contributions in Science

No. 231

4 A. Posterior surface of thigh uniform (Ecuador) . . . . L. knudseni

4B. Posterior surface of thigh patterned 5

5A. Posterior surface of thigh dark with discrete

light spots (northern South America) L. rhodomystax

5B. Posterior surface of thigh marbled, never with

distinct light spots 6

6A. Large, adults to 160 mm; males usually with

a single thumb spine (widespread) ............ L. pentadactylus

6B. Moderately large, adults to 80 mm; males with

two thumb spines (Peru) L. rhodonotus

Resumen

Se demuestra que Leptodactylus pumilio Boulenger, 1920, es un sinonimo
menor de Eleutherodactylus parvus (Girard). La especie Pentadactylus grupo
de Leptodactylus es redefinida y una nueva especie de este grupo del Ecuador
Amazonico es descrita. La presencia de pliegues dorsolaterales combinada con
la uniforme coloracion negra de la superficie posterior del muslo, distingue a
la nueva especie de los otros miembros del grupo. El cariotipo de la nueva
especie tiene un numero diploide de 22 cromosomas birrameos sin constric-
ciones secundarias. Se proporciona una clave para las especies del grupo
Pentadactylus.

Literature Cited

Boulenger, G. A. 1920. Descriptions of two new frogs from Brazil. Ann. Mag.
Nat. Hist. 9(5): 122-124.

Girard, C. F. 1853. Descriptions of new species of reptiles, collected by the U.S.
Exploring Expedition, under the command of Capt. Charles Wilkes, U.S.N.
Second part.— Including the species of batrachians, exotic to North America.
Proc. Acad. Nat. Sci. Phila. 6:420-424.

Heyer, W. R. 1968. Biosystematic studies on the frog genus Leptodactylus. Ph.D.
Dissertation, Univ. So. Calif. 234 p.

1970. Studies on frogs of the genus Leptodactylus (Amphibia, Lepto-

dactylidae). VI. Biosystematics of the Melanonotus group. Los Angeles Co.
Mus., Contrib. Sci. 191:1-48.

, and J. A. Peters. 1971. The frog genus Leptodactylus in Ecuador.

Proc. Biol. Soc. Wash. 84(19) : 163-170.

Holdridge, L. R. 1964. Life zone ecology. Tropical Science Center, San Jose,
Costa Rica. 124 p.

Patton, J. L. 1967. Chromosome studies of certain pocket mice, genus Perognathus
(Rodentia: Heteromyidae). J. Mammal. 48:27-37.

Accepted for publication April 17, 1972

■ •t

S 0 7, 7 3

CzLiCi

NUMBER 232
OCTOBER 17, 1972

PRELIMINARY REPORT ON LATE
CRETACEOUS MAMMALS FROM THE
EL GALLO FORMATION,
BAJA CALIFORNIA DEL NORTE, MEXICO

By Jason A. Lillegraven

CONTRIBUTIONS IN SCIENCE

NATURAL HISTORY MUSEUM • LOS ANGELES COUNTY

CONTRIBUTIONS IN SCIENCE is a series of miscellaneous technical papers
in the fields of Biology, Geology and Anthropology, published at irregular intervals
by the Natural History Museum of Los Angeles County. Issues are numbered sep-
arately, and numbers run consecutively regardless of subject matter. Number 1 was
issued January 23, 1957. The series is available to scientific institutions and scien-
tists on an exchange basis. Copies may also be purchased at a nominal price.
Inquiries should be directed to Virginia D. Miller, Natural History Museum of Los
Angeles County, 900 Exposition Boulevard, Los Angeles, California 90007.

INSTRUCTIONS FOR AUTHORS

Manuscripts for CONTRIBUTIONS IN SCIENCE may be in any field of Life
or Earth Sciences. Acceptance of papers will be determined by the amount and char-
acter of new information. Although priority will be given to manuscripts by staff
members, or to papers dealing largely with specimens in the collections of the Muse-
um, other technical papers will be considered. All manuscripts must be recommend-
ed for consideration by the curator in charge of the proper section or by the editorial
board. Manuscripts must conform to those specifications listed below and will be ex-
amined for suitability by an Editorial Committee including review by competent
specialists outside the Museum.

Authors proposing new taxa in a CONTRIBUTIONS IN SCIENCE must indi-
cate that the primary type has become the property of a scientific institution of their
choice and cited by name.

MANUSCRIPT FORM.— (1) The 1972 CBE Style Manual, third edition
(AIBS) is to be followed in preparation of copy. (2) Double space entire manuscript.
(3) Footnotes should be avoided if possible. Acknowledgments as footnotes will not
be accepted. (4) Place all tables on separate pages. (5) Figure legends and unavoid-
able footnotes must be typed on separate sheets. Several of one kind may be placed on
a sheet. (6) An abstract must be included for all papers. This will be published at the
head of each paper. (7) A Spanish summary is required for all manuscripts dealing
with Latin American subjects. Summaries in other languages are not required but are
strongly recommended. Summaries will be published at the end of the paper. (8) A
diagnosis must accompany any newly proposed taxon. (9) Submit two copies of
manuscript.

ILLUSTRATIONS. — All illustrations, including maps and photographs, will be
referred to as figures. All illustrations should be of sufficient clarity and in the proper
proportions for reduction to CONTRIBUTIONS page size. Consult the 1972 CBE
Style Manual, third edition (AIBS) in preparing illustration and legend copy for
style. Submit only illustrations made with permanent ink and glossy photographic
prints of good contrast. Original illustrations and art work will be returned after the
manuscript has been published.

PROOF. — Authors will be sent galley proof which should be corrected and
returned promptly. Any changes or alterations, other than typographical corrections,
will be billed to the author. Unless otherwise requested, page proof also will be sent
to the author. One hundred copies of each paper will be given free to each author or
divided equally among multiple authors. Orders for additional copies must be sent
to the Editor at the time corrected galley proof is returned. Appropriate order forms
will be included with the galley proof.

Virginia D. Miller
Editor

PRELIMINARY REPORT ON LATE CRETACEOUS MAMMALS
FROM THE EL GALLO FORMATION,

BAJA CALIFORNIA DEL NORTE, MEXICO1

By Jason A. Lillegraven2

Abstract: A preliminary study of the mammalian fossils
from the late Campanian (Late Cretaceous) “El Gallo Forma-
tion” west of El Rosario, Baja California del Norte, Mexico
suggests the presence of Mesodma, cf. M . formosa (Ectypodonti-
dae, Multituberculata), ?Stygimys sp., species probably new
(Eucosmodontidae, Multituberculata), Pediomys sp., species
probably new (Pediomyidae, Marsupialia), and a new genus of
indefinite familial affinities (Insectivora). The sample provides
the first knowledge of Mesozoic mammals from the west coast
of North America. Despite taxonomic differences from the
distant and better known mammalian local faunas of the Rocky
Mountain region, the composition of the El Gallo assemblage is
basically similar to taxa found in the Western Interior and does
not suggest a profound endemism.

Introduction

Prior to the beginning of the present study, Mesozoic mammals from
North America were unknown from rocks west of the Rocky Mountains.
Field workers in the summer of 1968 under the direction of Dr. William J.
Morris discovered remains of multituberculate mammals in the “El Gallo
Formation,” Baja California del Norte, Mexico (Fig. 1). Fossils of therian
mammals were discovered in the summer of 1970 by members of another
field party working under my supervision. The present paper is a preliminary
report to the scientific community of these significant finds. Hopefully, future
collecting and study will result in a monographic treatment of the potentially
extremely important mammalian local fauna.

The “El Gallo Formation” is thought to be middle to late Campanian in
age (see Morris, 1967: 1539). An unpublished potassium-argon date of
approximately 73 million years is now available from a tuff in the lower one-
third of the “formation” (Morris, personal communication). Although
principally nonmarine in origin, a small number of marine interbeds are
known. Rocks are well exposed in deeply dissected badlands and dip approxi-
mately 10° regularly to the northeast. Only a small percentage of the exposed

1 Editorial Committee for this Contribution

William A. Clemens
Richard C. Fox
David P. Whistler

2 Research Associate, Section of Vertebrate Paleontology, Natural History Museum
of Los Angeles County; and Department of Zoology, San Diego State University,
San Diego, California 921 15.

1

2

Contributions in Science

No. 232

area is accessible to motor vehicles. Depositional facies shift from dominantly
conglomeratic near the basement source area approximately six miles to the
east of the fossiliferous area to complex but generally finer clastic facies in
the west. Petrified wood is common throughout. The “El Gallo Formation”
was proposed by Kilmer in his unpublished Ph.D. dissertation (1963). No
description or type section has ever been published, nor is the dissertation
available through the microfilm services. Thus, until described in publication
form, the unit must be considered informally.

Localities

Known fish-, amphibian-, lizard-, and dinosaur-bearing localities from
the “El Gallo Formation” are numerous and are recorded in the files of the
Vertebrate Paleontology Section, Natural History Museum of Los Angeles
County. Although dinosaur bones may be expected in nearly any sedimentary

1972

Late Cretaceous Mammals

3

rock in the area, it is interesting and important to note that all small-vertebrate
localities discovered to date are in gray to black silty claystone beds. For
unknown reasons, fossils of microvertebrates are rare in the paler and coarser-
textured rocks.

Fossil mammals have been discovered at three localities at various
stratigraphic levels, all of which are located within the middle one-third of
the formation, well above the dated tuff. Because there are no surveyed maps
of the area, localities were plotted on aerial photographs and described in field
notes augmented by Polaroid photos. These are also on file in the Natural
Flistory Museum of Los Angeles County. The numbers of the mammal-bearing
localities are: LAV-7 1 70, LAV-7 171, LAV-7 1 72.

Methods

Known productive layers were quarried by breaking rocks into walnut-
sized or smaller pieces while looking for freshly exposed bones. Bone-bearing
clods were wrapped for later preparation in the laboratory. The others lacking
exposed bones were dropped into burlap bags and carried to a soaking-tub
at camp. Rocks were soaked 24 hours in kerosene, which was then siphoned
off and replaced by water. The extent of breakdown of the rocks was then
dramatic, for it was practically nonexistent after soaking only in water. The
resultant mud was scooped into a table height large stacked screen-box system.
The upper seive was of standard gauge window screening, and the lower of
40 wires per inch bronze wire cloth. A small pump carried brackish water
from a lagoon via a hose to above the mud, and gentle spraying washed away
all but a concentrate of rocks and skeletal fragments. The concentrate was
air dried on long burlap strips and sorted for fossils at a later time. Although
considerable care to avoid breakage of the fossils was exercised through all
stages of the washing process, I feel the “washing table” technique used is less
desirable than others. It is necessary to scoop the fossil-bearing mud from the
soaking-tub onto the screens, and even the gentlest of water spraying is
damaging to small delicate bones and teeth. I feel that the quarry matrix
should be soaked, washed, and dried in the same screen box and that during
washing the fossil-bearing mud should be agitated gently while completely
submerged in water.

All measurements (which follow the citation of individual teeth in the
various “Referred or available specimens” sections) are in millimeters and
were made with an EPOI Shopscope at San Diego State University using the
orientations specified by Lillegraven (1969: 16). Abbreviations of measure-
ments are as follows:

A-P Antero-posterior length Post-W Posterior width

W Greatest width W-Tri Width of trigonid

Ant-W Anterior width W-Tal Width of talonid

4

Contributions in Science

No. 232

The abbreviation LACM is used throughout the descriptions to indicate
specimens cataloged in the Natural History Museum of Los Angeles County
(Vertebrate Paleontology).

Systematic Descriptions
Class Mammalia
Subclass Allotheria
Order Multituberculata

Family Ectypodontidae Sloan and Van Valen, 1965
[original name emended. Van Valen and Sloan, 1966]

Genus Mesodma Jepsen, 1940
Mesodma, cf. M. formosa (Marsh), 1889b

Holotype: Yale Peabody Museum 11812, left P4 (Marsh, 1889b, pi. 8,
fig. 36-39).

Referred specimens: LACM 27588, M1 fragment (W 1.21); LACM
27589, M2 (A-P 1.33, W 1.33); LACM 27590, M2 (A-P 1.38, W 1.26).

Localities: LAV-7 1 70 and LAV-7 171.

Distribution of Mesodma formosa: Upper part of Edmonton Formation,
Alberta; Hell Creek Formation, Montana, and South Dakota; type Lance
Formation, Wyoming; possibly Kirtland and Fruitland Formations, New
Mexico (see list, Fassett and Hinds, 1971: 19); “El Gallo Formation,” Baja
California del Norte.

Comments: The three specimens here identified as Mesodma formosa
cannot be distinguished from specimens of that taxon from Upper Cretaceous
deposits of the Rocky Mountain region. The size ranges and descriptions
follow closely and, until more evidence is forthcoming, I feel no new names
should be defined despite the differences in ages of deposits. Because of the
lack of equivalent dental elements, comparison is impossible with M. senecta
(see Fox, 1971a).

Family Eucosmodontidae (Jepsen, 1940)

Genus Stygimys Sloan and Van Valen, 1965

? Stygimys sp.

Referred specimens: LACM 27591, M2 (A-P 1.96, W 1 .70) (Figure 2, A.,
B., C.); LACM 27592, U.

Localities: LAV-7170 and LAV-7172.

Distribution of genus: Hell Creek Formation, Montana; various lower
and middle Paleocene localities. Rocky Mountain region; “El Gallo Forma-
tion,” Baja California del Norte.

Comments: The M2 here referred to IStygimys sp. (Fig. 2) is strikingly
similar in overall morphology with specimens (e.g., LACM 27593) of Stygimys
kuszmauli from the Bug Creek Anthills Local Fauna of the Hell Creek
Formation, Montana (Sloan and Van Valen, 1965). However, distinct dif-
ferences do exist in that the El Gallo specimen is significantly smaller than

1972

Late Cretaceous Mammals

5

Figure 2. A.-C. 1 Stygimys sp., species probably new, LACM 27591 (LAV-7170)
right M2: A. lingual view; B. occlusal view; C. labial view. Approximately 13X.

the Hell Creek material (the latter are approximately A-P 2.6, W 2.3). Also,
the El Gallo specimen has smoother cusp sides with much less ornamentation
than the usual Hell Creek M2’s referred to Stygimys. The greater epi- and
inter-cusp ornamentation on the Hell Creek specimens is quite possibly a
specialization advanced from the primitive condition.

Additional, but admittedly weak, evidence suggesting the assignment of
the M2 to Stygimys, or at least the Taeniolabidoidea, is a fragmentary lower
incisor from Locality LAV-7172 of the “El Gallo Formation” showing a
distinct rodentlike eucosmodontid enamel distribution. Stygimys and
Catopsalis are the only known North American Cretaceous genera possessing
the eucosmodontid-type incisor enamel pattern (see Sloan and Van Valen,
1965:224). Eucosmodontid multituberculate teeth have recently been recov-
ered from the Upper Cretaceous Kirtland and Fruitland formations of New
Mexico (see list, Fassett and Hinds, 1971: 19).

Although I consider the El Gallo specimen to be distinct at least at the
specific level from previously described material, I believe it would be wise to
refrain from adding a new name to the taxonomic literature until a larger
sample is available to allow the writing of a secure diagnosis.

Additional Multituberculate Teeth

Four other multituberculate teeth have been recovered that should be
mentioned but are, in my opinion, unidentifiable generically at the present
time. LACM 27594 (LAV-7172) is an isolated P3 (A-P 1.19, W 1.02) with
four cusps. It resembles in basic structure the P3 of Mesodma formosa
illustrated by Lillegraven (1969:22, Fig. 8, 2) but is proportionately shorter
anteroposteriorly. Two P4’s (LACM 27595 from LAV-7170 and 27596 from
LAV-7172) are represented by the posterior halves only, and no possibility
exists for making serration counts or lobe descriptions. Finally, a fragmentary
questionable lower right incisor tip (LACM 27597 from LAV-7172) has been
recovered. Little can be said about it except that it lacks a eucosmodontid
enamel pattern but has a deep longitudinal trough along what I interpret to be
the dorsolateral surface of the tooth.

6

Contributions in Science

No. 232

Subclass Theria
Infraclass Metatheria
Order Marsupialia
Family Pediomyidae Clemens, 1966
Genus Pediomys Marsh, 1889a
Pediomys sp.

Only available specimen: LACM 27598, M1 (A-P 1.67 [est.] , Ant-W
1.44, Post-W 1.61) (Figure 3, A., B.).

Locality : LAV-7 172.

Distribution of genus: "El Gallo Formation,” Baja California del Norte;
Oldman (unpublished) and Milk River formations, Alberta; Judith River
Formation, Montana (in press, A. Sahni); upper part of Edmonton Formation,
Alberta; Hell Creek Formation, Montana and South Dakota; Lance Formation,
Wyoming; North Horn Formation, Utah; possibly Kirtland and Fruitland
Formations, New Mexico (see list, Fassett and Hinds, 1971:19).

Figure 3. A.-B. Pediomys sp., species probably new, LACM 27598 (LAV-7172),
left M1: A. labial view; B. occlusal view. Approximately 17X.

Comments: The M1 (Figure 3) almost certainly represents a heretofore
undescribed small species of the genus Pediomys. The overall morphology
most closely resembles the teeth of P. elegans (see Clemens, 1966), but a series
of significant differences exist. The transverse measurements of LACM 27598
are proportionately less than in any other pediomyid save P. exiguus of the
Milk River Formation of Alberta (see Fox, 1971b). As in most specimens of
P. elegans, stylar cusp B and the stylar shelf labial to the paracone are lacking,
but, in contrast, stylar cusp C in LACM 27598 is slightly larger and more
robust than stylar cusp D. Stylar cusp D in LACM 27598 is long, low, and
anteroposteriorly twinned, giving the appearance of beading. Wear facets
are observed both on stylar cusps C and D. Unfortunately, the anterolabial
corner of LACM 27598 is broken away. Lingual cingula are lacking. Details
of the protocone, conules, paracone, and metacone of LACM 27598 are
similar to those of the teeth of P. elegans (see Clemens, 1966:37). Wear facets
are illustrated in Figure 3 and indicate strong shearing function along the
postmetacrista in the usual therian manner. The tooth is strongly three-rooted
and shows numerous differences from teeth identified as marsupial DP3’s

1972

Late Cretaceous Mammals

7

(e.g., Clemens, 1966, fig. 30 and Lillegraven, 1969, fig. 23, 5). I believe the
tooth to be part of the molar series.

Although the size and general proportions of LACM 27598 are near
those of specimens of Pediomys exiguus, I believe different species are
represented. P. exiguus possesses a stylar cusp B, although reduced, usually
lacks a stylar cusp C, and has an undivided bladelike stylar cusp D (Fox,
1971b: 153).

I can see no serious objections to allying the species represented by
LACM 27598 closely with Pediomys elegans. I would not unite them as the
same species because significant differences in morphology, in geochrono-
logic age, and in geographic location exist. On the other hand, the probable
new species is known from only a single upper molar, and I consider it prudent
to wait until the hypodigm increases before entering a new specific name into
the taxonomic literature.

Infraclass Eutheria
Order Insectivora
Family indefinite
New genus

Available specimens: LACM 27599, M2 (A-P 2.57, W-Tri 1.58, W-Tal
1.65) (Figure 4, A., B., C.); LACM 27600, mandibular fragment with talonid
of Mi (W-Tal 1.48), M2 (A-P 2.29, W-Tri 1.50, W-Tal 1.46 [est.]), M3 (A-P
2.43, W-Tri 1.45, W-Tal 1.33) (Figure 5, A., B., C.).

Locality: LAV-7172.

Distribution: Known only from “El Gallo Formation,” Baja California
del Norte.

Descriptions: LACM 27599 (Figure 4) is tentatively considered to be M2
and is essentially unworn. The tooth is fairly robust in general construction
with a moderately high crowned trigonid. The protoconid is considerably
higher than the metaconid. The paraconid is low, somewhat anteriorly-
projecting, and anteroposteriorly compressed. The paraconid is well separated

Figure 4. A.-C. Insectivora, family indefinite, new genus, LACM 27599 (LAV-7172),
right M2: A. labial view; B. occlusal view; C. lingual view. Approximately 10X.

8

Contributions in Science

No. 232

from the metaconid and is labial to it. A short, weak cingulum is present at
the anterior base of the protoconid. The talonid is slightly wider than the
trigonid and the talonid cusps are well defined and separated. The talonid
basin is deeply concave. The hypoconulid is close to the entoconid but cannot
be strictly said to be twinned with it. The hypoconid is the highest talonid
cusp, the hypoconulid the lowest. A cingulum descends steeply from the
labial edge of the hypoconulid lateroventrally to the posterolabial base of the
hypoconid. The cristid obliqua meets the posterior midline of the base of the
protoconid.

LACM 27600 (Figure 5) is a mandibular fragment with a last molar and
two preceding molars. These are tentatively identified as M1-3. All three teeth
are severely worn and Mi is represented only by the posterior margin of the
talonid. Cusps are distinguishable as individual units only on M3 but the
general cusp arrangement appears to have been similar to that on LACM
27599. Cingula on the available parts of all three teeth are as on LACM 27599.

Figure 5. A.-C. Insectivora, family indefinite, new genus, LACM 27600 (LAV-7172),
left mandibular fragment with Mi-3: A. lingual view; B. occlusal view; C labial
view. Approximately 9X.

1972

Late Cretaceous Mammals

9

The hypoconulid of the M3 is more strongly produced posteriorly than in
any other Cretaceous therian known to me, and the proportional width of the
talonid of the M3 is considerably greater than in most Cretaceous therians.

Comments: LACM 27599 and 27600 are identified as eutherian because:
(1) the paraconid is proportionately small and labial !y placed; (2) the hypo-
conulid on the M3 is strongly produced posteriorly; (3) the hypoconulid is
not closely twinned with the entoconid. The combination of these features is
common among placental mammals but rare among marsupials. The signifi-
cance of the position of the hypoconulid (criterion “3” above) is not completely
certain. As stated in the description, it is not as closely twinned with the
entoconid as in most Mesozoic marsupials, yet is nearer the entoconid than
in most Cretaceous eutherians yet described.

The affinity of the new species at lower categorical levels is uncertain. I
placed the species in the Order Insectivora as an act of conservatism. Familial
relationships are totally obscure. The specimens suggest somewhat greater
similarity with molars of Late Cretaceous palaeoryctid taxa (e.g., Cimolestes
magnus, see Lillegraven, 1969) than with other known Cretaceous groups, but
evidence is insufficient to warrant even a tentative familial assignment. Con-
siderable similarity in general form and size also exists with the “ Champ -
Garimond tooth” discovered in Upper Cretaceous rocks of France (Ledoux
et al., 1966). The identification of that specimen, however, has also yet to be
determined (see McKenna, 1969:228). No striking resemblances have been
recognized with described Asiatic Cretaceous eutherians (e.g., see Rician-
Jaworowska, 1968) or the one upper molar described by Fox (1970) from the
Milk River Formation of Alberta, Canada. Although a strongly developed
hypoconulid on the M3 is a feature common to most early primates, the El
Gallo specimens have unusually broad and elongated talonids, paraconids
well separated from the metaconids, and other features decidedly different
from the most primitive known primates.

Both specimens are very tentatively referred to the same genus, and
perhaps they even represent the same species. The specimens are unquestion-
ably representative of a previously unknown genus. However, because of the
scanty material at hand, I have declined to name the taxon. Despite any
nomenclatorial inconvenience that may be caused, I feel it prudent to wait
until adequate reference material becomes available from future field work.

Discussion

The known El Gallo specimens give a tantalizing but misty first glimpse
of the Late Cretaceous mammalian fauna of the West Coast of North America.
Although all but one of the species discovered so far are probably new, most
of the genera seem referable to those well known from the Rocky Mountain
region. The El Gallo peri- Pacific Late Cretaceous collection represents
ecological, geographical, and temporal settings previously unsampled. One
would thus expect taxonomic differences from the distant and better known

10

Contributions in Science

No. 232

Rocky Mountain assemblages and, indeed, they have been found. However,
the known assemblage of El Gallo mammals does not suggest great and
profound endemism of the Late Cretaceous fauna of Baja California. At the
present stage of our knowledge, I see no particular reason to suggest geo-
graphic isolation of the peninsula from the remainder of the continent.

Acknowledgments

The entire project was visualized, implemented, and supervised by
Dr. William J. Morris, Department of Geology, Occidental College. Field
and laboratory support was generously given by the Vertebrate Paleontology
Section, Natural History Museum of Los Angeles County and by a continuing
grant from the National Geographic Society. Dr. Ismael Ferrusquia V. of the
Instituto de Geologia, Universidad Nacional Autonoma de Mexico participated
in the field work in the summer of 1970, personally found the therian mandible,
and contributed in many ways to the success of the summer. Dr. Ferrusquia
is currently pursuing further investigations on the El Gallo fauna of small
vertebrates. The devoted efforts and friendship of the family of Sr. Pedro
Fonseca of El Rosario are deeply appreciated. A number of field assistants
added greatly to the success of the expeditions. Principal among these were
Messrs. Alan Tabrum, Gregg Franz, Richard Bergreen, and Bruce Burns.
Illustrations for the paper were prepared by Miss Linda Thompson.

Thanks also go to Drs. Donald E. Savage and William A. Clemens, Jr.
of the University of California, Berkeley, and to Dr. Richard C. Fox of the
University of Alberta for reading the manuscript and suggesting changes.
My wife, Bernice Ann Lillegraven, was helpful in many aspects of the prepara-
tion of the manuscript.

Most importantly, the governments of Mexico and Baja California and
the citizens of El Rosario are gratefully acknowledged for their generous and
concerned interest shown in the project.

Resumen

Un estudio preliminar de los aprovechables fosiles de mamfferos de la
“formacion El Gallo” del ultimo Campaniano (ultimo Cretaceo) al oueste de
El Rosario, Baja California del Norte, Mexico, sugiere la presencia de
Mesodma, cf. M. formosa (Ectypodontidae, Multituberculata), IStygimys
sp., especie probablemente nueva (Eucosmodontidae, Multituberculata),
Pediomys sp., especie probablemente nueva (Pediomyidae, Marsupialia), y
uno genero nuevo de incierto afinidad de la familia (Insectfvora). Esta muestra
representa el primero conocimiento de los mamfferos de la Secundaria de la
costa oeste de Norte America. A despecho de las diferencias taxonomicas de
las distantes y mas conocidas faunas de mamfferos de la region de Montanas
Roquenas, la composicion de la coleccion de El Gallo es fundamentalmente
semejante a taxa que es hallada an el interior del oeste y no sugiere una endemia
profunda.

1972

Late Cretaceous Mammals

1 1

Literature Cited

Clemens, W. A., Jr. 1966. Fossil mammals of the type Lance Formation, Wyoming.

Part II. Marsupialia. Univ. Calif. Publ. Geol. Sci. 62. v/+ 122 p.

Fassett, J. E., and J. S. Hinds. 1971. Geology and fuel resources of the Fruitland
Formation and Kirtland Shale of the San Juan Basin, New Mexico and Colorado.
U. S. Geol. Surv. Prof. Pap. 676. 76 p.

Fox, R. C. 1970. Eutherian mammal from the early Campanian (Late Cretaceous)
of Alberta, Canada. Nature 227:630-631.

1971a. Early Campanian multituberculates (Mammalia: Allotheria) from

the Upper Milk River Formation, Alberta. Canadian J. Earth Sci. 8: 916-938.

1971b. Marsupial mammals from the early Campanian Milk River

Formation, Alberta, Canada. In (D. M. Kermack and K. A. Kermack, eds.)
Early Mammals, Zool. J. Linnean Soc. 50, Suppl. 1: 145-164.

Jepsen, G. L. 1940. Paleocene faunas of the Polecat Bench Formation, Park County,
Wyoming. Am. Philos. Soc., Proc. 83: 217-340.

Kielan-Jaworowska, Z. 1968. Results of the Polish-Mongolian Palaeontological
Expeditions, Part I. Preliminary data on the Upper Cretaceous eutherian
mammals from Bayn Dzak, Gobi Desert. Palaeontologia Polonica 19: 171-191.
Kilmer, F. H. 1963. Cretaceous and Cenozoic stratigraphy and paleontology,
El Rosario area, Baja California, Mexico. Unpublished Ph.D. dissertation,
Univ. Calif., Berkeley.

Ledoux, J.-C., J.-L. Hartenberger, J. Michaux, J. Sudre, and L. Thaler. 1966.
Decouverte d’un mammifere dans le Cretace superieur a dinosaures de Champ-
Garimond pres de Fons (Gard). Comptes rendus des seances de FAcademie
des Sciences, Paris 262: 1925-1928.

Lillegraven, J. A. 1969. Latest Cretaceous mammals of upper part of Edmonton
Formation of Alberta, Canada, and review of marsupial-placental dichotomy
in mammalian evolution. Univ. Kansas. Paleontol. Contrib., Art. 50 (Verte-
brata 12). 122 p.

Marsh, O. C. 1889a. Discovery of Cretaceous Mammalia: Am. J. Sci., ser. 3, 38:
81-92.

1889b. Discovery of Cretaceous Mammalia, Part 2. Am. J. Sci., ser. 3,

38: 177-180.

McKenna, M. C. 1969. The origin and early differentiation of therian mammals.
Ann. N. Y. Acad. Sci. 167: 217-240.

Morris, W. J. 1967. Baja California: Late Cretaceous dinosaurs. Science 155: 1539-
1541.

Sloan, R. E., and L. Van Valen. 1965. Cretaceous mammals from Montana.
Science 148: 220-227.

Van Valen, L., and R. E. Sloan. 1966. The extinction of the multituberculates.
Syst. Zool. 15: 261-278.

Accepted for publication April 4, 1972

5'cfJ. I3

NUMBER 233
OCTOBER 17, 1972

A NEW GENUS OF CYPRINODONTID
FISH FROM NUEVO LEON, MEXICO

By Robert Rush Miller and Vladimir Walters

CONTRIBUTIONS IN SCIENCE

NATURAL HISTORY MUSEUM • LOS ANGELES COUNTY

CONTRIBUTIONS IN SCIENCE is a series of miscellaneous technical papers
in the fields of Biology, Geology and Anthropology, published at irregular intervals
by the Natural History Museum of Los Angeles County. Issues are numbered sep-
arately, and numbers run consecutively regardless of subject matter. Number 1 was
issued January 23, 1957. The series is available to scientific institutions and scien-
tists on an exchange basis. Copies may also be purchased at a nominal price.
Inquiries should be directed to Virginia D. Miller, Natural History Museum of Los
Angeles County, 900 Exposition Boulevard, Los Angeles, California 90007.

INSTRUCTIONS FOR AUTHORS

Manuscripts for CONTRIBUTIONS IN SCIENCE may be in any field of Life
or Earth Sciences. Acceptance of papers will be determined by the amount and char-
acter of new information. Although priority will be given to manuscripts by staff
members, or to papers dealing largely with specimens in the collections of the Muse-
um, other technical papers will be considered. All manuscripts must be recommend-
ed for consideration by the curator in charge of the proper section or by the editorial
board. Manuscripts must conform to those specifications listed below and will be ex-
amined for suitability by an Editorial Committee including review by competent
specialists outside the Museum.

Authors proposing new taxa in a CONTRIBUTIONS IN SCIENCE must indi-
cate that the primary type has become the property of a scientific institution of their
choice and cited by name.

MANUSCRIPT FORM.— (1) The 1972 CBE Style Manual, third edition
(AIBS) is to be followed in preparation of copy. (2) Double space entire manuscript.
(3) Footnotes should be avoided if possible. Acknowledgments as footnotes will not
be accepted. (4) Place all tables on separate pages. (5) Figure legends and unavoid-
able footnotes must be typed on separate sheets. Several of one kind may be placed on
a sheet. (6) An abstract must be included for all papers. This will be published at the
head of each paper. (7) A Spanish summary is required for all manuscripts dealing
with Latin American subjects. Summaries in other languages are not required but are
strongly recommended. Summaries will be published at the end of the paper. (8) A
diagnosis must accompany any newly proposed taxon. (9) Submit two copies of
manuscript.

ILLUSTRATIONS. — All illustrations, including maps and photographs, will be
referred to as figures. All illustrations should be of sufficient clarity and in the proper
proportions for reduction to CONTRIBUTIONS page size. Consult the 1972 CBE
Style Manual, third edition (AIBS) in preparing illustration and legend copy for
style. Submit only illustrations made with permanent ink and glossy photographic
prints of good contrast. Original illustrations and art work will be returned after the
manuscript has been published.

PROOF. — Authors will be sent galley proof which should be corrected and
returned promptly. Any changes or alterations, other than typographical corrections,
will be billed to the author. Unless otherwise requested, page proof also will be sent
to the author. One hundred copies of each paper will be given free to each author or
divided equally among multiple authors. Orders for additional copies must be sent
to the Editor at the time corrected galley proof is returned. Appropriate order forms
will be included with the galley proof.

Virginia D. Miller
Editor

A NEW GENUS OF CYPRINODONTID FISH
FROM NUEVO LEON, MEXICO1

By Robert Rush Miller2 and Vladimir Walters3

Abstract: Megupsilon aporus, a new genus and species of
cyprinodontid fish related to Cyprinodon, is described from a
large series of individuals from an interior basin in Nuevo Leon,
Mexico. It is unique in having a huge Y-chromosome in the male
and in the sexually dimorphic chromosome number (male
2n = 47, female 48), as well as in lacking pores in the cephalic
sensory canal system, possessing two distinctive behavioral
traits (jaw-nudge and opercular rotation) not developed in
Cyprinodon, and having blackened scales on the side in the
nuptial male which also lacks a black terminal band on the
caudal fin. Megupsilon inhabits shallower water than does the
species of Cyprinodon with which it is sympatric. It also has a
much shorter gut than Cyprinodon and is carnivorous, whereas
the local Cyprinodon is herbivorous. The new genus is a relict,
representing an earlier invasion of the basin than does the species
of Cyprinodon.

Introduction

Cyprinodontoid fishes comprise nearly one-third of the known freshwater
fish fauna of Mexico (approximately 115 of 390 species). Of these, the autoch-
thonous Goodeidae and the Cyprinodontidae together have about as many
species as do the Poeciliidae, whereas the fourth family of the group, the
Anablepidae, is monotypic. The novelty described here is the third known
endemic Mexican genus of the Cyprinodontidae ( Garmanella Hubbs, 1936
and Cualac Miller, 1956 are the other two); its discovery further emphasizes
the richness and diversity of the continental fish fauna of Mexico. The new
genus is distinguished from all other members of the family karyotyped thus
far by the very large Y-chromosome in the male and the sexually dimorphic
chromosome number (Uyeno and Miller, 1971). It is confined to a single,
spring-fed pond on a high, endorheic plateau in Nuevo Leon, northeastern
Mexico.

1 Editorial Committee for this Contribution

Robert J. Lavenberg
Robert K. Liu
Camm Swift

2 Museum of Zoology, The University of Michigan, Ann Arbor, Michigan 48104.

3Department of Zoology, University of California, Los Angeles, California; and

Research Associate in Ichthyology, Natural History Museum of Los Angeles
County, Los Angeles, California 90007.

1

2

Contributions in Science

No. 233

Xx AftAftn* AftrtftftftftAM
AAftftftft A<5 AAAAfiAAAAftAO

Type species. Megupsilon aporus, new species.

Diagnosis. A Cyprinodon-Yike killifish with uniserial tricuspid jaw teeth
from which it is distinguished by having: (1) a huge Y-chromosome in the male
(unique for cyprinodontoids) and a sexually dimorphic diploid chromosome

Figure 2. Photomicrographs of somatic chromosome complements of a, female
(2n = 48) and b, male (2n = 47), of Megupsilon aporus x 1900.

AAtftftOftftAAA $

XI AdfiOftq JiMAftfirtADD
AAAAftflftANftft lift
DOftftOA AA

Figure 1. Karyotype of Megupsilon aporus.

Megupsilon , new genus
Figures 1-4

1972

New Genus of Cyprinodontid Fish

3

number, 47 in the male and 48 in the female (Figs. 1-2); (2) the cephalic sensory
canal system represented by exposed neuromasts only (no trace of canals or
pores); (3) two distinctive behavioral traits (see below); (4) blackened scales on
the side between dorsal and anal fins in the male (Fig. 3); and (5) nuptial male
without black terminal border on caudal fin. In addition, the following com-
bination of characters separates this genus from all others having tricuspid
teeth that comprise the North American subfamily Cyprinodontinae (for
diagnosis, see Uyeno and Miller, 1962: 528): entire preorbital region scaleless;
pelvic fins and girdle lacking; intestine of adult usually shorter than body
length; gill rakers few (10-13); anal fin of female about as large as her dorsal
fin (Fig. 3). The pelvic fins and girdle are lacking also in Cyprinodon diabolis
and in the Old World species Aphonias (Tellia) apodus, and the development
of squamation in the preorbital region is variable in Cyprinodon and lacking
in Floridichthys.

Relationships. The new genus is obviously closest to Cyprinodon with
which it shares many traits, e.g., tricuspid teeth, body shape, size and position
of fins, squamation, and osteological characters. It has diverged sufficiently

Figure 3. Paratypes (UMMZ 189020) of Megupsilon aporus. Above, male, 24 mm
SL.; below, female, 27.5 mm SL. Photo by Louis P. Martonyi.

4

Contributions in Science

No. 233

STANDARD LENGTH (mm)

Figure 4. Size frequency of 216 Cyprinodon sp., UMMZ 189021, and 533 Megup-
silon aporus, UMMZ 189020, from El Potosf, all collected 25 March 1968. Stipple,
immatures; black, males; clear, females.

1972

New Genus of Cyprinodontid Fish

5

that it is behaviorally and reproductively incompatible with Cyprinodon and
both premating and postmating isolating mechanisms prevent its hybridization
with that genus. Robert K. Liu (personal communication, 1970) has observed
two traits, jaw-nudge and opercular rotation, found in no species of Cyprinodon
tested and has failed to obtain hybrids in forced matings between the two
genera.

Etymology. The generic name is from the Greek prefix peya (mega-),
from peyaa (megas) meaning big, great, and vxpiXov (upsilon), name of the
Greek letter Y ( v ), in reference to the huge Y-chromosome; gender is neuter.
The specific trivial, aporus, is from the Latin, meaning without pores, in
reference to the lack of pores in the sensory cephalic canal system. We are
indebted to Carl L. Hubbs for proposing the generic name.

The material used in the following description is deposited in The
University of Michigan Museum of Zoology (UMMZ), University of Cali-
fornia, Los Angeles (UCLA), and the Natural History Museum of Los Angeles
County (LACM).

Megupsilon aporus, new species
Figures 1-3

Types. Holotype, a breeding male, UMMZ 1 890 1 8, 2 1 .4 mm SL, collected
by R. R. Miller and H. L. Huddle at El Potosi, Nuevo Leon, Mexico, 25 March
1968. Allotype, an adult female, UMMZ 189019, 26.6 mm SL, taken with the
holotype. Paratopotypes: UMMZ 189017, an adult male, 19 mm SL, collected
by Miller and Huddle at the type locality, 23 February 1961 ; UMMZ 189020,
510 juvenile to adult (including a male and female cleared and stained),

12- 36 mm SL, taken with the holotype; LACM 32147-1, 25 juvenile to adult,
11-28 mm SL, (ex UMMZ 189020); UCLA, W68-21, 124 juvenile to adult,

13- 28 mm SL, collected by Vladimir Walters and John Bleck at the type
locality, 15 February 1968.

Additional Material (not designated as paratypes). UCLA, W68-74, 151
juvenile to adult, 7-31 mm SL, collected by Vladimir Walters and Bruce J.
Turner at the type locality, 29 June 1968. Food studies were performed on 60
of these specimens.

Diagnosis. See generic diagnosis (genus is monotypic).

Description. The generic diagnosis of this species includes most of the
important specific characters. Form and pigmentation are portrayed in Figure
3 and other diagnostic features appear in Figures 1 and 2. Proportional
measurements are given in Table 1. Methods of counting and measuring are
those used by Miller (1948: 9-13). The last two closely approximated rays in
both dorsal and anal fins are counted as a single ray.

Dorsal rays: 9(10), 10(29), 11(11), x 10.02, all rays branched in 4 fish,
the first one unbranched in 44, and the first two rays unbranched in 2; anal
rays: 9(4), 10(39), 11(7), x 10.06, all rays branched in 35, the first ray
unbranched in 15; pectoral rays (both fins): 13(15), 14(60), 15(25), x 14.10;

6 Contributions in Science No. 233

Table 1

Proportional measurements, in thousandths of
standard length, of Megupsilon aporus.

Data for the holotype and allotype are included with the 20 adults.

10 Males

10 Females

Measurement

Holo-
type $

Allo-
type $

Range

Aver-

age

Range

Aver-

age

Standard length, mm

21.4

26.6

21.1-27.1

24.0

23.0-31.8

25.7

Predorsal length
Anal origin to

626

616

598-628

616

598-616

610

caudal base

397

380

391-419

404

366-393

378

Body, greatest depth

421

410

415-459

434

388-428

408

Greatest width

210

218

210-234

221

211-234

225

Head length

369

357

346-369

358

343-370

358

Depth

350

320

327-350

340

311-336

325

Width

Caudal peduncle

234

241

234-253

245

232-263

248

length

257

248

257-289

272

244-263

252

Least depth
Interorbital, least

206

199

194-222

208

180-199

191

bony width

93

86

93-103

98

82-95

89

Preorbital width

33

34

30-37

34

29-38

33

Opercle length

117

120

103-119

113

111-126

118

Snout length

84

83

81-96

87

76-87

82

Orbit length

107

105

106-114

i 10

97-114

107

Mouth width

112

120

107-122

116

109-135

123

Upper jaw length

126

128

118-134

128

120-138

130

Mandible length
Dorsal fin, basal

126

124

114-131

124

117-134

126

length

178

177

175-203

186

148-181

168

Depressed length
Anal fin, basal

285

278

268-303

284

236-278

260

length

140

132

122-144

133

118-138

129

Depressed length
Middle caudal rays.

271

244

244-271

254

239-268

250

length

233

229

214-236

227

214-244

225

Pectoral fin length

215

192

188-215

199

182-210

193

caudal rays: 16(1), 17(5), 18(31), 19(10), 20(3), x 18.18. The holotype has
dorsal i,9, anal 0,10, pectorals 15-15, and caudal 18.

Scales in lateral series: 24(8), 25(41), 26(1), x 24.86; scales between dorsal
and anal fins: 10(25), 1 1(24), 12(1), x 10.52; scales around caudal peduncle:
1 4?( 1 ), 15(4), 16(44), x 15.88; scales around body: 26(2), 27(3), 28(28), 29(3),
30(12), 31(0), 32(2), x 28.56; predorsal scales: 18(4), 19(14), 20(14), 21(12),
22(6), x 20.04. The holotype has 25 lateral scales, 10 between dorsal and anal,
16 around peduncle, 28 around body, and 22 predorsal.

Vertebral counts (including hypural complex), taken from radiographs.

1972

New Genus of Cyprinodontid Fish

7

are: 25(8), 26(50), 27(3), x 25.92; of these the precaudal vertebrae number
11(42), 12(15) and the caudal vertebrae 13(1), 14(20), 15(33), 1 6(3). Holotype
11+15-26.

Gill rakers: 10(8), 11(24), 12(14), 13(4), x 1 1.28. Holotype, 11. All gill
rakers on the outer part of the first arch were counted, without distinction
between upper and lower limbs.

The branchiostegals numbered 4 in 6 specimens and 5 in 34; only one
fish had the formula 4-4. In the typical count, 4 branchiostegals insert on the
ceratohyal and 1 on the epihyal.

Coloration and Dimorphism. The life colors of the new genus were noted
in both field and laboratory; the sexes show marked dichromatism (typical also
of Cyprinodon ): nuptial males have steel blue iridescence on the back and sides
anterior to the blackened area that lies between the dorsal and anal fins; the
caudal peduncle, however, has a golden bronze sheen, seen also on top of the
head, and the caudal fin is watery orange, with no trace of the terminal black
border typical of Cyprinodon; the dorsal and anal fins are chalky bluish white,
the base of the dorsal orange. There is a conspicuous, vertical black bar on
the eye above and below the pupil that disappears on preservation. There is
also an orange spot on the posterior part of the opercle, noted only in the male.
Adult females are golden olivaceous over the entire body and have a weak and
often interrupted midlateral stripe, from the upper angle of the gill opening to
the base of the caudal fin, that is no wider than three-fourths the diameter of
the eye; rarely there is a tendency to develop several teardrop-shaped extensions
from this stripe toward the anal fin.

The male differs most notably from the female in having the side of the
body heavily blackened between the tip of the extended pectoral fin and the
bases of the dorsal and anal fins (Fig. 3); this mark varies in development,
apparently being most intense and expansive in alpha males. Neither young
nor adult possess a dorsal ocellus, found in most species of Cyprinodon. The
anal fin of the female is as large as or larger than her dorsal fin, whereas in the
male the dorsal fin is larger than the anal fin (as typical for both sexes
of Cyprinodon ).

As shown in Table 1, there is marked sexual dimorphism in the measure-
ment of anal origin to caudal base, head depth, caudal peduncle length, least
depth of caudal peduncle, least bony width of interorbital, basal length of
dorsal fin, and depressed length of dorsal fin. Except for the interorbital
measurement, sexual dimorphism is similar in Cyprinodon. In addition, males
of Cyprinodon have notably longer anal fins than do females, whereas these
fins are virtually the same length in both sexes of Megupsilon. The functional
significance of this difference may be related to breeding behavior.

Individuals of the new genus are small, attaining a maximum standard
length of only 36 mm (1 female); males are smaller than females and may
mature at 15 mm SL (Fig. 4). The smaller male size may be correlated with the
absence of territorial behavior in this genus (see below). The sympatric species

8

Contributions in Science

No. 233

Figure 5. Spring-fed pond at El Potosf, type locality of Megupsilon aporus. View
northeast, 23 February 1961 (from Kodachrome by R. R. Miller).

of Cyprinodon at El Potosf reaches a larger size and the two sexes are not
significantly different in their maximum lengths.

Discussion. Megupsilon is known only from a spring-fed pond (Fig. 5)
near the northern edge of the small settlement of El Potosf, 95 airline km due
south of Monterrey, on the west side (rain shadow) of the Sierra Madre
Oriental, in Nuevo Leon. The elevation is about 1,880 m, and the highest
adjacent mountains (Cerro Potosf) are about 3,640 m. The pond lies in the
endorheic basin named La Hediondilla, which is a high, arid plateau extending
northward for about 65 km and southward some 50 km from Potosf. We were
told that the pond is the only permanent water in the entire basin, which is
lowest toward the southeast. At high level, the pond covers somewhat more
than 1 hectare and, in places along its eastern side, is 3.5 to 4 m deep. Its water
is very clear though easily roiled because of the firm clay that overlies a
limestone base. Vegetation is abundant, particularly Ceratophyllum which
forms dense masses in the southeastern sector; Potamogeton is restricted to
water deeper than about 1 m, and unidentified “grasses” are restricted to water
shallower than about 1 m; floating masses of green and blue-green algae
(unidentified) occur among the “grasses” and Ceratophyllum; Nasturtium is
also present.

An abrupt limestone cliff (Fig. 5) is at the northeastern edge of the pond.
The water is moderately alkaline (pH 7. 2-7. 4, indicator strips) and moderately

1972 New Genus of Cyprinodontid Fish 9

Table 2

Temperature Measurements/1

Date

Time

Temperature (° C)

23 Feb. 1961

1630

20.6 air, 19.4 water

14 Feb. 1968

2230

18.9 water

15 Feb. 1968

1000

20.0 water

25 Mar. 1968

1100

17.8 air, 18.0 water

28 June 1968

1545

26.0 water

28 June 1968

2235

16.5 air, 17.0 water

aWater temperatures taken 5 cm below the surface, at the south end of the pond.

hard (DH 11-15, approximately 197-269 ppm as CaO). Air and water temper-
atures are summarized in Table 2.

Each year, starting in July, the pond level is lowered about 1 m as water
is pumped out to irrigate the corn fields, according to the residents. This
considerably reduces the surface area of the pond. The pond slowly refills, and
by October covers the area shown in the photograph; water level then remains
stable until the following summer. The commemorative plaque on the wall of
the pumphouse states that this structure was dedicated in 1955 and, according
to the residents, the partial dam which parallels the limestone cliff and serves
to delimit the deeper portion of the pond from the shallower areas was built
in 1960. The annual man-caused changes in the level of the pond may have
enabled “grasses” to colonize those pond areas which become dry land
in summer.

On 23 February 1961 Miller and Huddle collected a single Megupsilon
and 315 Cyprinodon whereas subsequent collections made in Feburary,
March and June, 1968, revealed that Megupsilon was 2 or 3 times more
abundant than Cyprinodon. The 1968 collections indicate that Megupsilon
predominates in the grassy areas of the pond and in the Ceratophyllum whereas
the Cyprinodon, especially the adults, inhabits water deeper than 1 m. It appears
to us that yearly pumping of the pond has resulted in an increase in Megupsilon
habitat and a decrease in Cyprinodon habitat. During pluvial times (Wisconsin
glaciation), when the now restricted pond probably formed a sizable marsh
and lake, the habitat suitable for Megupsilon would have been extensive.

One other species of fish, the goldfish (Carassius auratus), is present in
the pond. Most were greenish bronze but one bright golden one was noted in
1961 and a number of golden individuals were seen in 1968; the brightly-
colored goldfish were confined to the deepest part of the pond and were large,
perhaps the original propagules. A dwarf species of crayfish, Cambarellus
alvarezi Villalobos (1952), is endemic to this pond.

Mr. Robert J. Naiman, while a graduate student at UCLA, measured gut
length and studied dietary preferences of the 2 cyprinodontids of El Potosi
(Tables 3-4). Megupsilon has a much shorter digestive tract than does
Cyprinodon:

10

Contributions in Science

No. 233

x Gut Length

Species

(as % SL)

Range

N

Size Range

Megupsilon adults

88%

53-130%

55,

16-33 mm SL

Cyprinodon adults

211%

137-348%

36,

27-54 mm SL

Megupsilon juveniles

78%

53-100%

5,

13-15 mm SL

Cyprinodon juveniles

112%

90-133%

14,

10-16 mm SL

Mr. Naiman’s data indicate that Megupsilon is carnivorous and feeds
mainly on larval chironomids whereas Cyprinodon is herbivorous and feeds
mainly on filamentous algae. The average adult Megupsilon contains 3.96
times more animals than does the average adult Cyprinodon, and Megupsilon
juveniles, on the average, contain 4.32 times more animals than do Cyprino-
don juveniles. On the other hand Cyprinodon adults ingest considerably more
plant matter than does Megupsilon; the mean fullness value (filamentous algae
plus vascular plants) for Cyprinodon is 22.07 times that for Megupsilon and
since an adult Cyprinodon gut is 4.3 times the length of an adult Megupsilon
gut (x gut length in adults is 86.5 mm vs. 20.3 mm, respectively) Cyprinodon
must ingest about 100 times more plant matter than does Megupsilon. Both
species were found to contain appreciable amounts of unicellular algae such
as diatoms and desmids but no attempt was made to estimate quantities.

When Walters and Bleck arrived at the pond on 14 February 1968,
Megupsilon was observed to be actively swimming about at 2230 hrs. No

Table 3

Feeding Preferences of the El Potosf Cyprinodontidsa

Food Category

Megupsilon aporus, juveniles5
Mdn No./Fish x No./Fish

Cyprinodon sp., juveniles0
Mdn No./Fish x No./Fish

Chironomid larvae

11.0

11.8

0.50

1.21

Other insects plus

arachnids

0.3

1.2

0.14

0.50

Copepods (Cyclops)

0.3

4.6

0.14

2.50

Larger crustaceans

0.1

0.2

0.00

0.00

Eggs (cyprinodont?)

0.1

0.4

0.00

0.00

Insect eggs

0.1

not counted

0.00

0.00

Filamentous algae

0.1d

0.5d

9.0d

7.4d

Vascular plants

0.0d

0.0d

0.04d

0.29d

aFishes collected by seining at 0900-1000, 29 June 1968.

bN = 5, 13-15 mm SL; 100% with food in gut; no helminth parasites found.

CN = 14, 10-16 mm SL; 92.9% with food in gut; no helminth parasites found.
dFullness values. For plant matter, the fullness of the gut was estimated on an
arbitrary scale of 0 (gut devoid of algae/vascular plants) to 10 (gut stuffed with
algae/vascular plants).

1972

New Genus of Cyprinodontid Fish

1 1

Table 4

Feeding Preferences of the El Potosi Cyprinodontidsa

Megupsilon aporus, adultsb Cyprinodon sp., adultsc

Food Category

Mdn No./Fish

x No./Fish

Mdn No./Fish

x No./Fish

Chironomid larvae

9.00

15.71

0.93

4.47

Other insects plus
arachnids

2.45

2.25

0.40

1.28

Copepods ( Cyclops )

0.58

4.58

0.01

0.03

Larger crustaceans

1.13

2.87

0.08

0.56

Eggs (cyprinodont?)

0.22

1.20

0.10

0.39

Gastropods

0.03

0.07

0.00

0.00

Filamentous algae

0. 15d

0.40d

9.64d

8.90d

Vascular plants

0.02d

0.04d

0. 19d

0.8 ld

aFishes collected by seining between 0900-1000, 29 June 1968.
bN = 55; 20 males, 35 females, 16-33 mm SL; 100% with food in gut; 60.0% with
helminth parasites.

CN = 36; 12 males, 24 females, 27-54 mm SL; 100% with food in gut; 63.9% with
helminth parasites.

d Fullness values. For plant matter, the fullness of the gut was estimated on an arbi-
trary scale of 0 (gut devoid of algae/vascular plants) to 10 (gut stuffed with algae/
vascular plants).

reproductive activity was noted then or the following morning, which was
marked by light rain and overcast sky. Walters and Turner noted that
Cyprinodon males were maintaining territories in deep water on June 28-29,
but such behavior was not observed for Megupsilon.

The two killifishes are endemic to the El Potosi pond today. The pond
undoubtedly represents the last remnant of a larger body of water which may
have filled much of La Hediondilla during Pleistocene pluvial periods. At
some past Pleistocene time, the hypothetical lake must have had a drainage
connection to the north or northwest to permit entry by the ancestor of the
El Potosi Cyprinodon. Although this form has not been studied, it appears to
belong to the group of species allied to Cyprinodon eximius Girard, which
today occurs in isolated drainages and in the Rio Conchos basin, of northern
Mexico, as well as in certain Rio Grande tributaries in Texas.

The population of Megupsilon aporus can only be regarded as relict and
representative of a much earlier cyprinodontine invasion of the Mexican
Plateau. That it is most closely related to Cyprinodon is indicated by the many
shared morphological characters. Another relict cyprinodontine, Cualac tes-
sellatus Miller, inhabits a warm spring area (La Media Luna) near Rio Verde
in San Luis Potosi.

12

Contributions in Science

No. 233

On the morning of 6 July 1972 Walters revisited the spring pond, accom-
panied by Robert E. Brown, Jr., Richard Haas, Robert K. Liu, and Sylvia H.
Walters. Conditions had changed since the last visit. The pump has been
inoperable for several years and the spring’s flow is now tapped year-round by
sluices. Since pond area is now fairly constant there has been a change in the
aquatic vegetation. Ceratophyllum demersum, restricted to the area of the
pump house in 1 968, now covers most of the pond with a thick mat ; in shallower
areas this vegetation was moribund but in fruit, possibly reflecting elevated
summer water temperatures. Wide-angle Infrared Ektachrome photographs,
taken with a Wratten 12 filter from the hillside about 25 feet above the spring,
show the moribund areas as white to pale pink vs. red for healthy areas. In cooler
areas the Ceratophyllum is partially overlain by Ranunculus sp. Grasses are
diminished.

Water temperatures, measured between 9 AM and 12 noon with a YSI
telethermometer at several scattered locations, were 22-23 °C (surface), 16-
19°C (shallow depths), and 18°C at the deepest point. Oxygen content,
measured at the same times with a Hach Kit, ranged from 4. 5-7. 5 ± /0.5 mg/
1 = 2. 8-5. 2 ml/1 ; the lower readings were taken in shade, near Ceratophyllum.
Four minnow traps, baited with chicken liver and placed in deep water below
the Ceratophyllum mat for 90 minutes and then in shallow water in the Cerato-
phyllum mat for 90 minutes yielded several hundred Cyprinodon sp., 8
Megupsilon aporus, and 2 dwarf crayfish. The trapping results were surprising,
in view of the dietary differences between the two fishes as indicated by earlier
gut analyses.

Megupsilon aporus was seen to be abundant immediately below and in the
Ceratophyllum mat. Cyprinodon sp. abounded in open water, from the surface
to the deepest part of the spring; territorial males were tightly packed in shallow
water along the western side of the pond. Crayfish abounded in the Cerato-
phyllum. The goldfish population seemed unchanged. No specimens were
preserved; all trapped fish were released.

Acknowledgments

We are indebted to Robert K. Liu for allowing us to publish his observa-
tions on the behavior of Megupsilon, and to Robert J. Naiman for the data on
gut length and food preferences. Our colleague, Teruya Uyeno, prepared the
chromosomes. Biologo Juan Luis Cifuentes L., Direccion General de Pescas
e Industrias Conexas, kindly issued permits for collecting in Mexico. Field
work by the senior author was supported by NSF grant GB-6272X and that
by the junior author by University of California Faculty Research Grant No.
1780; laboratory studies were supported by NSF GB 8212 (to The University
of Michigan Museum of Zoology for Research in Systematic and Evolutionary
Biology).

1972

New Genus of Cyprinodontid Fish

13

Resumen

Megupsilon aporus, un nuevo genero y especie de la familia Cyprinodon-
tidae mas cercamente relacionado a Cyprinodon, se describe de un estanque
aislado en Nuevo Leon Mexico. Solamente otro pez, una especie de Cyprino-
don, es indigeno del mismo manantial. Este nuevo genero se distingue por
medio de su dimorfismo sexual en numero de cromosomas, 2n — 47 en el
macho y 2n = 48 en la hembra, y el macho tambien con una enorme cromosoma
Y. Ademas Megupsilon solamente tiene neuromastos expuestos (carece canales
o poros) en el sistema canal sensorio cefalico, sin aletas o cenidor pelviano, el
intestino del adulto mas corto que el largo del cuerpo, pocos rastrillos
branquiales (10-13), el macho nupcial sin margen negra terminal en la aleta
caudal pero con una region enegredida en el lado entre las aletas dorsal y anal,
y la aleta anal de la hembra aproximadamenta tan grande como su aleta dorsal.
Ensena dos caracteristicas de comportamiento que no se encuentran en
Cyprinodon y no es territorial. Es carnivoro y prefiere agua mas o menos poco
profunda. El nuevo genero es una reliquia representando una invasion mas
temprana que la del especie simpatrica de Cyprinodon.

Literature Cited

Miller, R. R. 1948. The cyprinodont fishes of the Death Valley system of eastern
California and southwestern Nevada. Univ. Mich. Mus. Zool., Misc. Publ.
68: 1-155.

Uyeno, T., and R. R. Miller. 1962. Empetrichthys erdisi, a Pliocene cyprinodontid
fish from California, with remarks on the Fundulinae and Cyprinodontinae.
Copeia 1962 (3): 519-531.

1971. Multiple sex chromosomes in a Mexican cyprinodontid fish. Nature

231:452-453.

Villalobos, A. 1952. Estudios de los cambarinos Mexicanos. X. Una nueva especie
del genero Cambarellus del estado de Nuevo Leon. Anal. Inst. Biol. 22
(2): 525-532.

Accepted for publication April 5, 1972

NUMBER 234
OCTOBER 30, 1972

c5^ 7* 7^

c?Lttf

INDO-WEST PACIFIC FISHES
FROM THE GULF OF CHIRIQUI, PANAMA

By Richard H. Rosenblatt,
John E. McCosker,
and Ira Rubinoff

CONTRIBUTIONS IN SCIENCE

NATURAL HISTORY MUSEUM • LOS ANGELES COUNTY

CONTRIBUTIONS IN SCIENCE is a series of miscellaneous technical papers
in the fields of Biology, Geology and Anthropology, published at irregular intervals
by the Natural History Museum of Los Angeles County. Issues are numbered sep-
arately, and numbers run consecutively regardless of subject matter. Number 1 was
issued January 23, 1957. The series is available to scientific institutions and scien-
tists on an exchange basis. Copies may also be purchased at a nominal price.
Inquiries should be directed to Virginia D. Miller, Natural History Museum of Los
Angeles County, 900 Exposition Boulevard, Los Angeles, California 90007.

INSTRUCTIONS FOR AUTHORS

Manuscripts for CONTRIBUTIONS IN SCIENCE may be in any field of Life
or Earth Sciences. Acceptance of papers will be determined by the amount and char-
acter of new information. Although priority will be given to manuscripts by staff
members, or to papers dealing largely with specimens in the collections of the Muse-
um, other technical papers will be considered. All manuscripts must be recommend-
ed for consideration by the curator in charge of the proper section or by the editorial
board. Manuscripts must conform to those specifications listed below and will be ex-
amined for suitability by an Editorial Committee including review by competent
specialists outside the Museum.

Authors proposing new taxa in a CONTRIBUTIONS IN SCIENCE must indi-
cate that the primary type has become the property of a scientific institution of their
choice and cited by name.

MANUSCRIPT FORM.— (1) The 1972 CBE Style Manual, third edition
(AIBS) is to be followed in preparation of copy. (2) Double space entire manuscript.
(3) Footnotes should be avoided if possible. Acknowledgments as footnotes will not
be accepted. (4) Place all tables on separate pages. (5) Figure legends and unavoid-
able footnotes must be typed on separate sheets. Several of one kind may be placed on
a sheet. (6) An abstract must be included for all papers. This will be published at the
head of each paper. (7) A Spanish summary is required for all manuscripts dealing
with Latin American subjects. Summaries in other languages are not required but are
strongly recommended. Summaries will be published at the end of the paper. (8) A
diagnosis must accompany any newly proposed taxon. (9) Submit two copies of
manuscript.

ILLUSTRATIONS. — All illustrations, including maps and photographs, will be
referred to as figures. All illustrations should be of sufficient clarity and in the proper
proportions for reduction to CONTRIBUTIONS page size. Consult the 1972 CBE
Style Manual, third edition (AIBS) in preparing illustration and legend copy for
style. Submit only illustrations made with permanent ink and glossy photographic
prints of good contrast. Original illustrations and art work will be returned after the
manuscript has been published.

PROOF. — Authors will be sent galley proof which should be corrected and
returned promptly. Any changes or alterations, other than typographical corrections,
will be billed to the author. Unless otherwise requested, page proof also will be sent
to the author. One hundred copies of each paper will be given free to each author or
divided equally among multiple authors. Orders for additional copies must be sent
to the Editor at the time corrected galley proof is returned. Appropriate order forms
will be included with the galley proof.

Virginia D. Miller
Editor

INDO-WEST PACIFIC FISHES FROM THE
GULF OF CHIRIQUF PANAMA1

By Richard H. Rosenblatt2, John E. McCosker2, and Ira Rubinoff3

Abstract: Recent collections indicate the presence of a
number of Indo-west Pacific fishes in the Gulf of Chiriqui. The
Gulf of Chiriqui is not subject to seasonal upwelling as is the
adjacent Gulf of Panama, and supports a relatively rich develop-
ment of hermatypic corals. Twenty-four percent (40) of the reef
fish species collected there also occur in the Indo-west Pacific,
and of them, nine were previously unrecorded at or near the
American mainland: Myripristis murdjan, Ctenochaetus cyano-
guttatus, Gymnothorax flavimarginatus, G. buroensis, G. an-
dulatus, Enchelynassa canina, Uropterygius tigrinus, Malacan-
thus hoedti, and Hemipteronotus taeniourus. The last six are
heretofore unreported from the eastern Pacific, although none
is restricted to the Gulf of Chiriqui.

Eastern Pacific records of the following Indo-west Pacific
species are regarded as invalid, being based either on misidenti-
fication or mislabelings: Brachysomophis crocodilinus, Gymno-
thorax chilospilus, Callechelys marmoratus, Myrichthys macalo-
sus, Myripristis berndti, Lutjanus kasmira, Runulci tapeinosoma,
Abudefduf saxatilis vaigiensis, and Antennatus bigibbus.

The ranges of the eastern Pacific endemic species Gymno-
thorax castaneus, Petrotyx hopkinsi, and Paraclinus altivelis are
extended to Panama. Xyrichthys panamensis Fowler 1944, is
synonymized with Hemipteronotus pavoninus (Valenciennes,

1839).

Many of the transpacific migrants are localized and limited
in their eastern Pacific distributions. Some are seemingly closely
associated with the development of hermatypic corals. There is
no evidence that any are displacing eastern Pacific endemnic
species.

The number of new records in the Gulf of Chiriqui collec-
tions reflects the inadequacy of current knowledge of the distri-
bution of the fishes of the eastern tropical Pacific.

Introduction

Recent collecting efforts by the Scripps Institution of Oceanography and
the Smithsonian Tropical Research Institute in the Gulf of Chiriqui, western
Panama, have disclosed the presence of a large number of Indo-west Pacific
species adjacent to or along the continental coastline in the eastern tropical

Review Committee for this Contribution
William A. Bussing
Robert J. Lavenberg
C. Richard Robins
James C. Tyler

Contribution from the Scripps Institution of Oceanography, University of Cali-
fornia at San Diego, La Jolla, California 92037

3Smithsonian Tropical Research Institute, P.O. 2072, Balboa, Canal Zone

1

7

Contributions in Science

No. 234

Pacific. Our results are interesting in that many fishes of Indo-west Pacific
origin which have previously been reported only from the oceanic Galapagos,
Revillagigedo, Cocos, and Clipperton islands are maintaining populations in
the coral reef communities in the Gulf of Chiriqui.

The Gulf of Chiriqui lies west of the Gulf of Panama and is not subject to
the seasonal upwelling conditions which profoundly affect the fauna of Panama
Bay and the Perlas Islands (Schaefer et al., 1958; Forsbergh, 1969). Pacific
coastal waters west of the Azuero Peninsula, therefore, present a warmer and
more stable thermal regime (Renner, 1963) which facilitates extensive develop-
ment of certain hermatypic corals (Glynn, in press). The presence of extensive
Pocillopora bank reefs (Fig. 1) to depths of 10-15 meters provides a habitat
similar, but not identical, to that of the islands of the central Pacific. These
reefs, in contrast to well-developed Caribbean or Indo-Pacific formations,
comprise relatively few species of Pocillopora, possibly three or four. Associ-
ated with them, however, are several species of Porites, Pavona, and the hydro-
coral Millepora which contribute to the habitat diversity, to which the in-
creased Indo-west Pacific components in the vertebrate and invertebrate fauna
may be related. The structure and extent of coral reef development in the Gulf
of Chiriqui is discussed in Glynn et al. (in press). Notable Indo-west Pacific
invertebrates in the Gulf of Chiriqui include the crown of thorns starfish,
Acanthaster cf. plane i, the painted shrimp Hymenocera picta, and the fire
corals Millepora intricata and M. platyphylla (Glynn, in press). Eastern Pacific
records for Hymenocera and Millepora are based on specimens from the Gulf
of Chiriqui, these forms being as yet unreported from Clipperton, Galapagos,
and the Revillagigedo islands. A similar restricted distribution pattern also
exists for certain fishes.

Collections

The eastern Gulf of Chiriqui contains seven major island groups. The
largest is Coiba which is ca. 30 km in length. The outermost island, Montuosa,
is 60 km from the mainland and separated by a channel 80 m deep. We have
either collected at or made observations using SCUBA at each island group
and several mainland localities (Fig. 2) on three separate occasions, during
March and September of 1970 and April of 1971. More than 30 days were
spent in the field while aboard the vessels R/V Alpha Helix, R/V Tethys,
and USN LST Traverse County. A collection of fishes made by C. H. Birkeland
and T. Spight at Isla Viradores Sur, Costa Rica (10°34'50"N, 85°43'30"W),
is included in this study. Accessory material from other Pacific island and Gulf
of California localities was provided through the extensive collecting efforts
of the Scripps Institute of Oceanography (SIO), and the University of California
at Los Angeles (UCLA). Fishes discussed in this paper are presently housed at
SIO, UCLA, the Smithsonian Tropical Research Institute (STRI), the Univer-
sity of Miami Marine Laboratory (UMML), the Harvard Museum of Com-
parative Zoology (MCZ), the Universidad de Costa Rica (UCR), and the
California Academy of Sciences (CAS). In this study we refer to the offshore

1972

Fishes from the Gulf of Chiriqui, Panama

3

O ON

G-g

. s-
C/3 cd
cd *5
o ^
<u

oo ^

J <N

C/3

2 C

v> C
M >>

45

cd ^
U «
cd qj
^ CL,

r 1 .

pj 2

c n o

<u ^
« *

<- ^
rt P
X> C

O

-2

? -2
£Z;
£ S

r" sf

• «3
Cd -£3

O ^

r *

£ 8

O Co

Cd

X 3
& £

4

Contributions in Science

No. 234

83° 82° 81°

Figure 2. Major collection localities in the Gulf of Chiriqui, Panama. 100 and 1000
fathom contours taken from hydrographic chart H.O. 1018 (1968 edition). 1, Puerto
Armuelles; 2, Islas Ladrones; 3, Isla Montuosa; 4, Isla Parida; 5, Isla Balanos and
Isla Berraco; 6, Islas Secas; 7, Isla Brincanco and 8, Isla Uva, Islas Contreras; 9,
Bahia Honda; 10, Isla Canal de Afuera; 1 1, Isla Rancheria; 12, Isla Coiba, Bahia de
Las Damas; 13, Isla Jicaron; 14, Isla Jicarita; 15, Isla Cebaco; 16, Islas Naranjas.

islands of the eastern Pacific ocean. These include: Isla del Coco, Clipperton
Island, Islas de Revillagigedo, and Islas Galapagos.

Gulf of Chiriqui Fishes

We have discovered nine Indo-west Pacific fish species previously un-
recorded at or near the American mainland. Six species are first reported from
the eastern Pacific in this paper. The Indo-west Pacific fishes of the Gulf of
Chiriqui can be placed in two categories (Table 1) based on their distribution,
and most likely, their dependence upon the coral reef habitat. These cate-
gories are arbitrary in some cases, but for the most part the distinction is rather
clear cut.

The fishes that are part of the coral reef community of the gulf island
groups include 165 species; of these we find that 40 (24 percent) also occur
in the Indo-west Pacific region. This high percentage is comparable only to the
Clipperton fish fauna, and is probably associated with the extensive coral
development at both localities.

Other fishes collected in the Gulf of Chiriqui represent range extensions
for the eastern tropical Pacific. A single specimen of Paraclinus altivelis

1972

Fishes from the Gulf of Chiriqui, Panama

5

Table 1

Eastern Pacific distributions of Indo-west Pacific and circumtropical shorefish
species. * Indicates species found in the Gulf of Chiriqui.
f Indicates circumtropical species.

/. Broadly distributed in eastern tropical Pacific

f * Aetobatus narinari (Euphrasen)
Chanos chanos (Forsskal)
f Albula vulpes (Linnaeus)

* Euleptorhamphus viridis

(Van Hasselt)

t Ablennes hians (Valenciennes)

* Kuhlia taeniura (Cuvier)
t * Priacanthus cruentatus

(Lacepede)

t*A lugil cephalus Linnaeus

* Alectis ciliaris (Bloch)

* Gnathanodon speciosus

(Forsskal)

*Scarus ghobban Forsskal
*S. rubroviolaceous Bleeker
*Sectator ocyurus

(Jordan and Gilbert)

* Oxycirrhites typus Bleeker

* Ci rrh i tick thys oxy cephalus

(Bleeker)

* Doryrhamplus melanopleura

(Bleeker)

* Acanthurus xanthopterus
(Valenciennes)

* Fist ularia petimba Lacepede

t Canthidermis maculatus (Bloch)
Chilomycterus affinis (Gunther)
f *Diodon holacanthus Linnaeus
t *D. hystrix Linnaeus
*Arothron hispidus (Linnaeus)

* A. meleagris

(Bloch and Schneider)
*Ostracion meleagris Shaw

II. Limited to offshore islands and/or certain mainland localities

* Triaenodon obesus (Riippell)

* Echidna nebulosa (Ahl)

*£. zebra (Shaw)

Gymnothorax buroensis
(Bleeker)

* G. flavimarginatus (Riippell)
G. pictus (Ahl)

*G. undulatus (Lacepede)

* Enchelynassa canina

(Quoy and Gaimard)

* Uropterygius tigrinus (Lesson)
Holotrachys lima

(Valenciennes)

* Myripristis murdjan (Forsskal)
Aphareus furcatus (Lacepede)

* Malacanthus hoedti Bleeker
*Caranx melampygus Cuvier

Forcipiger flavissimus
Jordan and McGregor
* Hemipteronotus pavoninus
(Valenciennes)

* H. taeniourus (Lacepede)

* Thalassoma lute see ns

(Lay and Bennett)

Calotomus spinidens
(Quoy and Gaimard)

* Aulostomus chinensis

(Linnaeus)

* Acanthurus triostegus Linnaeus
*A. glaucopareius Cuvier
'fCtenochaetus cyanoguttatus

Randall

'•'Zanclus canescens (Linnaeus)
Antennarius drombus
Jordan and Evermann
t Xanthichthys r ingens
(Linnaeus)

f *Melichthys niger (Bloch)
t * Alutera scripta (Osbeck)
Canthigaster amboinensis
(Bleeker)

6

Contributions in Science

No. 234

(Lockington), previously known only from deep water in the Gulf of Cali-
fornia (Rosenblatt and Parr, 1969), was collected in ten m at Isla Canal de
Afuera (SIO 71-52). Numerous specimens of Gymnothorax castaneus Jordan
and Gilbert (which we regard as distinct from G. dovii Gunther) were collected
at several Gulf of Chiriqui and Panama Bay locations and represent a southern
extension from the previously known range in Mexico. A single specimen of
the brotulid Petrotyx hopkinsi Heller and Snodgrass from Isla Uva (SIO 70-
135) extends the recorded range of the species from the Galapagos Islands,
although it has also been taken between Cape San Lucas and Espiritu Santo
Island, Lower California (SIO material). The collections also include a new
species of chaenopsid (Stephens and Rosenblatt, MS) and a new species of
dactyloscopid, both of which are distinctively different from known genera.

Transpacific Shore Fishes

Briggs (1961, 1 964) has listed 62 transpacific shore fishes. His list includes
certain records that our studies indicate are invalid for various reasons. These
are discussed below:

Br achy somo phis crocodilinus (Bennett). The eastern Pacific occurrence
of this species rests on a report by Gunther (1870) of a single specimen listed
as “Galapagos Islands. From the Haslar Collection.” Incorrect provenances of
Haslar Hospital collection material has already led to several zoogeographic
improbabilities (Kresja, 1960). In light of this, and lacking other records, we
remove B. crocodilinus from the fauna of the eastern Pacific.

Gymnothorax chilospilus Bleeker. Herre’s (1936) record of this species
from Eden Island Galapagos, was based on a small specimen of Muraena
lentiginosa Jenyns. We have examined Herre’s specimen (SU 24399, now at
CAS) and compared it with other material of M. lentiginosa. Herre’s record of
Gymnothorax undulatus (Lacepede), also based on M. lentiginosa, is dis-
cussed later in this paper.

Callechelys marmoratus (Bleeker). Fowler’s (1932) record of this species
from Charles Island, Galapagos, pertains to the recently described eastern
Pacific species C. galapagensis McCosker and Rosenblatt, 1972.

Myrichthys maculosus (Cuvier). Fowler’s (1938) record of M. maculosus
from Narborough Island, Galapagos is referable to M. tigrinus Girard, an
eastern Pacific endemic. The two nominal species are identical in external
appearance. However, eastern Pacific populations have a significantly lower
number of vertebrae than central and western Pacific material (McCosker, in
preparation).

Myripristis berndti Jordan and Evermann. Although Greenfield (1965)
did not place Briggs’ (1964) record of M. berndti in the synonymy of M. murd-
jan (Forsskal) he does include the three Cocos Island specimens recorded by
Briggs in his material of M. murdjan.

Lutjanus kasmira (Forsskal). The eastern Pacific endemic L. viridis
Valenciennes is very similar to the Indo-west Pacific L. kasmira. Seale (1940)

1972

Fishes from the Gulf of Chiriqui, Panama

7

regarded the two as synonymous in recording L. kasmira from the Galapagos
and Cocos islands. However, Jordan and Evermann (1898) had noted morpho-
logical differences between L. kasmira and L. viridis and regarded the latter as
distinct. Our material indicates differences in color pattern between the two
species. In L. viridis there are five distinct blue stripes, the lowest behind the
pectoral base; in L. kasmira this band is absent. The upper three stripes in
L. viridis are almost horizontal, contacting the dorsal profile at the base of the
ninth dorsal spine, between the ninth and tenth dorsal soft rays, and the anterior
one-third of the caudal peduncle respectively. In L. kasmira the corresponding
points are the sixth dorsal spine, the fifth or sixth dorsal soft ray, and the end of
the soft dorsal. Also in L. viridis the fourth stripe runs forward below the eye
to the upper lip, rather than ending at the preopercular margin. Seale’s (1940)
record then should be considered a misidentification of L. viridis, and L.
kasmira removed from the eastern Pacific list.

Runula tapeinosoma (Bleeker). Clark’s (1936) Galapagos record of Petro-
scirtes tapeinosoma was without doubt based on a specimen of the wide ranging
eastern Pacific Plagiotremus azaleus (Jordan and Boilman).

Abudefduf saxatilis vaigiensis (Quoy and Gaimard). The taxonomy of the
Abudefduf saxatilis species complex is confused. The Atlantic, Indo-west
Pacific, and eastern Pacific populations have been considered to represent
distinct species or subspecies (A. saxatilis (Linnaeus), A. vaigiensis and A.
troschelii (Gill) respectively) or sometimes united under the oldest name, A.
saxatilis. Herre’s listing of Galapagos material with specimens from the western
Pacific under the name A. saxatilis is insufficient reason to establish the pres-
ence of the Indo-west Pacific form at that locality.

Scarops jordani (Jenkins) and Scarus rubroviolaceus Bleeker. These
nominal species have recently (Rosenblatt and Hobson, 1969) been shown to
be synonymous. The older name is S. rubroviolaceus.

Amanses carolae (Jordan and McGregor). This species has been shown
by Randall (1964) to be synonymous with Cantherines dumerilii (Hollard),
known from east Africa, the Seychelles, Lord Howe Island, the central Pacific
and Hawaii.

Antennatus bigibbus (Lacepede). The specimen on which the Revillagigedo
Island record was based (BC 57-160) was included by Rosenblatt (1963) in his
material of the eastern Pacific endemic Antennatus strigatus (Gill). A. bigibbus
has not yet been taken in the eastern Pacific.

Our findings, in general, agree with the concept of the eastern Pacific
barrier to shorefish distribution as proposed by Ekman (1953) and amplified by
Briggs (1961, 1964). Most of the Indo-Pacific elements in western Panama
possess larval stages adapted to long distance pelagic transport, or juveniles
and adults which may be able to accompany floating debris across the equa-
torial Pacific using the north equatorial current system (Hubbs and Rosenblatt,
1961).

The often mentioned but poorly understood phenomenon of offshore

8

Contributions in Science

No. 234

insular confinement in the eastern Pacific (Snodgrass and Heller 1905; Briggs,
1961, 1967; McCosker, 1971; Rosenblatt and Walker, 1963) deserves further
mention. It is important to note that Indo-west Pacific migrants are not only
confined to the offshore islands, but are also usually less abundant than the
congeneric species of the indigenous fauna. An example is the sympatric
association of the squirrelfishes Myripristis murdjan and M. leiognathus Valen-
ciennes. The former, an Indo-Pacific emigrant, is, in the Gulf of Chiriqui,
always found with, but less abundant than, the latter, a widespread eastern
Pacific species. The same situation seems to pertain at Clipperton Island,
except that the abundant eastern Pacific endemic there is M. clarionensis. A
similar picture is also found in the Indo-Pacific morays in the eastern Pacific,
except at Clipperton Island.

The evidence that the direction of movement across the Pacific has been
from west to east has been presented by Briggs (1961) and Hubbs and Rosen-
blatt (1961). More recent findings have done little to alter their conclusions.
It is, however, difficult to argue a west Pacific origin for Sectator ocyurus. The
species has been recorded only from Hawaii and the Marquesas, on the fringes
of the area, and might have crossed from east to west.

Briggs (1961, 1967, 1969, 1970) has in part ascribed the greater success
of the Indo-west Pacific species in crossing the eastern Pacific barrier to their
status as “dominant species." He (1967: 575) has stated that “It seems clear
that the unusually stable ecosystems and high level of competition (in the Indo-
west Pacific region) provide the proper environment for the evolution of domi-
nant species that can successfully invade the other regions.’’

Inherent in this argument is the concept that competition between species
leads to an increase in general “fitness’’ and the ability to compete in a new
habitat with different competitors. This might be true if competition (overlap
of requirement(s) for resource(s) in short supply) inevitably led to the extinction
of all competitors but one, leaving a generalist occupying a broad niche. How-
ever, the widespread phenomenon of character displacement (Brown and
Wilson, 1956) indicates that a more common result of competitive interaction
is coexistence, with competition reduced by narrowing of niche breadth.
Competition thus is more likely to produce specialists than generalists. The
richness of the Indo-west Pacific fauna, especially in sympatric congeneric
species, indicates that competitive interactions have had the latter result. For
example Chave (in press) has carefully studied partitioning of the environment
by six species of Apogon in Hawaii. Although all six occur together, there are
differences in substrate preference, time of feeding, position in the water column
while feeding, and food organisms taken. Her observations indicate that
resources are partitioned in such a way as to reduce competition. Hobson’s
(1968) observations on Apogon retrosella, an eastern Pacific endemic which
overlaps in part of its range with a single congener, A. parri, indicate much less
restriction in several of these parameters. It is found over rocks as well as over
sand patches at night, and feeds benthically as well as in midwater. Although it

1972

Fishes from the Gulf of Chiriqui, Panama

9

is difficult to predict the results of invasions (MacArthur and Wilson, 1967,
Chap. 5), there is no a priori reason to suppose that any one of the Hawaiian
species of Apogon, each with a narrow range of substrate and food preferences,
would be able to replace A. retrosella if introduced into the habitat of that
species.

The data indeed indicate that eastward migrants have not displaced
eastern Pacific endemics. As our previous discussion has shown, a large number
of eastward migrants are limited in their eastern Pacific distributions. The
Muraenidae are instructive in this regard. There are 15 endemic species of
muraenids, distributed among six genera, in the eastern tropical Pacific. As
might be expected from their pelagic larval stage, the muraenids are repre-
sented by more species of migrants than any other family. Seven species dis-
tributed among four genera have crossed the east Pacific barrier. However,
none of these is widespread and abundant along the mainland coast.

The success of Indo-west Pacific forms in colonizing the eastern Pacific
seems to be related to several factors, among them the ability to survive in the
coral-poor, more variable environment of the eastern Pacific, as well as to the
presence of endemic competitors. The idea that these species are behaving as
“competitively dominant species” is unwarranted, and not supported by
evidence.

The paradox that the major equatorial currents flow from east to west but
the major faunal movements have been from west to east is more apparent than
real. The North Equatorial Current is relatively weak to the east. Movement of
water from the mainland of Central America is not strongly unidirectional and
more a drift than a current for much of the year (Wyrtki, 1965). In addition a
considerable part of the north equatorial current is derived from the California
Current, which would not be carrying tropical elements. The South Equatorial
Current, which is strong and consistent near its eastern source, originates from
the cold Peru Current which flows along the South American coast, where the
fauna is essentially temperate (Myers, 1941; Ekman, 1953; Morrow, 1957). It
is not surprising that these currents have not been major highways for tropical
shore-fish dispersal.

The present impoverishment of the coral reef habitat in the eastern tropical
Pacific appears to be limiting the diversity of corallophilic fishes and other in-
shore faunal elements (as Emerson, 1967, has suggested for the Panamic
molluscan fauna). The presence of a suitable reef habitat may be a key to the
success of Indo-west Pacific elements in the Gulf of Chiriqui. A similar associa-
tion of Indo-west Pacific fishes with notable coral development has been des-
cribed for Isla Jaltemba, Nayarit, Mexico by Greenfield et al. (1970), and an
association between coral and certain eastern Pacific scarids has been demon-
strated by Rosenblatt and Hobson (1969: 438). As was pointed out in the
latter paper, the causative factors in this relationship are not clear. It may be
that hermatypic corals and the associated fishes have similar requirements
with respect to the physical environment. For example, Myripristis murdjan

10

Contributions in Science

No. 234

would seem by its distribution to be a strongly corallophilic form. However, it
is a nocturnal planktivore which seemingly utilizes coral only as a shelter during
the day. Additionally, Indo-west Pacific species form a conspicuous component
of the fish fauna at the region of Cape San Lucas, lower California, an area of
much poorer coral development than the Gulf of Chiriqui. The interrelation-
ships between the biotic and physical factors in determining these associations
clearly can only be elucidated by detailed studies.

In conclusion, we suggest that our findings of this large number of Indo-
west Pacific species in western Panama is representative of the poor state of
knowledge of fish distribution throughout western Central America, (Rosen-
blatt and Rubinoff, 1972), and may require reevaluation of the role of distance
in maintaining the geographic isolation of many species of shore fishes with
vagile embryonic or larval stages.

ANNOTATED LIST OF INDO-WEST PACIFIC REEF-ASSOCIATED
FISHES IN THE GULF OF CHIRIQUI

Hemirhamphidae

1 . Euleptorhamphus viridis (Van Hasselt) — Indo-Pacific, widespread
in the eastern Pacific.

Muraenidae

2. Echidna zebra (Shaw) — known from the Indo-west Pacific and Hawaii;
in the eastern Pacific, from Isla del Carmen to Cabo San Lucas, Isla Jaltemba
Mexico, Clipperton Island, nearshore island localities from Costa Rica (UCR
14-38), the Gulf of Chiriqui, and the Perlas Archipelago.

3. Echidna nebulosa (Ahl) — known from the Indo-west Pacific and
Hawaii; and the eastern Pacific from Bahia Muertos (SIO 61-253), Bahia San
Lucas (SIO 67-136), and Manzanillo (UCLA 56-232), Mexico, Cocos Island,
the Gulf of Chiriqui, and the Gulf of Panama.

4. Gymnothorax buroensis (Bleeker) — known from the Indo-west Pacific
and Hawaii. In the eastern Pacific, from Clipperton Island (UCLA 58-289),
Cocos Island, Isla del Cano and Isla Murcielago, Costa Rica (UCR 423-58
and 382-29), and a single specimen (SIO 71-48) collected in 10 meters in a
Pocillopora bank reef at Islas Secas, Gulf of Chiriqui. New record for the
eastern Pacific.

5. Gymnothorax flavimarginatus (Riippell) — abundant in Indo-west
Pacific and Hawaii, and offshore eastern Pacific islands of Clipperton, Cocos,
and Isla del Cano, Costa Rica (UCR 423-125). Observed and photographed,
but not collected at Islas Secas and Islas Contreras, Gulf of Chiriqui.

6. Gymnothorax undulatus (Lacepede) — Indo-west Pacific and Hawaii.
In the eastern Pacific, known only from Isla del Cano, Costa Rica (UCR 423-
59) and the Gulf of Chiriqui. We have collected and/or observed this species

1972

Fishes from the Gulf of Chiriqui, Panama

1 1

at Islas Naranjas, Islas Contreras (SIO 70- 1 35, SIO 7 1 -40), Islas Secas (SIO 70-
136, SIO 70-140), and Isla Coiba (MCZ 44103). New record for the eastern
Pacific. Galapagos listings for this species are based on Herre’s misidentification
of a juvenile Muraena lentiginosa (SU 24382).

7. Enchelynassa canina (Quoy and Gaimard) — Indo-west Pacific and
Hawaii. In the eastern Pacific known from Clipperton Island (SIO 59-12,
UCLA 56-240) and Isla Montuosa, Gulf of Chiriqui (SIO 70-358). New record
for the eastern Pacific.

8. Uropterygius tigrinus (Lesson) — Hawaii, Johnston, and the Society
Islands. In the eastern Pacific, from Isla Espiritu Santo, Gulf of California
(SIO 61-276), Isla Clarion, Islas de Revillagigedo (UCLA 55-131), and Islas
Contreras, Gulf of Chiriqui (SIO 70-135, SIO 71-40). New record for the
eastern Pacific.

Holocentridae

9. Myripristis murdjan (Forsskal) — Red Sea and Indo-west Pacific;
eastern Pacific from the major islands groups, nearshore island localities from
Costa Rica, and the Gulf of Chiriqui.

Kuhliidae

10. Kuhlia taeniura (Cuvier) — Indian Ocean to central Pacific. In the
eastern Pacific, from Cape San Lucas to Colombia. Observed at Isla Montuosa
and other localities in the Gulf of Chiriqui. The name K. arge Jordan and
Bollman is available for the eastern Pacific population. In the absence of a
critical study we tentatively regard it as conspecific with the western Pacific
form.

- Priacanthidae

11. Priacanthus cruentatus (Lacepede) — Pantropical; in the eastern
Pacific, from Cabo San Lucas, Isla Jaltemba, and Islas Tres Marias, Mexico,
the major offshore islands, Panama Bay, and the Gulf of Chiriqui.

Mugilidae

1 2. Mugil cephalus Linnaeus — Cosmopolitan in warm seas; in the eastern
Pacific from Monterey, California, to Chile.

Branch i ostegidae

13. Malacanthus hoedti Bleeker — Indian and tropical Pacific Oceans.
This species, a new record for the eastern Pacific, was observed and collected
at numerous localities in the Gulf of Chiriqui (SIO 70-138, SIO 71-42, SIO 7 1-
53) where it is a common associate of the sand bottom and contiguous reef
community at depths of 10-25 meters. The finding of Malacanthus initiated a
search for additional material in existing collections; as a result of this inspec-
tion we now know that M. hoedti in the eastern Pacific ranges from Costa Rica

12

Contributions in Science

No. 234

Figure 3. Malacanthus hoedti. A 244 mm individual from Isla Cavada, Islas Secas
(SIO 70-138).

(Isla Viradores Sur, sight record) to Gorgona Island, Colombia (Argosy 27,
now at UMML). In the Gulf of Chiriqui we have observed M. hoedti at numer-
ous stations, both near the mainland (Bahia Honda) and at several island
groups (Islas Naranjas, Brincanco, Uva, and Canal de Afuera). M. hoedti was
encountered in pairs (not known to be male-female pairs in that the sexes are
not externally distinguishable) at all localities. When approached by a diver, the
fish would retreat into a burrow head-first. The burrow entrances were at the
edges of large rocks, and the shallow burrows run beneath the rocks and termi-
nate in an enlargement. We have compared our material with a series from
Hawaii (CAS 24823) and a single specimen from the Caroline Islands (CAS
24824). All agree in morphology, number of vertebrae, and coloration,
especially in the distinctively banded caudal (compare Fig. 3 with Berry, 1958,
Fig. 7). There are, however, differences in the mean numbers of dorsal and
anal rays (Table 2). The differences are significant at the P< .05 level but not
at P< .01. Differences of this magnitude could indicate separation of the

Table 2

Total dorsal and anal rays in Malacanthus hoedti. Data for Central Pacific
material include counts from Berry (1958).

E. Pacific
Cent. Pacific

54

55

Total dorsal rays
56 57 58 59 60 61

62

X

95% Conf.
interval

1

1

2 1 3

56.5

± 1.3

112-42

1

59.4

± 1.2

48

49

Total anal rays
50 51 52 53

54

X

95% Conf.
interval

2

-

3 3

49.9

± 1.0

2 2 3 3

1

51.9

±0.9

E. Pacific
Cent. Pacific

1972

Fishes from the Gulf of Chiriqui, Panama

13

populations at the specific or subspecific level. Flowever, there is broad overlap
of the ranges of the dorsal and anal counts. More importantly, our concept of
M. hoedti (sensu stricto) is based on the Hawaiian population (10 of 11 speci-
mens). Until adequate samples from throughout the entire range of the species
are available, it would be premature to give formal taxonomic recognition to
differences between the Hawaiian and eastern Pacific populations.

Carangidae

14. Alectis ciliaris (Bloch) — Indo-west Pacific and Hawaii, widespread
in eastern Pacific. Observed and taken at several localities in the Gulf of
Chiriqui.

15. Caranx melampygus Cuvier — Indo-west Pacific and Hawaii; in the
eastern Pacific, from the major offshore islands and the Cape San Lucas region
of Baja California. Observed and photographed over the reefs at several locali-
ties in the Gulf of Chiriqui.

16. Gnathanodon speciosus (Forsskal) — Indo-west Pacific and Hawaii,
and widespread in the eastern tropical Pacific. Observed and collected at
numerous localities in the Gulf of Chiriqui (SIO 70-136).

Labridae

17. Hemipteronotus pavoninus (Valenciennes) — Indo-west Pacific and
Hawaii; in the eastern Pacific, known from Cabo San Lucas, Baja California,
several island localities in the Gulf of Chiriqui, and Isla Pedro Gonzalez,
Archipielago de las Perlas (as Xyrichthys panamensis Fowler, 1944). We
follow Randall (1965) in placing Iniistius and Xyrichthys in the synonymy of
Hemipteronotus.

18. Hemipteronotus taeniourus (Lacepede) — Indo-west Pacific and
Hawaii; in the eastern Pacific, from Punta Pescadero, Gulf of California (SIO
59-225, SIO 61-252), the Gulf of Chiriqui, and the Archipielago de las Perlas.
New record for the eastern Pacific.

19. Thalassoma lutescens (Lay and Bennett) — Indo-west Pacific; in the
eastern Pacific from San Jose del Cabo (SIO 61-237), the Gulf of Chiriqui, and
the major offshore island groups.

Scaridae

20. Scarus ghobban Forsskal — Red Sea and Indian Ocean to eastern
Pacific. In Panama, common in the Gulf of Chiriqui and the Archipielago de
las Perlas.

21. Scarus rubroviolaceus Bleeker — East Africa to central Pacific and
Hawaii; in eastern Pacific, at the major offshore island groups, Cabo San Lucas,
and in Panama, in the Gulf of Chiriqui and the Archipielago de las Perlas.

Kyphosidae

22. Sectator ocyurus (Jordan and Gilbert — Randall (1961) notes that this

14

Contributions in Science

No. 234

species is a senior synonym of S. azureus Jordan and Evermann from Hawaii.
Known from Hawaii and the Society Islands in the Indo-west Pacific, and in
the eastern Pacific, from Cabo San Lucas to Costa Rica, the Gulfs of Chiriqui
and Panama, and Isla La Plata, Ecuador.

Cirrhitidae

23. Cirrhitichthys oxycephalus Bleeker — Red Sea and Indo-west Pacific;
in the eastern Pacific it extends from the Gulf of California to Colombia, and
the major offshore islands.

24. Oxycirrhites typus Bleeker — Randall (1963) and Morris and Morris
(1967) have discussed the range and synonymy of this species, now known
from the Indo-west Pacific and Hawaii, and in the eastern Pacific from Los
Frailes, Baja California to Isla Gorgona, Colombia. We have observed it in
relatively shallow water (15-20 m) associated with the gorgonian Lophogorgia
cf. alba, at Isla Coiba in the Gulf of Chiriqui, Isla Taboguilla in Panama Bay,
and Isla Viradores Sur, Costa Rica.

Syngnathidae

25. Doryrhamphus melanopleura (Bleeker) — Indo-west Pacific; wide-
spread and common in the eastern Pacific from the Gulf of California to
Panama.

Fistulariidae

26. Fistularia petimba Lacepede — Indo-west Pacific; in the eastern
Pacific from the Gulf of California to Panama.

A ulostomatidae

27. Aulostomus chinensis Smith and Swain — Indo-west Pacific; in the
eastern Pacific from Clipperton, Revillagigedo, and Cocos Islands, and Islas
Contreras in the Gulf of Chiriqui.

Acanthuridae

28. Acanthurus triostegus (Linnaeus) — Indo-west Pacific and Hawaii;
in the eastern Pacific from Cabo San Lucas, Isla Jaltemba, and Islas Tres
Marias, Mexico, to the Gulf of Chiriqui and the offshore island groups.

29. Acanthurus glaucopareius Cuvier — Indo-west Pacific and Hawaii;
in the eastern Pacific from the major offshore islands (except the Galapagos),
Isla Jaltemba, Isla Viradores, and the Gulf of Chiriqui.

30. Acanthurus xanthopterus Valenciennes — Indo-west Pacific and
Hawaii; in the eastern Pacific, from the Gulf of California to Panama. This is
the only surgeonfish species observed at the Perlas Archipelago.

31. Ctenochaetus cyanoguttatus Randall/Briggs (1961:554) lists the
distribution as “Cocos Island. Line Islands to the Marquesas and west to
Aldabra in the western Indian Ocean." This species has been collected in the
Gulf of Chiriqui (SIO 71-40), at Isla del Cano, Costa Rica (UCR 423), and
photographed at Isla Viradores, Costa Rica.

1972

Fishes from the Gulf of Chiriqui, Panama

15

32. Zanclus canescens (Linnaeus) — Widespread in the Indo-west Pacific;
in the eastern Pacific from Las Frailes, Gulf of California (SIO 61-243), Isla
Jaltemba, Islas Tres Marias, the Gulf of Chiriqui, and the offshore islands.

Diodontidae

33. Diodon holacanthus Linnaeus — Circumtropical; widespread in the
eastern tropical Pacific.

34. Diodon hystrix Linnaeus — Circumtropical; widespread in the eastern
tropical Pacific.

Tetraodontidae

35. Arothron hispidus (Linnaeus) — Indo-west Pacific and Hawaii; in
eastern Pacific from Cabo San Lucas to Panama and the offshore islands.

36. Arothron meleagris (Bloch and Schneider) — Indo-west Pacific and
Hawaii; in the eastern Pacific it ranges from Cabo San Lucas to Ecuador and
the offshore islands. Recent evidence (Tyler, Randall, and McCosker, in
preparation) indicates that the polychromatic A. setosus (Smith) is conspecific
with the wide ranging Indo-Pacific species A. meleagris.

Balistidae

37. Melichthys niger (Bloch) — A circumtropical species usually associa-
ted with oceanic islands (Berry and Baldwin, 1968). This species is present at
the offshore islands within the Gulf of Chiriqui (Isla Ladrones and Isla Mon-
tuosa) where it replaces Sufflamen verres (Gilbert and Starks) on the reef.

38. Alutera scripta (Osbeck) — A circumtropical species. In the eastern
Pacific, at the offshore islands and Cabo San Lucas. In Panama, it is infre-
quently seen in the Gulf of Chiriqui and the Archipeilago de las Perlas.

Ostraciontidae

39. Ostracion meleagris Shaw — Indo-west Pacific and Hawaii; in the
eastern Pacific from Cabo San Lucas, Bahia Banderas, Isla Jaltemba, the off-
shore islands, and the Gulf of Chiriqui.

Resumen

Las colecciones recientes nos indican la presencia de un numero de
especies de peces del Indo Pacifico Occidental en el Golfo de Chiriqui. El
veinticuatro por ciento (40) de las especies de peces de arrecifes tambien se
encuentran en el mar Indo Pacifico Occidental. El Golfo de Chiriqui esta fuera
del efecto de afloramiento, como si lo esta el Golfo de Panama; siendo asf rela-
tivamente mas rico en el desarrollo de corales hermatfpicos.

Nueve de las especies del Indo Pacifico Occidental que no han sido regis-
trado en o cerca del continente Americano fueron colectadas: Myripristis
murdjan, Ctenochaetus cyanoguttatus, Gymnothorax flavimarginatus, G.
buroensis, G. undulatus, Enchelynassa canina, Uropterygius tigrinus, Mala-
canthus hoedti, y Hemipteronotus taeniourus. Las ultimas seis de las especies

16

Contributions in Science

No. 234

mencionadas no han sido reportadas como del Pacifico Oriental; aunque
ninguna se encuentra confinada al Golfo de Chiriqui.

Los datos de las siguientes especies del Pacifico Oriental son clasificados
como nulos, basandose en el hecho de que no han sido correctamente identifi-
cados o erroneamente registrados; Br achy somo phis crocodilinus, Gymno-
thorax chilospilus, Callechelys marmoratus, Myrichthys maculosus, Myri-
pristis berndti, Lutjanus kasmira, Runula tapeinosoma, Abudefduf saxatilis
vaigiensis y Antennatus bigibbus.

La distribucion de las siguientes especies endemicas del Pacifico Oriental
Gymnothorax castaneus, Petrotyx hopkinsi, y Paraclinus altivelis se ha exten-
dido hasta Panama. Xyrichthys panamensis Fowler, 1944, es sinonimo con
Hemipteronotus pavoninus (Valenciennes, 1839).

Muchos de los migratorios transpacifico estan restringidos y limitados en
su distribucion Pacifico Oriental. Algunos se encuentran aparentemente
en estrecha relacion asociados con el desarrollo de corales hermatfpicos. No
existe evidencia que nos indique que dichas especies esten desplazando especies
endemicas del Pacifico Oriental.

El numero de especies encontradas por primera vez en el Golfo de
Chiriqui refleja el poco conocimiento en lo que se refiere a la distribucion de
los peces del Pacifico tropico oriental.

Acknowledgments

We thank C. H. Birkeland, T. F. Dana, P. W. Glynn, S. McCosker, and
A. Rodaniche for field assistance, W. N. Eschmeyer (CAS), C. R. Robins
(UMML), and B. W. Walker (UCLA) for permission to examine and use
material in their care, The National Science Foundation (GB 4408), The
Smithsonian Tropical Research Institute, and the U. S. Armed Forces Southern
Command (Rodman Naval Station) for shiptime, and P. W. Glynn for his
critical reading of a draft of this manuscript. Special thanks are due B. W.
Walker, W. Baldwin, and K. S. Norris for information regarding fish distribu-
tion in the eastern tropical Pacific, and W. A. Bussing (UCR) who has allowed
us to cite unpublished Costa Rican records from his collections. Much of this
work was completed during the tenure of J. McCosker as a pre-doctoral fellow
with the STRI, whose support is gratefully acknowledged. These studies would
not have been possible without the continued cooperation of Lt. Col. R. D.
Paredes and Sr. J. L. Obarrio of the Republic of Panama.

1972

Fishes from the Gulf of Chiriqui, Panama

17

Literature Cited

Berry, F. H. 1958. A new species of fish from the western North Atlantic, Dikel-
lorhynchus tropidolepis, and relationships of the genera Dikellorhynchus and
Malacanthus. Copeia (2): 1 16-125.

, and W. J. Baldwin. 1966. Triggerfishes (Balistidae) of the eastern Pacific,

Proc. Cal. Acad. Sci. 34(9):429-474.

Briggs, J. C. 1961. The East Pacific Barrier and the distribution of shore fishes. Evol.
15:545-554.

1964. Additional transpacific shore fishes. Copeia (4):706-708.

1967. Relationship of the tropical shelf regions. Stud. Trop. Oceanogr.

Miami 5:569-578.

1969. The sea-level Panama Canal: potential biological catastrophe.

BioScience 19(l):44-47.

1970. Tropical shelf zoogeography. Proc. Calif. Acad. Sci., 4th Ser., 38(7):

131-138.

Brown, W. L. and E. O. Wilson. 1956. Character displacement. System. Zool. 5:
49-64.

Chave, E. H. 1973. Ecological separation of six species of Hawaiian cardinalfishes.
Copeia (in press).

Clark, H. W. 1936. The Templeton Crocker Expedition of the California Academy
of Sciences, 1932. No. 29. New and noteworthy fishes. Proc. Calif. Acad. Sci.,
fourth ser., 21(29):383-396.

Ekman, S. 1953. Zoogeography of the sea. Sidgwick and Jackson, London. 417 p.
Emerson, W. K. 1967. Indo-Pacific faunal elements in the tropical eastern Pacific,
with special reference to the mollusks. “Venus,” Jap. J. Malacol: 25(3):85-93.
Forsbergh, E. D. 1969. On the climatology, oceanography and fisheries of the
Panama Bight. Bull. Inter-Amer. Trop. Tuna Comm. 14(2):49-385.

Fowler, H. W. 1932. The fishes obtained by the Pinchot South Seas Expedition of
1929, with descriptions of one new genus and three new species. Proc. U.S. Nat.
Mus. 80(6): 1-16.

1938. The fishes of the George Vanderbilt South Pacific Expedition, 1937.

Monogr. Acad. Nat. Sci. Phila. no. 2. 349 p.

- 1944. Results of the Fifth George Vanderbilt Expedition (1941). Monogr.
Acad. Nat. Sci. Phila. no. 6:57-529.

Glynn, P. W. In Press. Observations on the ecology of the Caribbean and Pacific
coasts of Panama. The Panamic Biota: some observations prior to a sea-level
canal. Sym. Biol. Soc. Wash.

, R. H. Stewart and J. E. McCosker. In Press. Pacific coral reefs of

Panama: structure, distribution, and predators. Geol. Rundschau.

Greenfield, D. W. 1965. Systematics and zoogeography of Myripristis in the eastern
tropical Pacific. Calif. Fish and Game 5 1(4):229-247.

, D. Hensley, J. W. Wiley, and S. T. Ross. 1970. The Isla Jaltemba coral

formation and its zoogeographical significance. Copeia (1): 180- 181.

Gunther. A. 1870. Catalogue of the fishes in the British Museum Vol. VIII, Cata-
logue of the Physostomi. London, Taylor and Francis. 549 p.

Herre, A. W. C. T. 1936. Fishes of the Crane Pacific Expedition. Publ. Field, Mus.
Nat. Hist., Zool. ser. 21. 472 p.

Hobson, E. S. 1968. Predatory behavior of some fishes in the Gulf of California.

U.S. Dept. Interior, Bureau Sport Fish, and Wildlife, Res. Report 73. 92 p.
Hubbs, C. L. and R. H. Rosenblatt. 1961. Effects of the equatorial currents of the
Pacific on the distribution of fishes and other marine animals. Tenth Pac. Sci.
Cong., Abst. of Symposium Papers: 340-341.

18

Contributions in Science

No. 234

Jordan, D. S. and B. W. Evermann. 1898. The fishes of North and Middle America:
A descriptive catalogue . . . Panama. Part II. Bull. no. 47, U.S. Nat. Mus.
1241-1936.

Kresja, R. J. 1960. The eastern, tropical Pacific fishes of the genus Blenniolus,
including a new island endemic. Copeia (4):322-3 36.

MacArthur, R. H. and E. O. Wilson. 1967. The theory of Island Biogeography.
Princeton University Press. 203 p.

McCosker, J. E. 1971. A new species of Parapercis (Pisces: Mugiloididae) from the
Juan Fernandez Islands. Copeia (4):682-686.

, and R. H. Rosenblatt. 1972. Eastern Pacific snake-eels of the genus

Callechelys (Apodes: Ophicthidae). Trans. San Diego Soc. Nat. Hist. 17(2):
15-24.

Morris, R. A. and D. E. Morris. 1967. A rare hawkfish Oxycirrhites typus Bleeker
found in Hawaii. Ichthyologica 39(2):71-72.

Morrow, J. E. 1957. Shore and pelagic fishes from Peru, with new records and the
description of a new species of Sphoeroides. Bull. Bingham Oceanogr. Coll.
16(2): 1-55.

Myers, G. S. 1941. The fish fauna of the Pacific ocean, with especial reference to
zoogeograph ical regions and distribution as they effect the international aspects
of the fisheries. Proc. 6th Pac. Sci. Congr. 3, 1939:201-210.

Randall, J. E. 196 1. A record of the kyphosid fish Sectator ocyurus ( = azure us) from
the Society Islands. Copeia (3):357-358.

1963. Review of the hawkfishes (family Cirrhitidae). Proc. U.S.N.M.

1 14(3472):389-451.

1964. A revision of the filefish genera Amanses and Cantherhin.es. Copeia

(2):33 1-361.

1965. A review of the razorfish genus Hemipteronotus (Labridae) of the

Atlantic Ocean. Copeia (4):487-501.

Renner, J. A. 1963. Sea surface temperature monthly average and anomaly charts
eastern tropical Pacific ocean, 1947-58. U.S. Fish and Wildlife Serv. Spec. Rept.,
Fish. No. 442. 57 p.

Rosenblatt, R. H. 1963. Differential growth of the ilicium and second dorsal spine
of Antennatus strigatus (Gill) and its bearing on the validity of A. reticularis
(Gilbert). Copeia (2):462-464.

, and E. S. Hobson. 1969. Parrotfishes (Scaridae) of the eastern Pacific, with

a generic rearrangement of the Scarinae. Copeia (3):434-453.

, and T. D. Parr. 1969. The Pacific species of the clinid fish genus Paraclinus.

Copeia (1): 1-20.

, and I. Rubinoff. 1972. Pythonichthys asodes, a new heterenchelyid eel

from the Gulf of Panama, Bull. Mar. Sci. 22(2):355-364.

, and B. W. Walker. 1963. The marine shore fishes of the Galapagos Islands.

Occ. Pap. Calif. Acad. Sci. 44:97-106.

Schaefer, M. B., Y. M. M. Bishop, and G. V. Howard. 1958. Some aspects of up-
welling in the Gulf of Panama. Inter-Amer. Trop. Tuna Comm., Bull. 3(2):
77-132.

Seale, A. 1940. Report on fishes from Allan Hancock Expeditions in the California
Academy of Sciences. Allan Hancock Pacific Exped. 9(1): 1-46.

Snodgrass, R. E. and E. Heller. 1905. Shore fishes of the Revillagigedo, Clipper-
ton, Cocos and Galapagos Islands. Proc. Washington Acad. Sci. 6:333-427.

Wyriki, K. 1965. Surface currents of the eastern tropical Pacific ocean. Bull. Inter-
Amer. Trop. Tuna Comm. 9(5):27 1-304.

Accepted for publication September 14, 1972

NUMBER 235
DECEMBER 29, 1972

607. 7 3

Cz Lvc?

REVIEW OF THE INSECTIVORA
FROM THE EARLY MIOCENE SHARPS
FORMATION OF SOUTH DAKOTA

By J. H. Hutchison

CONTRIBUTIONS IN SCIENCE

NATURAL HISTORY MUSEUM • LOS ANGELES COUNTY

CONTRIBUTIONS IN SCIENCE is a series of miscellaneous technical papers
in the fields of Biology, Geology and Anthropology, published at irregular intervals
by the Natural History Museum of Los Angeles County. Issues are numbered sep-
arately, and numbers run consecutively regardless of subject matter. Number 1 was
issued January 23, 1957. The series is available to scientific institutions and scien-
tists on an exchange basis. Copies may also be purchased at a nominal price.
Inquiries should be directed to Virginia D. Miller, Natural History Museum of Los
Angeles County, 900 Exposition Boulevard, Los Angeles, California 90007.

INSTRUCTIONS FOR AUTHORS

Manuscripts for CONTRIBUTIONS IN SCIENCE may be in any field of Life
or Earth Sciences. Acceptance of papers will be determined by the amount and char-
acter of new information. Although priority will be given to manuscripts by staff
members, or to papers dealing largely with specimens in the collections of the Muse-
um, other technical papers will be considered. All manuscripts must be recommend-
ed for consideration by the curator in charge of the proper section or by the editorial
board. Manuscripts must conform to those specifications listed below and will be ex-
amined for suitability by an Editorial Committee including review by competent
specialists outside the Museum.

Authors proposing new taxa in a CONTRIBUTIONS IN SCIENCE must indi-
cate that the primary type has become the property of a scientific institution of their
choice and cited by name.

MANUSCRIPT FORM.— (1) The 1972 CBE Style Manual, third edition
(AIBS) is to be followed in preparation of copy. (2) Double space entire manuscript.
(3) Footnotes should be avoided if possible. Acknowledgments as footnotes will not
be accepted. (4) Place all tables on separate pages. (5) Figure legends and unavoid-
able footnotes must be typed on separate sheets. Several of one kind may be placed on
a sheet. (6) An abstract must be included for all papers. This will be published at the
head of each paper. (7) A Spanish summary is required for all manuscripts dealing
with Latin American subjects. Summaries in other languages are not required but are
strongly recommended. Summaries will be published at the end of the paper. (8) A
diagnosis must accompany any newly proposed taxon. (9) Submit two copies of
manuscript.

ILLUSTRATIONS. — All illustrations, including maps and photographs, will be
referred to as figures. All illustrations should be of sufficient clarity and in the proper
proportions for reduction to CONTRIBUTIONS page size. Consult the 1972 CBE
Style Manual, third edition (AIBS) in preparing illustration and legend copy for
style. Submit only illustrations made with permanent ink and glossy photographic
prints of good contrast. Original illustrations and art work will be returned after the
manuscript has been published.

PROOF. — Authors will be sent galley proof which should be corrected and
returned promptly. Any changes or alterations, other than typographical corrections,
will be billed to the author. Unless otherwise requested, page proof also will be sent
to the author. One hundred copies of each paper will be given free to each author or
divided equally among multiple authors. Orders for additional copies must be sent
to the Editor at the time corrected galley proof is returned. Appropriate order forms
will be included with the galley proof.

Virginia D. Miller
Editor

REVIEW OF THE INSECTIVORA
FROM THE EARLY MIOCENE SHARPS
FORMATION OF SOUTH DAKOTA1

By J. H. Hutchison2

Abstract: Five of the seven reported insectivore taxa from
the Sharps Formation (early Arikareean) of South Dakota are
considered valid: Ocajila makpiyahe, Proscalops evelynae,

Quadrodens wilsoni, Domnina greeni, D. dakotensis. One addi-
tional genus of shrew, Trimylus, is added to the fauna. Quad-
rodens is the senior synonym of Palaeoscalopus and is regarded
as a talpid.

Introduction

In two papers covering the early Miocene (early Arikareean) vertebrate
faunas from the Sharps Formation of the Wounded Knee area of south-
western South Dakota, Macdonald (1963, 1970) described six new species
and three new genera of insectivores. A survey of the more recent paper and
subsequent examination of the figured material indicates a need for some
taxonomic revision and re-allocation of several of his specimens.

Methods

Measurements were made with a Gaertner measuring microscope and
are given to the nearest hundredth of a millimeter (mm). Length of lower teeth
equals the maximum possible length of parallel planes normal to a best fit line
along the lingual margin of the tooth, with entoconid perpendicular to plane
of view. Width equals the maximum width between parallel planes parallel to
the length line. All specimens are conserved in either the Museum of Geology,
South Dakota School of Mines and Technology (SDSM) or the Natural History
Museum of Los Angeles County (LACM).

Systematics

FAMILY Erinaceidae Fischer von Waldheim, 1817
SUBFAMILY Galericinae Pomel, 1848
TRIBE Echinosoricini (Cabrara, 1925) Gill, 1872
GENUS Ocajila Macdonald, 1963
Ocajila makpiyahe Macdonald, 1963

Macdonald (1963) initially described Ocajila makpiyahe on a single
dentary fragment with M2-M3, but he made no subfamily assignment of the

1 Review Committee for this Contribution
William A. Clemens
Jason A. Lillegraven
David P. Whistler

2Museum of Paleontology, University of California, Berkeley, California 94720

1

2

Contributions in Science

No. 235

genus. Van Valen (1967: 262) placed Ocajila in the tribe Echinosoricini and
suggested that the type jaw “may represent the otherwise unknown lower
dentition of Brachyerix but is more probably a synonym of the echinosoricine
Lanthanotherium .” The lower dentition of Brachyerix has subsequently been
identified (Rich and Rich, 1971) and is quite unlike Ocajila. The reduction of
the paralophid and low profile of the molars support the placement in the
Echinosoricini and close relationship with Lanthanotherium.

Subsequently Macdonald (1970) referred two new specimens to Ocajila
makpiyahe, one of which he figured. He stated (p. 19) that “the Mi of LACM
9380 represents the first record of this tooth. It is an enlarged version of M2
with no significant variations in the pattern.” LACM 9380 is referable to the
soricid species Trimylus (see below). The second referred specimen (LACM
9491) is referable to O. makpiyahe but contributes no new information.

FAMILY Talpidae Fischer von Waldheim, 1817
SUBFAMILY Proscalopinae K. M. Reed, 1961
GENUS Proscalops Matthew, 1901
Arctoryctes Matthew, 1907

Before any discussion of the evolutionary position of Proscalops evelynae
from the Sharps Formation is undertaken, the probable stratigraphic position
of the types of the named species of Proscalops needs clarification. The type
localities of P. tertius K. M. Reed, 1961, P. terrenus (Matthew, 1907), and
P. secundus Matthew, 1909 lack precise data as to location or formation or
both. However, study of isolated teeth (to be published elsewhere) from known
formations in South Dakota and adjacent states provides a reference to which
the probable chronostratigraphic position of those type specimens may be
determined. K. M. Reed (1961) states that P. tertius came from the “ “White
River fm.,” possibly Brule, “Badlands, South Dakota.” ” The presence of Oligo-
scalops in the lower member (Scenic Member) of the Brule Formation (K. M.
Reed, 1961, and unpublished) and the slightly more advanced character of P.
evelynae from the Sharps Formation which overlies the Brule Formation
indicate that P. tertius is probably from the Poleslide member of the Brule
Formation. This assumption is strengthened by a specimen of the P4-M1,
LACM 1493 from locality 1990, from the Poleslide Member in the Wounded
Knee area that is essentially identical with P. tertius; however, the possibility
that the type of P. tertius is from the lower part of the Sharps Formation or is
conspecific with P. evelynae (see below) cannot be objectively ruled out with
the present small sample sizes.

Macdonald (1963, 1970) has concluded that the type skull of P.
secundus may have come from the upper portion of the Monroe Creek Formation
or the lower Harrison Formation. Although lack of lower teeth of this type
prohibit a refined interpretation of its evolutionary stage, Macdonald’s con-
clusion is in agreement with the supposed position of P. secundus in the bio-
stratigraphic series that has been based upon upper molar and premolar

1972

Insectivora From The Sharps Formation

3

Figure 1. Proscalops evelynae (Macdonald, 1963), LACM 21416, incomplete
rostrum with right I3 to P4 and lingual moieties of M4-M2 and left C to M1; A, palatal
view; B, lateral view of left side. Scale line equals 1 mm.

4

Contributions in Science

No. 235

specializations. P. terrenus is a form species based on the humerus. Specimens
referred to this species range in age from Whitneyan (Poleslide Member of
Brule Formation) to Hemingfordian (Split Rock Formation, Wyoming). Thus,
the concept of P. terrenus more than covers the entire chronostratigraphic
span of all the species of Proscalops named on skulls and jaws. Macdonald
(1963: 170) stated in regard to the type humerus of P. terrenus that “We can
assume that the specimen was found anywhere between Porcupine Creek and
the top of the divide east of Wounded Knee creek. The reference to “Upper
Rosebud” [by Matthew, 1907] probably precludes the possibility that the
type came from beds below the Harrison.” Although Macdonald (1970:24)
later suggested on less objective criteria that the type came from the Sharps
Formation, I accept his original placement. P. terrenus is probably a valid
species. Proscalopine teeth from the Harrison Formation near Agate, Nebraska,
represent either an advanced species of Proscalops (more advanced than P.
secundus ) or a primitive species of Mesoscalops.

At present there are many reasons to assume that the sequence Oligo-
scalops galbreathi (C. A. Reed, 1956) — P. tertius — P. evelynae > P. secundus
— Mesoscalops K. M. Reed represents a phyletic lineage with gaps. O. gal-
breathi was first named on the basis of a humerus as Arctoryctes galbreathi
C. A. Reed, 1956 then later named again as Oligoscalops whitmanensis K. M.
Reed, 1961 on the basis of a skull from Wyoming and a referred jaw from the
same locality as the humerus in northeastern Colorado; there seems to be no
reason to assume that the humerus belongs to a different species than the skull.
P. miocaenus Matthew, 1901 may also belong in the sequence and on size
and available dental characters (K. M. Reed, 1961) would fall between Oligo-
scalops and P. tertius.

Proscalops evelynae (Macdonald), 1963
Domninoides evelynae Macdonald, 1963
Arctoryctes terrenus Matthew, 1907 in part, Macdonald, 1963
Proscalops evelynae (Macdonald), Hutchison, 1968, including
Arctoryctes terrenus of Macdonald, 1963
Proscalops evelynae (Macdonald), Macdonald, 1970
Arctoryctes terrenus Matthew in part, Macdonald, 1970

A previously unpublished rostrum with C-M1 and fragments of the M2
(LACM 21416 from locality 6898 [Fig. 1]), collected by Mr. Robert Machris,
aids in comparing Proscalops evelynae (previously known only from the type,
incomplete mandible, and from humeri) with other proscalopines known from
skulls.

The type mandible (SDSM 5338) of Proscalops evelynae was originally
described under another genus and has not previously been compared in detail
to known jaws of other species of Proscalops. There is a chronological trend in
Proscalops towards increasing hypsodonty of the molars. P. evelynae appears
to be slightly more hyposodont, has greater extension of the enamel down the

1972

Insectivora From The Sharps Formation

5

labial side of the roots, and has narrower anterior cingula than in P. tertius,
P. miocaenus, and Oligoscalops. In size of teeth P. evelynae agrees closely with
P. tertius but is significantly larger than P. miocaenus and Oligoscalops (see
Macdonald, 1963, and K. M. Reed, 1961 for measurements). The low longi-
tudinal crest (entocristid) at the base and lingual opening of the talonid valley
is similar in all three species. P. evelynae differs markedly from Mesoscalops
scopelotemos K. M. Reed, 1960, which has greater crown height, enamel
extension, very high entocristid, and better developed cingular shelf between
the labial bases of the protoconid and hypoconid.

The rostrum (LACM 21416) lacks the tip of the snout and posterolateral
margins of the palate, thus only the left M1 and lingual shelves of the right
M1-2 are preserved. Preceding the P4s, three teeth and two alveoli (one with
root) are preserved on the left side and four teeth on the right side. Following
the dental terminology of K. M. Reed (1961), these are I3, C, P2-3. The I3“P3
are unicuspid and single rooted with ovate to drop-shaped cross-sectional
outlines. I3 is the smallest tooth and canine the largest between the I3 and P4
with P2-P3 subequal in size (Table 1). The P4 supports a single labial blade
and lingual shelf with cusp. A minute cusp on the anterior side and near the
base of the paracone represents the remnant parastyle as in Proscalops tertius,
P. miocaenus, and P. secundus but not Mesoscalops in which it is absent.
K. M. Reed (1961:286) states that the parastyle is absent in all Proscalops and
Mesoscalops but in the types of all Proscalops species my observations indicate

Table 1

Measurement (in mm) of the upper teeth of Proscalops evelynae,
LACM 21416

Left

Right

P4* length

2.27

width

2.28

—

P3* length

0.87

0.87

width

0.70

0.68

P2 length

0.84

0.87

width

0.60

0.61

PI length

0.93

0.94

width

0.74

0.69

C length

—

0.55

width

—

0.48

*P4 maximum length between parallel planes perpendicular to the line connecting
the parastyle and posterior tip of ectoloph (this is not the parameter used but
undefined by K. M. Reed, 1961). Unicuspid tooth length is the maximum cross-
sectional diameter

6

Contributions in Science

No. 235

Figure 2. Proscalops evelynae (Macdonald, 1963), LACM 9362, damaged right
humerus; A, anterior view; B, posterior view. Scale line equals 1 mm.

a small vertical ridge or distinct minute cuspation occurred in the ancestral
area of the parastyle. The lingual cusp is rather posteriorly situated as in P.
tertius but more shelflike, although, not to the extreme as in P. secundus. There
is no indication of a second cusp behind the main lingual cusp as in the unworn
P4 of Mesoscalops. The molars as preserved agree closely with P. tertius in
presence of well-developed metaconules (hypocone of K. M. Reed), proto-
conules (protostyle of K. M. Reed), and minute “hypostyles.” Despite the
qualitative differences in degree of angulation and development mentioned
by K. M. Reed (1961) between the molars of P. tertius and P. secundus , I
believe that at present it is difficult to distinguish such features on worn teeth;
the relative proportions or distinctness of these cusps change significantly due
to differential wear and to stage of wear. Unworn teeth and an analysis of
wear progression in larger samples of the various species are needed to evaluate
these features. The rostrum agrees in detail with those already described by
K. M. Reed (1961) for other species of Proscalops.

Measurements (following Reed and Turnbull, 1965) on 11 humeri (Fig.
2) from the LACM collections show a wide range in variation, with measure-
ments of the smaller specimens ranging from 74 to 88% of the largest; however,
the ratios of these measurements produced ranges essentially identical to those
calculated by Reed and Turnbull (1965) for “ Arctoryctes terrenus ” except

Figure 3. Quadrodens wilsoni Macdonald, 1970; A, LACM 9331 (Type) occlusal
view of Mi-M2; B, SDSM 6244 (referred specimen of Palaeoscalopus) reversed
occlusal outline of Mi-M2, hachures indicate edge of apparently anomalous shear
wear surface. C-E, SDSM 55135 (Type of Palaeoscalopus ); C, reversed occlusal
outline of Mi-M2; D, lingual view of dentary fragment with P4-M3; E, occlusal view
of P4 and antemolar alveoli; F, SDSM 6244, occlusal view of P4 and antemolar
alveoli. Scale lines equal 1 mm.

1972

Insectivora From The Sharps Formation

7

8

Contributions in Science

No. 235

those incorporating the measurement of the proximal shaft width. Reed and
Turnbull (1965:132) noted that this measurement is difficult to duplicate
objectively. The range in gross size is of the magnitude seen within some
living species ( Scalopus aquaticus ) but is unusual for a series from a specific
local area. However, nearly all the measurable specimens came from different
sites of unknown or differing stratigraphic levels. There is no consistent size
relationship with stratigraphic level when both are known.

Dentally Proscalops evelynae differs only slightly from P. tertius and
might be considered conspecific, but it is axiomatic that a better understanding
of the variability of both proposed species needs to be known. Considering the
probably older but uncertain age and locality of P. tertius and small samples
of the other species of Proscalops, it seems prudent to retain P. evelynae as a
distinct species for the time being.

Talpidae incertae sedis

Quadrodens wilsoni Macdonald, 1970:21

Quadrodens wilsoni Harksen, 1967, nomen nudum
Palaeoscalopus pineridgensis Harksen, 1967, nomen nudum
Palaeoscalopus pineridgensis Macdonald, 1970:23

Macdonald (1970) described Quadrodens wilsoni on the basis of a dentary
fragment containing the Mi and trigonid of the M2. He diagnosed the genus
on the basis of the large rectangular Mi with trigonid cusps confined to approx-
imately one third of the trigonid. He stated (p. 21) “This form seems to be
another variation of the “hedgehog” theme.” A few pages later he described a
new genus and species of talpid, Palaeoscalopus pineridgensis, on the basis of
two incomplete dentaries including the P4-M3. He diagnosed the Mi of this
form as having greatly reduced anterolabial cingulum. In discussion Mac-
donald stated that P. pineridgensis is the earliest record of a shrew-mole in
North America.

There is a great similarity between Macdonald’s figures (Figs. 6, 8) of
Quadrodens and Palaeoscalopus and subsequent examination of the types
revealed that these two forms are congeneric and probably conspecific. The
material referred to Palaeoscalopus is slightly smaller than the type of Quad-
rodens, but shows the same overall proportions. The three specimens (Fig. 3)
show some variation in the Mi outlines and Mi anterior cingula, but I regard
these as insignificant for generic allocation. A fourth specimen of Quadrodens
(LACM 9253) consisting of a dentary fragment with M2 and alveoli of P4-M3
was referred to Domnina greeni by Macdonald (1970:21). The structure of
the M2 and situation of the mental foramen is nearly identical in all four
specimens. Although more than one species may be represented, the sample
is too small to meaningfully define even two taxa. The name Quadrodens
wilsoni has page priority over Palaeoscalopus pineridgensis and is more des-
criptive without implication of relationships; thus I chose Quadrodens wilsoni
as the senior synonym.

1972

Insectivora From The Sharps Formation

9

(N ^

W

i o
h §

o

a

£

*

X

T3

i

x

60

c

<u

-a

fN ^
^X
60
G
<U

T3

£

X

X

Vs

a- £

60

C

<u

o

O fc

_ x>

S £

X 3

Q c

C/3

I «

£ x
"o £

a> 3

O. c

go

T5 ^

X o
CS G
C <D
O £
T3 X
o a>
aJ s-

s

m — 1

0 <u
0<

01

©. 2

© £

■-— <5j

<3 Go
Cj -w

© <U

^ £ a
»2 -5

s

©. £

-2

3 Co^3

S ^ <£

a .5 v

H

<N

(N

oo

OO

OO

ON

ON

On

— i

— H

II

II

II

— H

i— i

>— 1

ION

<N

X

X

ION

OO

00

00

ON

ION

xj-

m

m

r*N

xf

NO

(N

<N

ON

ION

ion

X

ON

s

s

S

u

C/3

GO

u

<

Q

Q

<

J

C/3

00

Co ,

X

GO

£ &
S s
Q c

c/3

e S
£ X

o £

« 3

X, C

C/3

ON — i — i

— m >on

— O
ion 00
m m
On On

C/3 U U

Q < <

^ J J

© ©

Q Q

10

Contributions in Science

No. 235

Macdonald has already fully described the teeth allocated to Quadrodens
and Palaeoscalopus; however, the variation in outline of Mi, variable expres-
sion of the Mi anterior cingulum, and slight range in size should be noted
(Table 2). The two preserved P4’s show some variation of the talonid. The
hypoconid of SDSM 55135 (Fig. 3E) is a simple distinct cusp flanked by only
one ridge which joins the hypoconid to the protoconid. However, SDSM 5244
(Fig. 3F) has a subdued hypoconid which forms the posterolabial rim of a
continuous ridge extending posteriorly from the base of the protoconid curving
lingually on the hypoconid to the lingual margin and then deflecting anteriorly,
thus forming (but not quite enclosing) a small talonid basin.

The dentary is most completely represented in SDSM 55135 and SDSM
6244 but has not previously been described in detail. It shows no transverse
curvature and is broadly convex ventrally with the deepest portion below the
molars. Its anterior part tapers quite sharply anteriorly, indicating a short
antemolar region and reduced dentition, although the tip of the dentary is not
preserved. The mandibular symphysis extends posteriorly to below the pos-
terior moiety of the P4. Alveoli anterior to P4 are preserved in both specimens
but are not identical. SDSM 6244 has one posterior alveolus preceded by two
incomplete but apparently subequal and longitudinally aligned smaller ones.
Five alveoli (both incomplete and complete) precede the P4 in SDSM 55135.
The posterior wall of the anterior-most alveolus indicates a rather large, long,
and anteriorly inclined (about 45°) root. This alveolus is followed by three
small and tightly crowded alveoli arranged in a equilateral triangle with two
of the alveoli labial. This clustering is followed by a larger centrally placed
alveolus and P4. There are several ways to interpret these alveoli but analogy
with talpids and progressive erinaceids suggests that the enlarged anterior
alveolus represents an incisor (I2) followed by two to four crowded antemolars
and a P4.

The family assignment of Quadrodens wilsoni is troublesome without
more data on the morphology of the mandible, skull, or skeleton. Chiroptera
are excluded from consideration on the grounds of their probable rarity in the
samples and degree of transverse curvature of their mandibles. Low profile
of the teeth, relatively equal height of the talonid and trigonid of the molars,
reduced molarity of the P4, and overall morphology eliminate from considera-
tion most of the more primitive and highly specialized insectivore suborders
(fide Van Valen, 1967) except the Erinaeceota. Most families of this suborder
except the Erinaceidae and Talpidae are either too generalized in molar
morphology and P4 reduction (Nesophonitdae, Adapisoricidae) or too spe-
cialized (Soricidae, Dimylidae) for close comparison. Plesiosoricids show a
greater and/or more primitive emphasis on development of prominent shearing
paralophids on the lower molars (especially the Mi) in contrast to the rather
bulbus and crushing-like Mi of Quadrodens. Macdonald’s error in describing
the same genus under two different families illustrates the difficulties of
separating the Talpidae from the Erinaceidae on the basis of molars alone
( Talpa incerta Matthew, 1924 is a brachyericine hedgehog). Members of both

1972

Insectivora From The Sharps Formation

1 1

families may be characterized by loss of a distinct hypoconulid and merging
of the hypolophid and entoconid of the molars. Most talpids and many erina-
ceids (Echinosoricinae, Erinaceinae) exhibit quadrate low crowned molars
without great disparity between the talonids and trigonids. Quadrodens xvilsoni
resembles some erinaceids and differs from known talpids in: 1) even convexity
of the lower margin of the dentary; 2) Mi larger than M2; and 3) basined (in
one specimen of two) talonid on P4. Quadrodens resembles the Talpidae and
differs from the Erinaceidae in: 1) lack of paraconid on P4; 2) transverse basal
posterior cingulum and posterolingual accessory cuspid on Mi-M2; and 3)
mental foramen small and not depressed into dentary. Although the anterior
part of the dentary was apparently short, there is no obvious indication in the
area of the P4 suggesting an enlarged incisor as in progressive erinaceids;
however, talpids frequently develop a reduced dentition with prominent but
not greatly hypertrophied lower incisor (I2).

Although the above comparisons may not lead to an obligatory conclusion
of talpid affinities, I believe that Quadrodens wilsoni is a talpid and, on theo-
retical zoogeographical grounds (Hutchison, 1968: 108), this species is probably
aligned with the Proscalopinae. .

FAMILY Soricidae Fischer von Waldheim, 1817
SUBFAMILY Heterosoricinae Viret and Zapfe, 1951
GENUS Domnina Cope, 1873
Domnina greeni Macdonald, 1963
Domnina greeni Macdonald, Hutchison, 1966
Domnina greeni Macdonald, Repenning, 1967
Domnina greeni Macdonald, Macdonald, 1970

This species is still known only from the type specimen; the specimen
subsequently referred to it (Macdonald, 1970) belongs to Quadrodens. The
teeth were described by Macdonald (1963) but the specimen is abraded and
much of the lingual side of the M2 is worn away. Macdonald’s (1963, Fig. 5,
teeth are incorrectly captioned M2_3) illustration of the type indicates a greater
longitudinal compression of the trigonid than exists on the specimen (Fig. 4C).
The mental foramen is partly preserved below and just anterior to the Mi
hypoconid.

For discussion of this species see that of Domnina dakotensis below.

Domnina dakotensis Macdonald, 1970
Domnina dakotensis Harksen, 1967, nomen nudum

This species, known only from the type mandible, was characterized by
its widely open Mi trigonid and closure of the Mi-M2 talonid valleys by the
entoconid crest (entocristid). In addition, there is a postsymphyseal foramen
below the level of the Mi hypoconid and the root of the large incisor extends
posteriorly as far as the Mi hypoconid.

Considering relative temporal and geographic proximity of Domnina

12

Contributions in Science

No. 235

Figure 4. A-B Domnina dakotensis Macdonald, 1970, LACM 9351 (Type), A,
lingual view of dentary fragment with Mi-M2; B, occlusal view of Mi-M2; C,
Domnina greeni Macdonald, 1963, SDSM 5895 (Type), occlusal view of Mi-M2.
Scale lines equal 1 mm.

1972

Insectivora From The Sharps Formation

13

greeni and D. dakotensis, discussion of relationships of these two poorly known
species is necessary. The relative openness of the Mi trigonid of D. dakotensis
is not particularly diagnostic and may be the same as in D. greeni when relative
differences in wear, preservation, and individual variation are considered.
Macdonald (1970) stressed the uniqueness of the closure of the trigonid valley
by the entoconid crest, but I fail to see any difference between the two species
in the closure or height of the crest (compare Fig. 4B with Macdonald, 1963,
Fig. 5). Indeed, Repenning (1967) characterizes the genus by the entoconid
“united to the metaconid by a high ridge (entoconid crest).” Although the
“diagnostic” characters are nullified, there are subtle differences between the
two specimens which may be of specific significance when better samples are
available. The postentoconid valley is comparatively well developed in D.
greeni (as in D. gradata Cope, 1873) but absent in D. dakotensis (Fig. 4A).
The presence or absence of this valley seems to be relatively constant in other
species of shrews; however, if these specimens represent transition to a closed-
valley condition, then variability in this character is significant. The molars of
D. dakotensis appear to be relatively longer (Table 3), more delicate, and have
perhaps a better-developed metastylar ridge on the metaconid than in D.
greeni , although preservation of the D. greeni specimen is not ideal for com-
parison. In the absence of a larger sample of either species, it seems prudent
to tentatively retain both species names on the basis of the above characters.

The closure of the postentoconid valley and delicacy of the molars suggests
that Domnina dakotensis might be on the lineage of Paradomnina Hutchison,
1966, but the relatively greater posterior extension of the incisor root, more
posterior position of the postsymphyseal foramen, and perhaps fewer ante-
molars are specialized characters over the later Paradomnina. There are no
serious obstacles to deriving D. dakotensis from D. gradata.

GENUS Trimylus Roger, 1885
Trimylus sp.

Macdonald (1970:19, Fig. 4) figured a dentary fragment, LACM 9380,
and referred it to Ocajila makpiyake. The deep robust jaw, position of the
mental foramen, and construction of the teeth showed remarkable similarity
to those features in the heterosoricine shrews, especially Trimylus. Subsequent
examination of this specimen confirms its assignment to Trimylus.

Macdonald did not describe this specimen in detail and some diagnostic
features are misleadingly illustrated or not figured. LACM 9380 consists of a
midsection of the horizontal ramus containing Mi-M2. The dentary (Fig. 5) is
deep and robust with a large mental foramen set below the ectoflexus of the
Mi in the posterior end of an elongate depression extending anterodorsad. A
prominent postsymphyseal foramen opens anterolabially near the ventral
margin of the dentary below the junction of the Mi-M2. The dentary is broken
off just anterior to this foramen with the break extending anterodorsally to
just in front of the Mi. No part of the symphysis is preserved. The cavity for

14

Contributions in Science

No. 235

Figure 5. A-C, Trimylus sp., LACM 9380, dentary fragment with Mi-M2; A, occlusal
view; B, lingual view; C, labial view. Scale line equals 1 mm.

the root of the large incisor extends as far as the symphyseal foramen but none
of its external margins are preserved. Remains of one antemolar alveolus (P4)
are preserved just anterior to Mi. There apparently is room for only one or
two additional alveoli between the P4 and Ii.

The molars are graded in size (Table 3) with the Mi about one-third
larger than M2. Mi resembles other Trimylus in its robust proportions and
major features of the trigonid and talonid (see Mawby, 1960, Wilson, 1960,

1972

Insectivora From The Sharps Formation

15

and Repenning, 1967). The anterior cingulum of Mi is well defined but
terminates labially to the lingual extremity of the paraconid. The hypolophid
is well fused to the entoconid but a small hypoconulid persists in early wear
high up the posterolabial flank of the entoconid. M2 resembles Mi but the
trigonid is more closed thus shortening and compacting the tooth. A small
(adventitous?) cuspid on the posterolabial flank of the paraconid occludes part
of the trigonid valley. A slight vertical ridge on the posterolabial wall of the
talonid indicates the position of the hypoconulid. Only the anterior root of the
M3 is preserved. No pigmentation is evident on the teeth.

The massive construction of the jaw, large incisor, posteriorly-situated
and recessed mental foramen, postsymphyseal foramen, strong size gradation
of the robust molars with nearly complete fusion of the entoconid and hypolo-
phid conspire to situate LACM 9380 firmly within the genus Trimylus
(Mawby, 1960, Repenning, 1967).

The Sharps Formation specimen agrees better with near contemporary
forms of Trimylus in North America than with later Miocene and Pliocene
species of North America and Europe in its more anterior location of the
mental foramen and incisor root and retention of vestiges of the hypoconulid
on the molars. The Sharps specimen differs from T. dakotensis Repenning,
1967 in nearly complete fusion of Mi and M2 hypolophjds to the entoconids,
slightly larger size, persistent anterior cingulum on Mi, and slightly more
anterior position of the mental and postsymphyseal foramina. The Sharps
specimen closely resembles T. compressus (Galbreath, 1953) in tooth mor-
phology but differs in slightly more anterior situation of the mental and
postsymphyseal foramina. The Sharps species differs from T. roperi in the
anterior position of the mental and (?) post-symphyseal foramina, more
anterior position of the Ii root, and possibly the greater prominence of the
hypoconulids.

Although nearest in time to Trimylus dakotensis (early Hemingfordian),
the Sharps specimen seems a little closer to T. compressus (Orellan) in those
meager features available, such as the greater coalescence of the Mi-M2 hypo-
lophids and entoconids and persistent Mi anterior cingulum. I think it prudent,
however, to leave specific allocation in abeyance until more diagnostic
material (i.e., antemolar region) is found and a better understanding of the
variation of the named species is obtained.

Summary

Of the seven insectivores reported from the Sharps Formation by Mac-
donald (1963, 1970), I recognize five of which two species are only tentatively
retained as distinct. The shrew, Trimylus, is an addition to the Sharps Forma-
tion fauna. The revised insectivore fauna is as follows:

Family: Erinaceidae

Ocajila makpiyahe Macdonald, 1963

16

Contributions in Science

No. 235

Family: Talpidae

Proscalops evelynae (Macdonald, 1963)

Quadrodens wilsoni Macdonald, 1963

(= Palaeoscalopus pineridgensis Macdonald, 1970)

Family: Soricidae

Domnina greeni Macdonald, 1963
Domnina dakotensis Macdonald, 1970
Trimylus sp.

Owing to the small sample sizes, I have only tentatively retained Pro-
scalops evelynae and Domnina dakotensis as distinct species on the basis of a
few dental characters of questionable significance.

Literature Cited

Galbreath, E. C. 1953. A contribution to the Tertiary Geology and Paleontology
of northeastern Colorado. Univ. Kans. Publ., Paleontol. Contrib., Vertebrata
4:1-120.

Harksen, J. C. 1967. Geology of the Porcupine Butte Quadrangle. S. Dak. Geol.
Survey map with text on reverse.

Hutchison, J. H. 1966. Notes on some Upper Miocene shrews from Oregon. Univ.
Oreg. Mus. Nat. Hist. Bull. 2:1-23.

1968. Fossil Talpidae (Insectivora, Mammalia) from the later Tertiary of

Oregon. Univ. Oreg. Mus. Nat. Hist. Bull. 1 1:1-1 17.

Macdonald, J. R. 1963. The Miocene faunas from the Wounded Knee area of
western South Dakota. Bull. Am. Mus. Nat. Hist. 125(3): 139-238.

1970. Review of the Miocene Wounded Knee faunas of southwestern

South Dakota. Bull. Nat. Hist. Mus., Los Angeles Co. Mus. 8:1-82.

Matthew, W. D. 1907. A lower Miocene fauna from South Dakota. Bull. Am. Mus.
Nat. Hist. 23(9): 169-219.

Mawby, J. E. 1960. A new occurrence of Heterosore x Gaillard. J. Paleontol.
34(5):950-956.

Reed, C. A. 1956. A new species of the fossorial mammal Arctoryctes from the
Oligocene of Colorado. Fieldiana, Geol. 10:305-31 1.

Reed, C. A., and W. D. Turnbull. 1965. The mammalian genera Arctoryctes and
Cryptoryctes from the Oligocene and Miocene of North America. Fieldiana:
Geol. 15(2):99-170.

Reed, K. M. 1960. Insectivores of the middle Miocene Split Rock Local Fauna,
Wyoming. Breviora, Mus. Comp. Zool. 116:1-11.

Reed, K. M. 1961. The Proscalopinae, a new subfamily of talpid insectivores. Bull.
Mus. Comp. Zool. 125(14):473-499.

Repenning, C. A. 1967. Subfamilies and genera of the Soricidae. Geol. Surv. Prof.
Pap. 565:1-74.

Rich, T. H. V., and P. V. Rich. 1971. Brachyerix, a Miocene hedgehog from western
North America, with a description of the tympanic regions of Paraechinus and
Podogymnura. Amer. Mus. Novitates 2477:1-58.

Van Valen, L. 1967. New Paleocene insectivores and insectivore classification.
Bull. Am. Mus. Nat. Hist. 135(5):2 17-284.

Wilson, R. W. 1960. Early Miocene rodents and insectivores from northeastern
Colorado. Univ. Kans. Publ., Paleontol. Contrib., Vertebrata 7:1-92.

Accepted for publication April 4, 1972

Printed in Los Angeles, California, by Anderson, Ritchie and Simon

IONIAN INSTITUTION NOlifUllSNI NVINOSH1IWS S3iaVU8n LIBRARIES

<T .. 2 £ 2

m r; x^oiusjix m ^ i" \^uus^y m

co £ — co \ £ co

)Shiiiaj$ S3iavaan libraries Smithsonian institution NoiiniusNi

CO 2 » *• . CO 2; .... CO

fev 5 , , it 5 2

> X^Oi. 11^2 ^ 'N >

2 00 * 2 </> ** 2

ISONIAN INSTITUTION NOIlDlllSNl NVIN0SH1HNS S3IBVdan LIBRARIES

~ CO — CO “ <

)SHims S3 iava an libraries Smithsonian institution NoiiniusNi

2 r- 2 f~ 2 r

m x^vosv^x m x^vosv^x ^ r

co £ co ± c

■IS0NIAN INSTITUTION NOlifUllSNI NVIN0SH1IWS S3iavaail LIBRARIES

2 \ CO 2 00 2 \ £

^ E E ^

^ ''Y^v >' 2 S

CO -v"' 2 CO 2 00

dshiiws S3 1 ava a 11 libraries Smithsonian institution NoiiniusNi

4/3 _ ^ CO — CO

UJ 1.1 A 2 1,1

HS0N1AN INSTITUTION NOlifUllSNI NVIN0SH1IINS S3IBVaail LIBRARIES

co

s.\

>

PO

^ _ Vb— Ui vo-s—sv - N" M m

dshiiws saiavaan u b rar i esc/5smithsonian_institution° NoiiniusNi

z w z ... cn

< 4!SC... S

> >

2 CO a 2 00 _

HSONIAN INSTITUTION NOIinillSNI NVIN0SH1IWS S3IBVaail LIBRARIES
=; co — co 2 (

a\-

CO

o

O — O

_j ~2L _l Z — 1

ONIAN INSTITUTION NOliniliSNI NVIN0SH1IINS S3 I M VH B 11 LI B RAR I ES SI

_

m ^Dcy ^ NgfruiisgX m N z cn

hiiws^sb I HVd a n~u brar i eswsmithsonian“institution NOliniliSNI N

CD Z 5^2 <

^ § /g^l 1

X f 1 X

[c^7 ^ | ^ i ■§

ONIAN JNSTITUTI0NWN0linXllSNI^NVIN0SHilWS^S3 I U VH a IT_ LI B R AR I ESJ3!

v .'si-'SL-' i . i >^rTcUTr\ /Cva so LlI

!T

< •:

=: \S54M8/ cq

H1IWSZS3 I dVd a II LiBRAR! ES__ SMITHSONIAN^ INSTITUTION ^NOSiniiiSNI^N

o

m W ^L»VAS*^

m x^vosvavx ^ £

IONIAN I N ST I TUT I ON^ NOliniliSNI NVIN0SH1IWS S3 I H V y 8 il 1 B R A R 1 E S S
z v co g ^ g

^ o *%% x <0% ° (s£lM% 5

§s § OsHk | £ III if) o

I# § t >

s > v 2 ^ 7' 2

auiws^sa i ava a n\i b rar i es^smithsonian ^institution ^NoiiniiiSNi_N

)/ X

jjjy. r n^asv^ 5 “r 5

SON I AN”" INSTITUTION NOlinillSNI^NVINOSHilWS S3 I.BVU 3 ll^LI BRAR I ES

/o,

X> V ^ ,

lo&a*

. ^ ^ - Ipw„ po

m xvdv/ ^ m ^ jj]

shiiws^sb i ava a n l i b rar i es ^Smithsonian institution NoiinmsNi
<2 ? 5 £ <

2 ^ * Z ^ *. Z <S>

SONIAN INSTITUTION NOliniliSNI NVIN0SH1IIAIS S3 I a Va a l*l_L I B R AR I ES ■

— co — , v . co z ,

^ x^sovTx LJ ^ w X^I^Tx co X^^Tx L_