co z <o *72. co * 3

LIBRARIES SMITHSONIAN INSTITUTION NOIinilJLSN! NVIN0SH1IWS S3I8V88I1
in zz co — 05

NOIinillSNI NVINOSHIII/NS S318V88I! LIBRARIES SMITHSONIAN INSTITUTION

^ ^ k i /#%\ 1 S ,

30

m v' vsg?' xjvasv^x m

cn £ o>

JBRARIES SMITHSONIAN INSTITUTION NOIinillSNI NVIN0SH1IWS S3I8V88I1
$£ z * oo z co

>joiifUiiSNi_NvmosHims S3 1 avaan_u brar i es Smithsonian jnstitution

£ 3 ^%) ^ 1 1 2

^ O — '^s^vasv^/' o '^*0 sv^x^ ^ O

Z ' ‘ -J Z _j 2

LIBRARIES SMITHSONIAN INSTITUTION NOIinillSNI NVIN0SH1IWS S3 I H VH 8 H
z r* z i- r

4- CO
70
>

70

m ~r N^oIi Is^X m ^ X^zusj^/ f

co £ co \ £ C

NOIinillSNI NVIN0SH1IINS S3I8V8 8S1 LIBRARIES SMITHSONIAN INSTITUTION

Z co Z , CO 2 y <

rf 2 * i I 1 vs#k

«> D.cJ^7 > 2> ^' > 2

in z co - z co

LIBRARIES SMITHSONIAN INSTITUTION NOIinillSNI NVIN0SH1IWS S3I8V88IT

CO ^ co — CO

cn

< >>

o:

m O ^ ^ " _J

NOliniUSNI^NVlNOSHlIINS S3 I 8 V 8 8 IT LIBRARI E S ~ SMITHSONIAN”* INSTITUTION *
r- > z r* z 1

<£rm^ *“ w O x<CvTrn7>v “ O /l0^ I

m s' vw ^ ^vJvasv^x m x^vqsva^x r; m

co £ co £ <0 =

LIBRARIES SMITHSONIAN INSTITUTION NOIinillSNI NVIN0SH1IWS SBiavaaiT.

CO 07 CO Z CO . 1

5.,^ /■

> -V-A^ s „■ £ >

Z CO •*" Z CO z 1

MOlinillCMI MV/IKinQU I I1A1C QTIMWMan J IQDAPIPQ Qy ITUCnWIflM INSTITUTION 1

r Tn Z w * Z ' t/5 z

AR I ES SMITHSONIAN _ I NSTITUTION^NOIXnXIXSNI_NVINOSHXIWS^ S3 I U Vb 8 IT_LI B F

? ,• , . § | i' “

O Vjftpc^X — X^Uu5^X O

2» «j 2 _j 2

IIIISNI^NVINOSHIIWS S3lBV8ail LIBRARIES' SMITHSONIAN INSTITUTION NOll

ARIES SMITHSONIAN INSTITUTION NOIIDIIISNI NVIN0SH11WS S3 I 8 VH 8 11 L I B F
co 2 co z co Z >

co

/:v " t

> 2 ' )vV > ’ 2

TXIISN^NVINOSHIIINS^SHI bVaanZLIBRAR lES^SMITHSONIAN INSTITUTION NOIX
“ v co r; co = co

\ . I » I 2 9 8 ^ 1 1 1

«— ► \^« - ^saMga*-' v p

*as^ O ■ ' W* _ X^H^SV^ o Xlftosv^x ~ o X^osvav>

2 _» 2 _I 2 -*

1ARIES SMITHSONIAN INSTITUTION NOIlfUllSNI NVINOSHiliNS S3I8VH8n LIB

z r .. z **"

m ^ ^ __ ^ ^

nillSNrNVINOSHilWS^SH I B V8 8 1 1~L 1 B R AR I ES ^SMITHSONIAN ~INSTITUTION NOIJ

2 CO Z • w 2 *••• to

< ><^sovaX 2 f? v • 5 x^sv/vT

X

qj CO

c ° “ * ~ ’ _ — — 2

2 ^ ^ > '*2^ S X >

VARIES SMITHSONIAN INSTITUTION NOIXfUIXSNI NVINOSHXIWS SHIbVban LIB

t n — to — co ~

O Vd* oty _ N^uasjix o

OXIXSNI- NVINOSHXIINS^SB I bVbail LIBRAR IES SMITHSONIAN^INSTITUTION NOIJ

r v z _ r- Z Z £

° ~ yf5o53x ° .jgg&v cn °

g Is

- %

►vash^x rn <'\W ^ XjvAsvv^ m X^vosv^ ^ m X^osv^

co £ co £ CO

VARIES SMITHSONIAN INSTITUTION NOIJLfUllSNI NVIN0SH1IWS S3 I 8 V8 8 IT __L I B I
co 2 to

^ s ™ yfrOONo

z «^.*v =5 <fll% § /®«WS V

iW I 2

^ Z /S' t = i

W > ^vAstA^ >* 2 X^os >’ 2

A' z In z co Z co

niilSNl NVIN0SH1IWS S3l8Vd8Il LIBRARIES SMITHSONIAN INSTITUTION NOll

S

LOS

1 1 1

ANGELES

COUNTY

MUSEUM

CONTRIBUTIONS
IN SCIENCE

on ,13

2 I

DECEMBER 3 1, 1969

TABLE OF CONTENTS
and

AUTHOR INDEX

1969

Nos. 155-175

^TTHSO/V^n

!T8 1 8J97I

Los Angeles County Museum of Natural History • Exposition Park
Los Angeles, California 90007

2

CONTRIBUTIONS IN SCIENCE

No. 155.

No. 156.
No. 157.
No. 158.

No. 159.
No. 160.

No. 161.
No. 162.

No. 163.
No. 164.

No. 165.
No. 166.

No. 167.
No. 168.
No. 169.

No. 170.

TABLE OF CONTENTS

Studies on the genus Leptodactylus (Amphibia, Leptodactylidae) III.

A redefinition of the genus Leptodactylus and a description of a new
genus of Leptodactylid frogs, by W. Ronald Heyer. 14 pp. , 5 figs.
January 31, 1969.

Detection of palynological introgression in Salvia (Labiatae), by
William A. Emboden, Jr. 10 pp. , 2 figs. February 27, 1969.

A new species of Anoura (Chiroptera: Phyllostomatidae) from Costa
Rica, by Andrew Starrett. 9 pp. , 2 figs. February 27, 1969.

The status and distribution in Uganda of the white -bellied
kingfisher, Alcedo leucogaster leopoldi, by Herbert Friedmann.

6 pp. , 1 fig. March 27, 1969.

Thoracic cirripedia of the San Diego Formation, San Diego County,
California, by Victor A. Zullo. 25 pp., 77 figs. March 31, 1969.
Observations on the behaviour of two species of honey-guides,

Indicator variegatus (Lesson) and Indicator exilis (Cassin), by A. L.
Archer and Robert M. Glen. 6 pp. March 27, 1969.

Further notes on the African molossid bat Tadarida aloysiisabaudiae,
by R. L. Peterson. 6 pp.,. 2 figs. Aprils, 1969.

The birds of the Sango Bay forests, Buddu County, Masaka District,
Uganda, by Herbert Friedmann and John G. Williams. 48 pp.

April 24, 1969.

Man and Smilodon; a preliminary report on fheir possible coexistence
at Rancho La Brea, by George J. Miller. 8 pp., 4 figs. June 26, 1969.
Upper Pennsylvanian vertebrates from the Sangre de Cristo Formation
of central Colorado, by Peter Paul Vaughn. 28 pp. , 5 tigs. June 26,
1969.

Habits and habitats of certain South American jays, by John William
Hardy. 16 pp. , 4 figs. June 26, 1969.

Early Permian vertebrates from southern New Mexico and their
paleozoogeographic significance, by Peter Paul Vaughn. 22 pp. ,

2 figs. June 26, 1969.

Mammalian organizational systems, by George F. Fisler. 32 pp. ,

1 fig. June 30, 1969.

Taxonomic review of the Mexican skinks of the Eumeces brevirostris
group, by James R« Dixon. 30 pp. , 8 figs. June 30, 1969.

A new stream breeding frog from Oaxaca, Mexico (Anura, Hylidae),
by Ian R. Straughan and John W. Wright. 12 pp. , 5 figs. June 30,
1969.

Taxonomic notes on the Myrm ecocystus melliger complex
(Hymenoptera: Formicidae), by Roy R. Snelling. 9 pp., 3 figs.

June 30, 1969.

TABLE OF CONTENTS

3

No. 171.
No. 172.
No. 173.

No. 174.

No. 175.

The Philippine subgenus Hoploprosopis of Hylaeus (Hymenoptera:
Colletidae), by Roy R. Snelling, 5 pp. , 4 figs. June 30, 1969.
Avian fossils from three Pleistocene sites in central Mexico, by
Hildegarde Howard. 1 1 pp. , 1 fig. June 30, 1969.

Fossil lanternfish otoliths of California, with notes on fossil
Myctophidae of North America, by John E. Fitch. 20 pp. , 4 figs.
June 30, 1969.

Results of the 1968 Avil expedition to Mt. Nyiru, Samburu District,
Kenya. Ornithology, by Herbert Friedmann and Kenneth E. Stager.
30 pp. , 8 figs. June 30, 1969.

System atics and evolution of neotropical salamanders of the
Bolitoglossa helmrichi group, by David B. Wake and Arden H. Brame,
Jr. 40 pp. , 16 figs. June 30, 1969.

4

CONTRIBUTIONS IN SCIENCE

AUTHOR INDEX

Archer, A* L„

No.

160

Brame, Arden H. , Jr.

No.

175

Dixon, James R.

No.

168

Emboden, William A. , Jr.

No.

156

Fisler, George F.

No.

167

Fitch, John E.

No.

173

Friedmann, Herbert

Nos.

158,

162, 174

Glen, Robert M.

No.

160

Hardy, John William

No.

165

Heyer, W. Ronald

No.

155

Howard, Hilde garde

No.

172

Miller, George J.

No.

163

Peterson, R. L.

No.

161

Snelling, Roy R.

Nos.

170,

171

Stager, Kenneth E.

No.

174

Starrett, Andrew

No.

157

Straughan, Ian R.

No.

169

Vaughn, Peter Paul

Nos.

164,

166

Wake, David B.

No.

175

Williams, John G.

No.

162

Wright, John W.

No.

169

Zullo, Victor A.

No.

159

LOS

ANGELES

CONTRIBUTIONS

COUNTY

MUSEUM

IN SCIENCE

SlUMBER 155

January 31, 1969

STUDIES

ON THE GENUS LEPTODACTYLUS (AMPHIBIA,

LEPTODACTYLIDAE ) III. A REDEFINITION OF THE GENUS
LEPTODACTYLUS AND A DESCRIPTION OF
A NEW GENUS OF LEPTODACTYLID FROGS

By W. Ronald Heyer

Los Angeles County Museum of Natural History • Exposition Park
Los Angeles, California 90007

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

INSTRUCTIONS FOR AUTHORS

Manuscripts for the LOS ANGELES COUNTY MUSEUM CONTRIBU-
TIONS 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 and the
form in which it is presented. Priority will be given to manuscripts by staff members,
or to papers dealing largely with specimens in the Museum’s collections. Manuscripts
must conform to CONTRIBUTIONS style and will be examined for suitability by
an Editorial Committee. They may also be subject to critical review by competent
specialists.

MANUSCRIPT FORM.— (1) the 1964 AIBS Style Manual for Biological
Journals is highly recommended as a guide. (2) Typewrite material, using double
spacing throughout and leaving ample margins, on only one side of x 11 inch
standard weight paper. (3) Place tables on separate pages. (4) Footnotes should be
avoided if possible. (5) Legends for figures and unavoidable footnotes should be
typed on separate sheets. Several of one kind may be placed on a sheet. (6) A factual
summary is recommended for longer papers. (7) A brief abstract must be included
for all papers. This will be published at the head of each paper.

ILLUSTRATIONS. — All illustrations, including maps and photographs, should
be referred to as “figures!’ All illustrations should be of sufficient clarity and in the
proper proportions for reduction to CONTRIBUTIONS page size. Permanent ink
should be used in making line drawings and in lettering (do not type on drawings);
photographs should be glossy prints of good contrast. Original illustrations will not
be returned unless specifically requested when the manuscript is first submitted.

PROOF. — Author will be sent galley proof which should be corrected and re-
turned promptly. Changes after the paper is in galley will be billed to the author. Un-
less specially requested, page proof will not be sent to the author. 100 copies of each
paper will be given free to a single author or divided equally among multiple authors.
Orders for additional copies should be sent to the Editor at the time corrected galley
proof is returned; appropriate forms for this will be included when galley is sent.

Dorothy M. Halmos
Editor

STUDIES ON THE GENUS LEPTODACTYLUS (AMPHIBIA,
LEPTOD ACT YLIDAE ) III. A REDEFINITION OF THE GENUS
LEPTODACTYLUS AND A DESCRIPTION OF
A NEW GENUS OF LEPTODACTYLID FROGS

By W. Ronald Heyer1

Abstract: Identification and subsequent study of Leptodac-
tylus pulcher Boulenger indicates that pulcher does not belong to
the genus Leptodactylus, nor can it be allocated to any described
leptodactylid genus.

The genus Leptodactylus is composed of approximately 30
species distributed in the lowlands and intermediate elevations of
Latin America. The genus is distinguished from all other New
World leptodactylids by the following combination of character-
istics: tympanum distinct; toes without web, or with a web vestige
only; maxillary and vomerine teeth present; no posterior projec-
tion of the vomer; mesosternum a bony style; terminal phalanges
claw-shaped. The genus is redefined using characters of external
morphology of adults and larvae; osteology; myology; breeding
call; egg morphology, placement, and development; and chromo-
somes.

A new genus, Barycholos, is described for Leptodactylus pul-
cher Boulenger. Barycholos is distinguished from all other New
World leptodactylids by the following combination of character-
istics: tympanum distinct; tarsal tubercle present; toe disks pres-
ent and lacking a groove on the outer circumference; mesosternum
a calcified style. The genus is defined using characters of external
morphology, osteology, and myology. It is presently known from
the single species pulcher, which is found along the coastal low-
lands of Ecuador. Barycholos is most nearly related to Eleuthero-
dactylus, but the relationship is not close. Barycholos pulcher is
redescribed on the basis of external morphology, and a lectotype
is designated from the type series.

Introduction

The determination of the limits of the genus Leptodactylus based on
re-examination of the named species is critical to a thorough biosystematic
knowledge of the genus. A series of frogs made available by Dr. James A.
Peters from Ecuador including representatives of Leptodactylus pulcher
Boulenger raised special questions regarding the limits of the genus. Subse-
quent study indicated that pulcher is not a Leptodactylus. In addition, the
species cannot be allocated to any described leptodactylid genus. The purpose

Research Associate in Herpetology, Los Angeles County Museum of Natural His-
tory. Present address Division of Amphibians and Reptiles, Field Museum of Natural
History, Chicago, Illinois 60605.

1

2

Contributions in Science

No. 155

of this paper is 1) to redefine the genus Leptodactylus; 2) to establish a new
genus forL. pulcher, and 3) to redescribe the species pulcher.

Leptodactylus Fitzinger, 1826.

Type species: Rana fusca Schneider, 1799.

Diagnosis : The following combination of characteristics distinguishes the
genus Leptodactylus from all other New World leptodactylid genera: tym-
panum distinct; toes without webs, or a web vestige only; maxillary and
vomerine teeth present; no posterior projection of the vomer; mesosternum a
bony style; terminal phalanges claw-shaped.

Description : External Morphology — Adult: Pupil rounded; tongue free
behind; tympanum distinct; vomerine teeth in transverse, arched, or curved
series, partly between or posterior to choanae; vocal slits present in male;
thumb spines present or absent in male; chest spines present or absent in male;
dorsolateral folds present or absent; parotoid glands distinct, diffuse, or absent;
ventrolateral glands present or absent; lumbar (inguinal) glands present or
absent; belly with discoidal fold not extending onto legs; toes fringed or free;
tarsal fold present or absent; usually no finger disks; toe disks present or (usu-
ally) absent, if present, not creased on the anterior border; upper surfaces of
terminal phalanges usually not divided; antebrachial tubercle absent; first
finger about equal to or much longer than second; toe webbing vestigial or
absent; outer metatarsals contained in flesh as sole of foot.

Larvae: Tadpoles generalized; mouth subterminal to almost terminal;
oral disk usually entire; oral papillae continuous except for an anterior gap;

li

denticle row formulae 2, l-i, or i-i; beak well developed; spiracle sinistral;

331-1

2

anal tube median; pattern almost uniform to blotched. (Larvae known for 10
species. See Fernandez and Fernandez, 1921, for example illustrations.)

Osteology: Dorsal process of premaxilla bifid; premaxilla toothed; maxilla
toothed; quadratojugal contacting maxilla; nasals in contact or not; spheneth-
moid complex lightly to heavily calcified or ossified, region between and
anterior to nasals and vomers calcified or not; sphenethmoid contacting optic
foramen or not; frontoparietals in contact at least in part; usually lacking a
fontanelle (a very small one present in a few species); frontoparietal with or
without posterolateral processes; prootic with distinct dorsal ridge, absent in
largest specimens of largest species; crista parotica partially to entirely ossified;
posterior portion of squamosal spur not contacting, contacting, or overlapping
crista parotica; anterior portion of squamosal spur not contacting maxilla;
vomer with three distinct spines, anterior spine spikelike or broadened; vomer
toothed; vomers in contact or not; vomerine teeth overlapping palatine or not;
vomer without distinct posterior process; palatine spearlike to broadened
laterally, with or without a raised ridge ventrally; orbitosphenoid calcified or

1969

A New Genus of Leptodactylid Frogs

3

PREMAXILLA

NASAL

MAXILLA

SPHENETHMOID

FRONTOPARIETAL

ORBIT

PTERYGOID

PROOTIC

SQUAMOSAL

QUADRATOJUGAL

QUADRATE
CRISTA PAROTIC A
EPIOTIC EMINANCE

OCCIPITAL CONDYLE
FORAMEN MAGNUM
PREMAXILLA
MAXILLA

VOMER

PALATINE

SPHENETHMOID
FRONTOPARIETAL
PARASPHENOID
PTERYGOID
OPTIC FORAMEN
ORBITOSPHENOID
OCULOMOTOR FORAMEN
PROOTIC FORAMEN

QUADRATOJUGAL
PSEUDOCOLUMELLA
CRISTA PAROTICA
SQUAMOSAL
QUADRATE

PERILYMPHATIC FENESTRA &
METOTIC FISSURE
PROOTIC

OCCIPITAL CONDYLE

Figure 1. Dorsal and ventral views of the skull of Leptodactylus pentadactylus . Dor-
sal view based on specimen number 435 in the osteological collection of Jay M. Sav-
age (JMS). Ventral view based on JMS 277. Hatching indicates dentigerous zones,
dark areas indicate foramina in the skull. Line equals 10 mm.

4

Contributions in Science

No. 155

Figure 2. Diagrammatic representations of the hyoids of four leptodactylid genera.
A —Barycholos pulcher, B—Eleutherodactylus bransfordii, C—Eleutherodactylus
fitzingeri, D=Eleutherodactylus fleischmanni, E—Lithodytes lineatus, E—Leptodac-
tylus marmoratus, G—Leptodactylus ocellatus. Processes mentioned in text labelled
on the hyoid of B. pulcher. Not drawn to scale.

1969

A New Genus of Leptodactylid Frogs

5

/

0M0

EPI

MESO

XIPHI

Figure 3. Diagrammatic representations of the sternal apparatus of four leptodactylid
genera. A=Barycholos, B—Eleutherodactylus, Q—Lithodytes , T)—Leptodactylus.
Sternal components labelled for Leptodactylus. Not drawn to scale.

cartilaginous, bordering to completely enclosing optic foramen; parasphenoid
dagger-shaped; posterior spur of parasphenoid extending to posterior portion
of skull or not; no odontoids on lower jaw (Fig. 1 ) .

Alary process of hyoid expanded distally or not, narrowly or broadly
stalked proximally; posterolateral process slender to moderately broad;
posteromedial process a slender bony shaft, expanding posteriorly to end in a
short cartilaginous epiphysis (Fig. 2). Eight presacral vertebrae; transverse
process of sacrum directed laterally or posteriorly, slightly expanded. Ilium
with thick crest, usually high; vastus prominence ovaloid, prominent or not.
Omosternum cartilaginous; episternum with a style, at least partly calcified;
mesosternum a bony style; xiphisternum cartilaginous (Fig. 3a). Phalangial
formula of hand 2-2-3-3, of foot 2-2-3-4-3; terminal phalanges claw-shaped.

Myology: Depressor mandibularis broadly originating from the dorsal
fascia, a small slip originating from the squamosal bone, and a few fibers
sometimes originating from the tympanic ring; adductor mandibularis externus
superficialis absent; adductor mandibularis posteriori subexternus present.

Tendons of the gracilis minor and gracilis major passing dorsad to the
tendon of the semitendinosus and adjoining to the posterodorsal tendon of the
sartorius muscle, or similar except for the tendon of the gracilis minor piercing
the tendon of the sartorius, or the tendons of the sartorius, gracilis minor and
major fused together in the same plane; tensor fasciae latae originating as far
anteriad as the iliacus externus muscle on the iliac bone, or not extending as
far anteriad as the iliacus externus; tensor fasciae latae wide, inserting on
cruralis only, cruralis and gluteus, cruralis and adductor longus, or cruralis,
gluteus, and adductor longus muscles; adductor longus small, about 1/5

6

Contributions in Science

No. 155

volume of sartorius to a moderately thick muscle, about equal in volume to the
sartorius; adductor longus always visible in superficial dissection; adductor
longus always inserting on knee.

Breeding Call: Call of single tone, but usually with distinct harmonics;
dominant harmonic being the fundamental (first) or the second harmonic;
frequency shift in dominant present or absent; difference between adjacent
harmonics equalling or not equalling the fundamental. (See Barrio, 1965,
1966, and Fouquette, 1960, for sonagrams.)

Egg Morphology, Placement, and Development: Eggs with or without
melanophores; eggs laid in foam nest; nest floating on water, nest placed in
potholes, or nest formed in holes or burrows (incubating chambers) ; develop-
ment usually involving an aquatic larval stage, but in at least one species, the
larvae metamorphosing within the incubating chamber.

Chromosomes: Haploid number 11; diploid number 22. (Based on four
species, see Barbieri, 1950.)

Content : The genus is composed of approximately 30 species allocated
to five species groups arranged into two subgenera (Heyer, 1968). The sub-
genus Leptodactylus is composed of the Fuscus, Melanonotus, Ocellatus, and
Pentadactylus species groups. The subgenus Adenomera is composed solely
of the Marmoratus species group.

Distribution : Extreme southern United States (Texas) and northwestern
Mexico through Middle America and South America; South America west of
the Andes south to northwestern Peru, east of the Andes south to central
Argentina, southern Bolivia, Paraguay, and Uruguay; Dominican Republic,
Puerto Rico, various Lesser Antillean islands, San Andres, Providencia,
Tobago, and Trinidad.

Barycholos n. gen.

Type species: Leptodactylus pulcher Boulenger, 1898.

Diagnosis : The only leptodactylid genera in the New World having a
bony or calcified style for the mesosternum are Barycholos, Hydrolaetare,
Leptodactylus, Lithodytes, Paratelmatobius, Physalaemus, and Pleurodema.
Barycholos has a tarsal tubercle, distinguishing it from Leptodactylus and
Lithodytes, which have cutaneous folds on the tarsus. Barycholos has a distinct
tympanum, distinguishing it from Paratelmatobius, which lacks a tympanum.
Barycholos has distinct toe disks, distinguishing it from Hydrolaetare, Physa-
laemus, and Pleurodema, which lack toe disks. Superficially, Barycholos
closely resembles some of the small species of Eleutherodactylus, but it does
not have a groove on the outer circumference of the toe disk; Eleutherodacty-
lus does.

Description : External Morphology: Pupil horizontal; tongue free posteri-
orly; tympanum distinct; vomerine teeth in long curved series, posterior to
choanae; vocal slits in males; no thumb adspersities in males; dorsal texture

1969

A New Genus of Leptodactylid Frogs

7

Figure 4. Dorsal and ventral views of the skull of Barycholos pulcher. Hatching indi-
cates dentigerous zones, dark areas indicate foramina in the skull. Line equals 5 mm.

8

Contributions in Science

No. 155

smooth; no distinct glands; belly with discoidal fold not extending onto legs;
fingers and toes free, not fringed; tarsal tubercle present, no tarsal fold; finger
disks small, toe disks larger, no groove on outer circumference of disk; upper
disk surface not divided; no antebrachial tubercle; first finger longer than
second; no webbing between toes; outer metatarsals contained in flesh as sole
of foot. (Based on specimens listed under the redescription of the species.)

Osteology: Dorsal process of premaxilla bifid; premaxilla toothed;
maxilla toothed; quadratojugal contacting maxilla; nasals in contact; anterior
sphenethmoid complex cartilaginous; sphenethmoid not contacting optic fora-
men; no frontoparietal fontanelle; frontoparietal without posterolateral proc-
esses; prootic with dorsal ridge; crista parotica cartilaginous; posterior portion
of squamosal spur overlapping crista parotica; anterior portion of squamosal
spur not contacting maxilla; vomer lacking distinct processes; vomer toothed;
vomerine teeth overlapping palatine; palatine spearlike; palatine without a
raised ridge ventrally; orbitosphenoid a tectum; orbitosphenoid enclosing optic
foramen; parasphenoid dagger-shaped; posterior spur of parasphenoid extend-
ing to posterior portion of skull; no odontoids on lower jaw (Fig. 4) .

Alary process of hyoid rudimentary; posterolateral process small, blunt;
posteromedial process a slender bony shaft, expanding posteriorly to end in a
short cartilaginous epiphysis (Fig. 2a).

Eight presacral vertebrae; transverse process of sacrum directed postero-
laterally, slightly expanded.

Ilium with a high crest; vastus prominence rounded, prominent.

Omosternum cartilaginous; episternum a calcified style; mesosternum a
calcified style, bifid posteriorly, each half bearing a cartilaginous xiphisternum
(Fig. 3a). Phalangial formula of hand 2-2-3-3, of foot 2-2-3-4-3; terminal
phalanges T-shaped. (Osteological observations based on GOV 8015 and
UMMZ 2881.)

Myology: Depressor mandibularis muscle originating in three almost
equal slips from the dorsal fascia, squamosal spur, and tympanic ring; adductor
mandibularis externus superficialis absent; adductor mandibularis posteriori
subexternus present.

Tendons of the gracilis minor and gracilis major passing dorsad to the
tendon of the semitendinosus and not adjoining to the tendon of the sartorius;
tensor fasciae latae not extending as far anteriad as the iliaeus externus; tensor
fasciae latae inserting on cruralis and gluteus; adductor longus thin, straplike,
inserting on adductor magnus muscle. (Myological observations based on
GOV 801 5.)

Content : One species, B. pulcher, with a known range along the coastal
lowlands of Ecuador.

Etymology : The name barycholos is from the Greek /3apuxo\os mean-
ing savage. The new genus is named in honor of Dr. Jay M. Savage.

Relationships'. The relationships among leptodactylid genera are unclear.

1969

A New Genus of Leptodactylid Frogs

9

John D. Lynch is currently reviewing the generic arrangements of members
of the Leptodactylidae. Until this task is completed, the relationships of the
distinctive Barycholos will remain uncertain.

Boulenger placed the species pulcher in the genus Leptodactylus on the
basis of the following combination of characteristics; vomerine teeth present,
toes not webbed, and the sternum with a style. Barycholos and Leptodactylus
are not closely related. The sterna of the two genera are distinctive (Fig. 3a),
and they differ in many other aspects of external morphology, osteology, and
myology. The two genera may also differ with respect to life histories. The
cleared and stained female of B. pulcher contained 43 large ova (about 2.8 mm
in diameter), indicating the likelihood of direct development.

In addition to the possible character of direct development, Barycholos
shares many characters with the composite genus Eleutherodactylus. The
shape of the squamosal spur, T-shaped terminal phalanges, and myological
characters are essentially identical in these two genera. Barycholos is most
closely related to Eleutherodactylus, but the relationship is not a close one.

The only other genus with which Barycholos might be confused is Litho-
dytes, but this is due, not to a similarity of characters, but to the fact that
Lithodytes has been treated in the literature as part of Eleutherodactylus .
Lithodytes shares many characteristics with Leptodactylus, indicating that the
relationships of Lithodytes lie with Leptodactylus rather than Barycholos.

A summary of diagnostic features of Barycholos, Eleutherodactylus,
Leptodactylus, and Lithodytes are presented in Table 1 and Figures 2 and 3.

Barycholos pulcher (Boulenger, 1898)

Leptodactylus pulcher Boulenger: 1898, 122, PI xiv, fig. 4 (Type locality
— Ecuador: Guayas; Puente del Chimbo, railway terminus about 70 miles
from Guayaquil, about 1,000 feet. Syntypes BMNH 1947.2.17.36-38).

Summary of characteristics : Snout rounded, subelliptical, or subovoid
from above; snout rounded in profile; canthus rostralis indistinct; loreal slightly
concave in cross section; tympanum distinct, horizontal diameter Vi diameter
of eye; male vocal slits paired, small, parallel to jaw, from posterior attachment
of tongue to angle of jaw; male external vocal sac large, single, including chest
region; vomerine teeth in long curved series, posterior to choanae; head longer
than wide; head length 33-38.2-42 per cent standard length (SL) ; head width
32-34.7-37 per cent SL; interorbital distance 10-10.9-12 per cent SL; first
finger (thumb) longer than second, second longer than fourth, first shorter
than third; fingers usually free, rarely finger two with a lateral dermal ridge;
finger tips swollen; two very large palmar tubercles at base of hand, three small
round palmar tubercles between the large palmar tubercles and thenar
tubercles; male without nuptial adspersities; no ulnar ridge; supratympanic to
shoulder fold present, rest of dorsum smooth or glandular appearing, only one
specimen with indication of five weak dorsolateral folds; well defined gland

10

Contributions in Science

No. 155

posterior to angle of jaw only; venter smooth; toe tips expanded into distinct
disks; toes entirely free, no lateral ridges or fringes; toes not webbed; sub-
articular tubercles large, globose or pointed; no metatarsal fold; distinct tarsal
tubercle, rest of tarsus smooth; foot with white tubercles; standard length to 29
mm; femur shorter than tibia; femur 40-43.8-50 per cent SL; tibia shorter than,
equal to, or longer than foot; tibia 44-49.6-56 per cent SL; foot equal to or
longer than femur; foot 43-49.0-55 per cent SL; upper lip barred; dark eye
stripe present or absent, if present sometimes expanding into broad dark
axillary band; dorsum with pattern of end to end hourglasses from interorbital
region to sacrum, hourglass pattern sometimes reduced or interrupted medially
by a mid-dorsal light stripe extending from the tip of the snout to the anus;
upper limb surfaces with dark transverse stripes, stripes sometimes faint;
venter scattered with melanophores, usually concentrated under the throat;
posterior surface of thigh profused with melanophores interrupted by small
dotlike spots.

Distribution : The known altitudinal range is from 30 to 600 m.

The coast of Ecuador has a dramatic climatic change along the North-
South axis. The north coastal region is covered with Tropical Rain Forest, the
south coastal region with Tropical Dry Forest. Barycholos pulcher apparently
inhabits the regions along the coastal lowlands intermediate between the
extreme climates and associated vegetations (Fig. 5). If the species is re-
stricted to the transition zone between the Tropical Rain and Dry Forests, its
geographic range may well be limited to western Ecuador.

Specimens Examined'. EL ORO: 7 km SE Buenavista, 30 m, JAP 2464-5;
Gualtaco, GOV 8012; 2 km S Pasaje, 30 m, JAP 2498; ESMERALDAS:
Hacienda Equinox, 38 km NW Santo Domingo de los Colorados, 300 m, JAP
1771-2; Rio Cupa, GOV 8017; Quininde, GOV 8016; GUAYAS: Bucay,
GOV 8018-9; 20 km W Guayas, 100 m, JAP 2524; 5.2 km E Milagro, Haci-
enda San Miguel, 60 m, JAP 2519-20, 2522-23, LACM 42080; MANABI:
Cordillera de Balzar, 9 km E San Placido, km 83 on Quevedo-Manta Hiway
from Manta, 60 m, JAP 8154, 8159-60; 2 km W Desvio, Quevedo-Manta
Hiway, 16 km W Pichincha, in Cordillera de Balzar, 75 m, JAP 8150; PICH-
INCHA: Puerto de Ila, GOV 8013-5, 3838; Rio Toachi, GOV 6474; 6 km
E Santo Domingo de los Colorados, Cmdr. Dyott Farm, JAP 8239, IJMMZ
2881; 18 km W Santo Domingo de los Colorados, J. Ramsey Farm, km 19,
Chone Rd, 200 m, JAP 4034-6, 4105-7, LACM 42081.

Remarks: Boulenger (1898) described Leptodactylus pulcher on the
basis of three specimens, but did not designate one of them as a holotype.
His description, however, is based upon a single specimen, BMNH
1947.2.17.36. This specimen is in the poorest shape of the three. BMNH
1947.2.17.38 is in good condition, and shows the characteristic dorsal pattern
figured by Boulenger. I hereby designate BMNH 1947.2.17.38, an 18.3 mm

< — o

1969

A New Genus of Leptodactylid Frogs

Figure 5. Map of Ecuador showing the distribution of Barycholos pulcher. Open
circle indicates type locality. Dashed line indicates 1,000 meter contour.

Comparison of certain characteristics between four leptodactylid genera. Jaw muscle
abbreviations and data for Eleutherodactylus from Starrett (1968) .

12

Contributions in Science

No. 155

+

Cj

a"

co

h(— )

O

73

03

Ph

Q

o

co

P-i

C3

43

cr

co

co

C

co

Ph

PJ

o

42

H

Q

o

K

cu

K)

<U

73

c

o

O

<u

Ph

c

n 1

o

co

aJ

4a

co

O 1

&

£

73

2

co

'S

C

o

4a

a3

0

c3

cr

s

u

"§

o

+

« <D

O C

u o
a

■§ 2

+

cc -2

Cu

o Pa

H

<

a

00

Q &

^ oo cr
H Q 73

73

QJ

P.

+

O

cq

<u

'PL

73 H
<
& a

aa WD
PP

Sp pH Q

o

o

co O)
<D 73
c o w

CO ^

<L> aj
O co

O b

*h

Dh H

c

o

<u a>

5 g c

<S § £»

Oh Obji

Si O I

C/3 -S o
O' *7 >H

% g O

73 73 m
<u 43 <u
O 03 3

H £ o

o

43

a>

o

o

!h

Ph

J-H

c3

<

<U

Ph

H

co

<U

too

c

13

4a

Ph

13

c

O 3
co 4a

g-g
q a

aJ co

1 3

rp o
73 tc

S £

£ §*

2 S

73 <D
< <U

Adductor mandibularis + +, — +

posteriori subexternus

Adductor longus adductor magnus adductor magnus knee knee

insertion muscle muscle

1969

A New Genus of Leptodactylid Frogs

13

specimen (specimen not dissected to determine sex), as the lectotype of
Leptodactylus pulcher Boulenger.

The presence or absence of vocal sacs and slits in males may vary within
the species. Only one of the specimens at hand (GOV 8017, 24.5 mm) has
vocal slits and a sac. Another specimen lacking vocal slits and sac (JAP 1771,
24.3 mm) was dissected and has large, well developed testes.

Dr. Walker (pers. comm.) kindly provided the following observations
on freshly preserved specimens; most individuals with a tawny-cinnamon area
in groin and another on proximal tibia near the knee.

Acknowledgments

Dr. Charles F. Walker (Museum of Zoology, University of Michigan,
UMMZ) has been generous in sharing his knowledge of leptodactylid genera
and commenting on the manuscript. Dr. James A. Peters (United States
National Museum, GOV, JAP) kindly made all of his Ecuadorian collections
available and allowed me to clear and stain one of the specimens of B. pulcher.
Dr. Alice G. C. Grandison (British Museum (Natural History), BMNH)
graciously loaned the syntypes of L. pulcher. Dr. John W. Wright (Los Angeles
County Museum of Natural History, LACM) made several helpful comments
on the manuscript.

To my major professor. Dr. Jay M. Savage, I owe a debt of thanks.

The research and report were completed while I was a National Science
Foundation Graduate Trainee at the University of Southern California.

Resumen

La identification y posterior estudio de Leptodactylus pulcher Boulenger
indican que pulcher no pertenece al genero Leptodactylus y que ademas, esta
especie no puede ser asignada a ningun genero descrito de leptodactilidos.

El genero Leptodactylus esta compuesto aproximadamente de 30 especies
distribuidos en las tierras bajas y elevaciones intermedias en Centro y Sur
America. El genero se distingue de todos los otros leptodactilidos del Nuevo
Mundo por la viguiente combinacion de caracteristicas: timpano distinguible;
dedos sin membrana o con vestigio de la misma unicamente; dientes maxilar y
vomeriano presentes; el vomer carece de proyeccion posterior; el meso-
esternon es un estilo oseo; falanges terminates en forma de garra. El genero
se define de nuevo haciendo uso de caracteristicas de morfologia externa de
adultos y larvas; osteologia; miologia; canto (entiempo de reproduction);
forma de los huevos; lugar en donde son puertos y desarrollo; y cromosomas.

Un nuevo genero, Barycholos, is descrito para Leptodactylus pulcher
Boulenger. Barycholos se distingue de todos los otros leptodactilidos del Nuevo
Mundo por la siguiente combinacion de caracteristicas: timpano distinguible;
tuberculo tarsal presente; discos presentes en los dedos, sin una ranura en la
circunferencia externa; el meso-esternon es un estilo calcificado. El genero es

14

Contributions in Science

No. 155

definido utilizando caracteristicas de morfologfa externa; osteologia y mio-
logia. A1 presente el genero comprende una sola especie, pulcher, la cual ha
sido encontrada a lo largo de las tierras bajas costeras en Ecuador. Barycholos
esta relacionado con Eleutherodactylus, pero la relacion no es muy cercana.
Barycholos pulcher es descrito en base a la morfologia externa y un leptotipo
se designa de la serie tipo.

Literature Cited

Barbieri, F. D. 1950. Observaciones sobre los cromosomas y espermatozoides en al-
gunos batracios del genero Leptodactylus. Acta Zool. Lilloana, 9:455-462.

Barrio, Avelino. 1965. Afinidades del canto nupcial de las especies cavicolas del
genero Leptodactylus (Anura — Leptodactylidae). Physis, 25:401-410.

. 1966. Divergencia acustica entre el canto nupcial de Leptodactylus ocellatus

(Linne) y L. chaquensis Cei (Anura, Leptodactylidae). Physis, 26:275-277.

Boulenger, G. A. 1898. An account of the reptiles and batrachians collected by Mr.
W. F. H. Rosenberg in Western Ecuador. Zool. Soc. London, Proc., 1898:107-
126.

Fernandez, Kati, and M. Fernandez. 1921. Sobre la biologia y reproduction de
algunos batracios argentinos. I. Cystignathidae. Sociedad Cientifica Argentina,
Anales, 91:97-140.

Fitzinger, Leopold. 1826. Neue Classification der Reptilien nach ihren natiirlichen
Verwandtschaften. Nebst einer Verwandtschafts-Tafel und einem Verzeich-
nisse der Reptilien — Sammlung des K. K. Zoologischen Museum’s zu Wien.
Vienna. 66 p.

Fouquette, M. J., Jr. 1960. Call structure in frogs of the family Leptodactylidae.
Texas J. Sci., 12:201-215.

Heyer, W. R. 1968. Biosystematic studies on the frog genus Leptodactylus. Ph.D.
dissertation, University of Southern California, 234 pp.

Schneider, J. G. 1799. Historiae amphibiorum naturalis et literariae. Jena. I, 264 pp.

Starrett, P. H. 1968. The phylogenetic significance of the jaw musculature in an-
uran amphibians. Ph.D. dissertation, University of Michigan, 179 pp.

Accepted for publication December 24, 1968.

LOS

ANGELES

CONTRIBUTIONS

COUNTY

MUSEUM

IN SCIENCE

JUMBER 156

February 27, 1969

?c> 7, 7 3

Cij l

DETECTION OF PALYNOLOGICAL INTROGRESSION IN
SALVIA (LABIATAE)

By William A. Emboden, Jr.

Los Angeles County Museum of Natural History • Exposition Park
Los Angeles, California 90007

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

INSTRUCTIONS FOR AUTHORS

Manuscripts for the LOS ANGELES COUNTY MUSEUM CONTRIBU-
TIONS 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 and the
form in which it is presented. Priority will be given to manuscripts by staff members,
or to papers dealing largely with specimens in the Museum’s collections. Manuscripts
must conform to CONTRIBUTIONS style and will be examined for suitability by
an Editorial Committee. They may also be subject to critical review by competent
specialists.

MANUSCRIPT FORM— (1) the 1964 AIBS Style Manual for Biological
Journals is highly recommended as a guide. (2) Typewrite material, using double
spacing throughout and leaving ample margins, on only one side of 8 Vi x 11 inch
standard weight paper. (3) Place tables on separate pages. (4) Footnotes should be
avoided if possible. (5) Legends for figures and unavoidable footnotes should be
typed on separate sheets. Several of one kind may be placed on a sheet. (6) A factual
summary is recommended for longer papers. (7) A brief abstract must be included
for all papers. This will be published at the head of each paper.

ILLUSTRATIONS.— All illustrations, including maps and photographs, should
be referred to as “figures!’ All illustrations should be of sufficient clarity and in the
proper proportions for reduction to CONTRIBUTIONS page size. Permanent ink
should be used in making line drawings and in lettering (do not type on drawings):
photographs should be glossy prints of good contrast. Original illustrations will not
be returned unless specifically requested when the manuscript is first submitted.

PROOF. — Author will be sent galley proof which should be corrected and re-
turned promptly. Changes after the paper is in galley will be billed to the author. Un-
less specially requested, page proof will not be sent to the author. 100 copies of each
paper will be given free to a single author or divided equally among multiple authors.
Orders for additional copies should be sent to the Editor at the time corrected galley
proof is returned; appropriate forms for this will be included when galley is sent.

Dorothy M. Halmos
Editor

DETECTION OF PALYNOLOGICAL INTROGRESSION IN
SALVIA (LABIATAE)

By William A, Emboden, Jr.1

Abstract: It has been proposed that Salvia, as it occurs in
western North America, arose through post Pleistocene hybridi-
zation and introgression. For reason of a scanty fossil record,
little is known of this postulated origin. Herein is presented a
technique for detecting instances of introgression using essential
features of pollen grains.

Introduction

Palynology is a relatively young discipline in Botany and many of the
techniques for studying palynological problems have not yet been formulated.
Herein I have presented an approach to the study of extant populations of
Salvia based upon the use of palynological features and stressing size relation-
ships. Pollen presents a variety of morphological features which may be used
as taxonomic characteristics, just as flowers and foliar appendages have been
used in the past. A palynological technique for surveying introgression as an
evolutionary phenomenon in the absence of usable floral or vegetative material
may prove to be a useful way to reconstruct the putative hybrid ancestry of
certain plant genera such as Salvia. This work could presumably be extended
to collections of fossil grains when there are sufficient numbers.

The genus Salvia as it occurs in southwest North America is grouped into
section Audibertia which is comprised of 19 species forming a dominant ele-
ment of the coastal sage community. This association represents a modern
derivative of the Madro-Tertiary Geoflora. Although the presence of Salvia in
this geoflora has been implied, no fossil records are known for vegetative mate-
rial. However, Salvia leaves are less sclerophyllous than most of the other mem-
bers of the association and are thus not as easily fossilized.

Chromosome numbers reported by Epling, Lewis, and Raven ( 1962) cor-
related with distribution records suggest that most members of section Audi-
bertia are relatively young and probably arose through Pleistocene disruption
creating new habitats which could be occupied by hybrid recombinants. The
ability of hybrids thus formed to backcross to parental species could result in
introgressed combinations more capable of existing in the newly-formed habi-
tats than were their progenitors. It is obvious that the California species of
Salvia have had common recent histories, for hybrids between species of di-

iResearch Associate in Botany, Los Angeles County Museum of Natural History,
and Associate Professor of Biology, San Fernando Valley State College, Northridge,
California 91324.

1

2

Contributions in Science

No. 156

Figure 1. Pollen stratification after the terminology of Erdtman.

verse morphology are not uncommon in nature; furthermore, they often show
a high degree of fertility and chromosomal homology. Backcrossing to one or
both of the parental species in disturbed areas leads to hybrid swarms and in
some cases introgression. Despite weak barriers to hybridization, the members
of Audibertia maintain their genetic and morphological identity except in dis-
turbed habitats. Usually these habitats are reconstituted in a few years and the
parental species once more dominate the area. The hybrid habitat as discussed
by Anderson (1948) is therefore a reality only in terms of a few generations.
Even the site of Anderson and Anderson’s famous Salvia study (1954) is no
longer introgressed, for the disturbance which resulted from the planting of
an olive orchard no longer exists, and the dense canopy of the olive trees has
eliminated hybrids which were found “in and among the olives.” (sic.)

Materials and Methods

Eleven interspecific hybrids have been produced by hand pollinations in
the greenhouses of San Fernando Valley State College in Northridge, Califor-
nia. A few of these are not found in nature as their ranges do not overlap or are
not in such proximity as to permit pollinators to effect pollination. Most, how-
ever, do occur in nature and are to be found as F1 hybrids. Introgression in
nature has been documented for only three species in this section. The fertility
of hybrids and backcrosses is rarely less than 50% and is sometimes as high as

1969

Palynological Introgression in Salvia

3

Figure 2. Reticulate stratification of the sexine of Salvia.

98%, as determined by staining pollen with lactophenol blue. Seed set seems
normal but the number of viable seeds is less than 3 % .

It has been ascertained by Emboden (1964) that species of Audibertia
have pollen morphology which is sufficiently distinct to distinguish them one
from another on the basis of size, shape and exine stratification. It was thought
that the pattern of introgression might be studied by using scatter diagrams,
which have been so useful in portraying introgression when morphological
characters have been used. If this should prove accurate in portraying con-
temporary introgression, it would be important in establishing introgression
as a factor in speciation where Post-Pleistocene fossil grains can be found in
abundance.

The hexacolpate pollen of Salvia is distinct from other genera of the
coastal sage community, and indeed from most pollen. This grain is thought to
have arisen from the tricolpate condition through a protrusion of the thinner
walls of the three colpi. Using the terminology of Erdtman (1958) (Figures 1
and 2), Salvia pollen may be described as medium to large, suboblate to pro-
late, having a reticulated exine of continuous or discontinuous muri which

4

Contributions in Science

No. 156

Figure 3. Pollen grains: upper-— Salvia mellijera; lower — S. apiana. Polar, right col-
umn, and equatorial, left column, views.

may be intersected by muroid ridges or may contain a few isolated pila. (Fig-
ure 3.) This description of pollen of section Audibertia is consistent with
Panelatti’s (1960) characterization of pollen from Central and South America,
and western Europe. It has been commonly observed that the size of contem-
porary and fossil pollen grains may vary with treatments; however Faegri and
Deuse (1960) state, “. . . there are important pollen types within which a
further taxonomic differentiation cannot be made by means of qualitative mor-
phology, and where more or less successful attempts have been made to differ-
entiate by means of size statistics,” Inasmuch as pollen grains of the California
Salvias exhibit size extremes that are often considerable, plotting polar axis
against equatorial axis (ordinate and abcissa) might give a picture of hybridiz-
ing entities as distinct groupings, with hybrids as intermediates and backcross-
ing showing some sort of continuum between the F1 individual and one of the
parents. Such information might give insight into the origins of some of the

1969

Palynological Introgression in Salvia

5

Figure 4. Flowers of Salvia apiana, S. mellifera and hybrids: 1. S. apiana; 2. S. melli-
fera; 3. Fi hybrid; 4, 5, 6, introgressants.

species in this section. Miocene grains found on the Kenai Peninsula of Alaska
compare favorably with those of a few relictual species of the present (Em-
boden, 1964).

Throughout coastal southern California, there are numerous areas in
which the overlapping ranges of S. mellifera and S. apiana have resulted in
the production of hybrids. Where major disturbances have been observed,
hybrids and backcrosses generally are found. Such a site is in the San Gabriel
Mountains above Arcadia. This is one of the two sites studied by Anderson
and Anderson (1954) . In contrast is a site in the Santa Susana Mountains three
miles north of San Fernando Valley State College on Zelzah Avenue, North-
ridge, where populations of S. apiana and S. mellifera coexist in an area of
about one square mile. The former occupies the steep well-drained slopes fac-
ing south and west, while the latter occupies a more mesic, north-facing ravine
and the lower slopes. A single putative F1 hybrid is known from this area.

Pollen grains were treated with lactophenol blue and allowed to remain
in solution for 24 hours. Fifty grains were measured from each individual. This
treatment introduces an increase in size of grains and is not comparable to
grains treated by acetolysis and mounted in silicone oil. Specimens of the plants

ngth of Corolla ( mm )

6

Contributions in Science

No. 156

• fc*23

13 15

'10

S*5# -1

*

24 , 17 14

^5

Zelza 1967

PUTATIVE FI

A13 AO A10

Al4 A15
A9

All A12

A6 A 1

46 8 10 12 14 16

Length of A nd roeciu m (mm)

Figure 5. Scatter diagram of populations of S. apiana (upper right) and S. mellifera
in the absence of introgression.

from which the pollen was taken are on deposit in the herbarium of San Fer-
nando State College, Northridge, California.

Results

Scatter diagrams based on floral characteristics of corolla length vs andro-
ecial length (characters found to be strongly correlated with other characters
of the respective species) demonstrate that an introgressed population such as
that in Arcadia is easily distinguished from populations which have no known
history of introgression (Figures 5 and 6). When pollen of individuals from
the Zelzah and Arcadia sites were compared on scatter diagrams using the
polar and equatorial axes of the pollen grains as ordinate and abcissa, a similar
phenomenon was observed (Figures 7 and 8). Introgression in the Arcadia
population was manifest as conspicuous backcrosses to the F2 hybrid and sub-

1969

Palynological Introgression in Salvia

7

o

zr 16
o
o
V

Arcadia 1967

•

•

5

•

1

34

•

16

•

22

•

•

8

20

• • •

•

18 3 13

. 33
12

•

• •

19

9 \7

25

27

Length of Androecium (mm)

Figure 6. Scatter diagram of an introgressing population of S. apiana and S. mellifera.

sequent hybridization within the sub-mellifera population. The area occupied
by S. mellifera individuals on the scatter diagram is considerably increased in
the direction of S. apiana.

In both sets of scatter diagrams, a comparison of the position of indi-
viduals reveals an impressive correlation. That is, palynological evidence
drawn from size relationships is able to portray introgression as clearly as
morphological characters and is strongly correlated with them.

Conclusion

It would seem from the evidence presented, that the postulated post Pleis-
tocene hybridization which gave rise to section Audibertia could be substan-
tiated by a careful analysis of fossil Salvia pollen. It has already been estab-
lished (Emboden, 1964) that both contemporary and fossil grains of Salvia
can be identified on the basis of size and stratification differences. This new
tool for analyzing palynological introgression would not only permit a depic-
tion of hybridization, but would establish the role of introgression in evolu-

P Axis

8

Contributions in Science

No. 156

55

/\TlO M9

M 7

• •

Ml M8

M5

PUTATIVE
FI HYBRID

Zelza 1967

A19 A14

A16 *

A* 7 A20

• • *

A13A12 All

A15

A18

60

70

E Axis

Figure 7. Populations of S. apiana and S. mellijera based on pollen measurements.
Introgression is absent.

tion. An extensive collection of fossil Salvia pollen is not at this time available.
When a number of these hexacolpate grains of two species can be taken from a
single site, the application of this technique should prove most valuable. A
knowledge of the origin of a dominant taxon of the coastal sage association
would be a major contribution toward understanding the evolutionary proc-
esses by which the whole association came into being.

Acknowledgment

I wish to thank Mrs. Frances Runyan for preparing the illustrations in
figures 1, 2, and 4, and Mr. Ira Pearlman for figure 3.

1969

Palynological Introgression in Salvia

9

Arcadia 1967

28

14 9

• *

15 27

, f 16 11 12

^ *17 33

23*%

30 — — —

45 50 55 60 65 70 75 80

E Axis /i

Figure 8. Palynological introgression between S. apiana and S. mellifem.

Literature Cited

Anderson, Edgar, 1948. Hybridization of the habitat. Evolution 2: 1-9.

Anderson, Edgar, and B. R. Anderson. 1954. Introgression of Salvia apiana
and Salvia mellifera. Missouri Bot. Garden, Ann., 41: 329-338.

Axelrod, D. I. 1958. Evolution of the Madro-Tertiary Geoflora. Bot. Rev.
24: 433-509.

Emboden, W. A. 1964. Pollen morphology of the genus Salvia, Section A udi-
bertia. Pollen et Spores, 6: 427-536.

Epling, Carl and H. Lewis, and P. H. Raven. 1962. Chromosomes of Salvia :
Section Audibertia. Aliso, 5: 217-221.

Erdtman, Gunnar. 1958. A Uniform Terminological Basis in Pollen and
Spore Morphology, Zurich.

10

Contributions in Science

No. 156

Faegri, Knut, and Paul Deuse, 1960. Size variations in pollen grains with
different treatments. Pollen et Spores, 2: 293-298.

Panelatti, Gisele. 1960. Quelques resultats de palynologie descriptive et
analytique pour le Maroc. Pollen et Spores, 2: 183-234.

Accepted for publication December 1, 1968.

LOS

ANGELES

COUNTY

MUSEUM

CONTRIBUTIONS
IN SCIENCE

Jumber 157

February 27, 1969

\507'7J

\Czt9Cf

A NEW SPECIES OF AN OUR A

( CHIROPTER A : PHYLLOSTOMATIDAE) FROM COSTA RICA

By Andrew Starrett

Los Angeles County Museum of Natural History • Exposition Park
Los Angeles, California 90007

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

INSTRUCTIONS FOR AUTHORS

Manuscripts for the LOS ANGELES COUNTY MUSEUM CONTRIBU-
TIONS 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 and the
form in which it is presented. Priority will be given to manuscripts by staff members,
or to papers dealing largely with specimens in the Museum’s collections. Manuscripts
must conform to CONTRIBUTIONS style and will be examined for suitability by
an Editorial Committee. They may also be subject to critical review by competent
specialists.

MANUSCRIPT FORM.— (1) the 1964 AIBS Style Manual for Biological
Journals is highly recommended as a guide. (2) Typewrite material, using double
spacing throughout and leaving ample margins, on only one side of 8V2 x 11 inch
standard weight paper. (3) Place tables on separate pages. (4) Footnotes should be
avoided if possible. (5) Legends for figures and unavoidable footnotes should be
typed on separate sheets. Several of one kind may be placed on a sheet. (6) A factual
summary is recommended for longer papers. (7) A brief abstract must be included
for all papers. This will be published at the head of each paper.

ILLUSTRATIONS. — All illustrations, including maps and photographs, should
be referred to as “figures!’ All illustrations should be of sufficient clarity and in the
proper proportions for reduction to CONTRIBUTIONS page size. Permanent ink
should be used in making line drawings and in lettering (do not type on drawings);
photographs should be glossy prints of good contrast. Original illustrations will not
be returned unless specifically requested when the manuscript is first submitted.

PROOF. — Author will be sent galley proof which should be corrected and re-
turned promptly. Changes after the paper is in galley will be billed to the author. Un-
less specially requested, page proof will not be sent to the author. 100 copies of each
paper will be given free to a single author or divided equally among multiple authors.
Orders for additional copies should be sent to the Editor at the time corrected galley
proof is returned; appropriate forms for this will be included when galley is sent.

Dorothy M. Halmos
Editor

A NEW SPECIES OF ANOURA

( CHIROPTER A : PHYLLOSTOMATIDAE) FROM COSTA RICA

By Andrew Starrett1

Abstract: A new species of glossophagine bat, closely re-
lated to Anoura cultrata Handley, is described. This species, here-
in named Anoura werckleae, was found associated with the mal-
vaceous tree, Wercklea lutea, and is so far known only from
Lower Montane Rain Forest in the Cordillera de Talamanca of
Costa Rica.

In 1961, Richard S. Casebeer and Ronald B. Linsky netted, at approxi-
mately 2500 m on the Pacific slope of the Cerro de la Muerte massif, in the
Cordillera de Talamanca of Costa Rica, an orange-brown adult male glosso-
phagine bat with yellow pollen in its fur. The only obvious sources of the
pollen were the large yellow flowers of Wercklea lutea Rolfe (Malvaceae)
growing along a nearby stream. In 1963, when the Los Angeles County
Museum of Natural History mammalian ectoparasite project (LACM-
USACR; see Starrett and Casebeer, 1968) field party was collecting in that
area, the same stream locality was visited in the hopes of obtaining more speci-
mens of this bat which had proven to represent an undescribed species of
Anoura. We succeeded in capturing an adult female, also with yellow pollen in
the fur on the head and shoulders, in a mist net set beneath flowering branches
of a Wercklea, and thus apparently confirmed the supposition that this new bat
visited flowers of this tree. Later microscopic comparison of pollen grains
taken from the fur of the second specimen with some which I collected from
the Wercklea flowers at the same time established this as fact.

Because of its association with the flowering tree, Wercklea lutea, this
species is named

Anoura werckleae, new species

Holotype : LACM 15186, adult male, alcoholic with skull removed,
collected July 12, 1961, by Richard S. Casebeer and Ronald B. Linsky, 6.8 mi.
S restaurant “La Georgina” along Interamerican Highway, 2500 m, Cerro de
la Muerte massif, Province of San Jose, Costa Rica, original number 1246 AS
(A. Starrett) .

Paratype : LACM 25438, adult female, alcoholic with skull removed,
collected July 16, 1963, by Andrew Starrett and Richard S. Casebeer (LACM-
USACR field party), type locality, original number USACR 0-2118.
Distribution : This species is known only from the type locality.

Research Associate, Los Angeles County Museum of Natural History, and Depart-
ment of Biology, San Fernando Valley State College, Northridge, California 91324.

1

2

Contributions in Science

No. 157

Diagnosis'. A robust member of the genus Anoura, closely related to
A. cultrata Handley (1960), with dense orange-brown pelage and with the
distal end of metacarpal III approximately even with the distal end of the first
phalanx of digit V when wing is folded. Braincase and rostrum broad, zygo-
mata well ossified. Postdental palate expansive, with incision in posterior
margin V-shaped. Basisphenoid pits expanded anterolaterally beneath ali-
sphenoids, partially covered by a raised shelf from the latter. P2 with occlusal
edge flattened and without obvious point, enlarged P2 similar in profile to P3,
the highest point not obviously posterior to the center of the tooth.

Description : Pelage dense, velvety, longest individual hairs 8.0 mm long
on dorsum and chest, about 4.0 mm on abdomen. Coloration generally rich
orange-brown throughout, somewhat paler ventrally; individual hairs of dor-
sum essentially tricolored with basal two-thirds whitish-yellow gradually dark-
ening distally to more orange (brownish-orange), a subterminal orange-brown
band (1 mm), and white or silvery tips (±1 mm). Fur on head similar but
shorter, lacking light tips. Individual hairs on chest similar to those of back but
with subterminal band less distinct, especially near the midventral line; hairs of
neck, chin and interramal area similar to those of chest but shorter. Hairs of
abdomen browner than those of dorsum, each with longer terminal band which
is light tan instead of white. Forearm heavily furred on proximal one-third
with hairs 2-3 mm long, propatagium furred to about same level, endopatagium
with hairs extending out to a line reaching from elbow to knee. Base of thumb
heavily furred, bases of metacarpals sparsely so. Tibia and most of uropata-
gium rather heavily furred dorsally (hairs 2-3 mm long) with uniform light
orange-brown hairs, more sparsely furred ventrally and distally on legs. Wing
membranes dark brown or blackish; uropatagium, ears and noseleaf light
brown, contrasting with color of wing membranes. (The foregoing color
descriptions were written originally after the paratype had been in preservative
only a short while. Comparison of the paratype with the holotype and with
color photographs taken of the latter in the field indicated that little if any
color change had occurred at the time the description was recorded.)

Head and body short, about 65 mm (in alcohol), ears and hind feet
moderately small, wing elements of roughly the same size as those of A. cul-
trata, but larger in relation to head and body length and differing somewhat in
proportions. Third metacarpal about equal in length to forearm and to meta-
carpal plus first phalanx of digit V when wing is folded. Thumb approximately
1 1 mm in length. Cranium with braincase long and expanded, rostrum rela-
tively broad, short and flat with a rather abrupt rise to top of braincase.
Zygomatic arches well-developed, complete, with jugal strong and well-ossi-
fied. Palate relatively short and broad, the postdental portion expansive with
V-shaped incisions in the posterior margins. Basisphenoid pits well formed and
expanded anterolaterally beneath alisphenoids, covered along anterior and
lateral margins and separated from alisphenoid depressions by a distinct raised
shelf or septum. Upper molar teeth appearing narrow relative to palate width.

1969

New AN OUR A From Costa Rica

3

generally quadrangular in occlusal view and arranged in parallel or somewhat
arched rows. P2 short, with occlusal surface flattened rather than pointed; P2
nearly symmetrical and similar to P3 in profile.

Measurements (Holotype and paratype, respectively, in mm) :

External: forearm, 40.7, 43.2; tail, 3, 3; hind foot, 12.2, 11.9; ear from
notch, 14, 14; tibia, 15.7, 16.8; thumb, 11.2, 11.3; metacarpal III, 40.6, 43.1;
metacarpal V, 32.7, 34.0.

Cranial: greatest length of skull, 25.8, 26.2; condylobasal length, 25.1,
25.4; length of maxillary toothrow (C-M3), 9.0, 9.2; palatal length, 12.5, 12.8;
zygomatic breadth, 10.8, 10.6; greatest rostral breadth, 5.3, 5.3; postorbital
constriction, 5.4, 5.3; breadth of braincase, 10.2, 10.1; alveolar breadth across
upper canines, 5.2, 5.0; breadth across M3-M3, 6.1, 6.1; length of mandibular
toothrow (C-M3), 9.4, 9.7.

Comparisons : Anoura werckleae is most similar to Anoura cultrata
Handley and the two species are distinguished from other members of the
genus in essentially the same ways (Handley, 1960). A. werckleae can be dis-
tinguished from A. cultrata by the following characteristics (those of A. cul-
trata in parentheses): orange-brown pelage (shiny blackish), distal end of
metacarpal III approximately equal with distal end of first phalanx of digit V
when wing is folded (metacarpal III not reaching end of phalanx), postdental
palate expansive, incision in posterior margin V-shaped (postdental palate less
expansive, incision U-shaped) (Fig. 1), basisphenoid pits expanded antero-
laterally beneath alisphenoids and partially covered by a raised shelf (basi-
sphenoid pits depressions with only slight anterior expansions and without
raised shelf) (Fig. 1), P2 rather nondescript, without point (P2 more like P3,
with obvious point), and P2 nearly symmetrical and similar in profile to P3
(highest point of P2 obviously posterior to center of tooth) . Anoura werckleae
differs from A. cultrata further in having: shorter thumb, absolutely and rela-
tive to forearm length; relatively lower, broader braincase and shorter, broader
rostrum when compared sex for sex (Fig. 2); stronger zygomatic arches;
broader postorbital constriction; greatest rostral breadth equal to or less than
postorbital constriction; upper molars quadrangular in occlusal view with
lingual faces flattened rather than uneven and rounded. This last character is
especially obvious in the M3s, which are essentially triangular in A . cultrata.

The one specimen of A. cultrata known from Costa Rica (Carter, et al. ,
1966) approaches A. werckleae in some measurements and dental features,
but in diagnostic characters, as well as in most of the others listed above, it
is purely A. cultrata and, in my opinion, does not represent a morphological
intermediate between the two species.

A new species of Anoura from South America, also belonging to the
A. cultrata group, currently being described by Dilford C. Carter (1968), is
readily distinguishable from A. werckleae on the basis of fur color (similar to
that of A. cultrata ) and size (smaller in external and cranial measurements).
In addition, Carter’s new species appears to have a relatively shorter rostrum

4

Contributions in Science

No. 157

A B

Figure 1. Posterior palate and basisphenoid region of cranium of A. Anoura werck-
Icae, sp. n., LACM 15186, $, (holotype). B. Anoura cultrata Handley, USNM
309398, $, (paratype).

and maxillary toothrow than does A. werckleae. In most other diagnostic
features, with the possible exception of the characters of the basisphenoid pits,
the South American form is separable from A. werckleae in the same ways as
is A . cultrata.

Specimens Examined: Anoura werckleae, LACM 15186, $, (holotype),
LACM 25438, 9, (paratype); Anoura cultrata , USNM 309398, $, and
USNM 309401, 9, (paratypes), TCWC 9829, $; Anoura sp. n. Carter,
TCWC 11881, 9, (paratype).

Remarks : In addition to being separable on the basis of color and other
morphological characters, the two Central American species of the Anoura
cultrata group, A. cultrata and A. werckleae, appear to have distinct ecological
distributions. A. werckleae is known only from 2500 m, near the upper limit
of Lower Montane Rain Forest, in Costa Rica. (Capitalized forest formation
and life zone designations are those of L. R. Holdridge, from Holdridge and
Budowski, 1959; Tosi, 1965). A. cultrata has been taken only in Costa Rica
and Panama, at elevations ranging from 594 to 1615 m (Carter et al., 1966;

1969

New ANOURA From Costa Rica

5

Figure 2. Crania of Anoura werckleae, sp. n. and Anoura cultrata Handley, dorsal
view. A. A. werckleae, LACM 15186, $ , (holotype). B. A. werckleae, LACM 25438,
$ , (paratype). C. A. cultrata, TCWC 9829, $ , (Costa Rica). D. A. cultrata, USNM
309398, $ , (paratype). E. A. cultrata, USNM 309401, $ , (paratype).

6

Contributions in Science

No. 157

Handley, 1966), in evergreen forest which is probably mostly within the wetter
portions of the Subtropical (Pre-montane) Zone. (The elevation of the type
locality of A. cultrata was recorded as 975 m in Handley, 1960, and was
apparently corrected to 594 m in Handley, 1966; see also Fairchild and
Handley, 1966.) The third species of Anoura known from Costa Rica and
Panama, A. geoffroyi Gray, has an altitudinal distribution similar to that of
A. cultrata ; the two species have been recorded from the same locality at least
once in Panama (LACM data and Handley, 1966). If, as the small amount of
available evidence may suggest, A. werckleae replaces one or both of the other
species of Anoura at higher elevations, then the new species may well turn out
to be limited in distribution to the Cordillera Central of Costa Rica and the
Cordillera de Talamanca, of which the Chiriqui massif in western Panama
essentially forms the eastern end.

Wercklea lutea is apparently restricted to wet situations, such as stream
edges, in Costa Rica (and perhaps western Panama), possibly only in Lower
Montane Forest. Although it is possible that this tree blooms throughout much
or all of the year, when I visited the type locality of A. werckleae in April,
1966, in an unsuccessful attempt to obtain more specimens of this new bat, the
Wercklea flowers were much less abundant than when I had seen them in July
and August of previous years and were generally restricted to higher branches
in the trees. At that time (April), I observed insects and at least two unidenti-
fied species of hummingbirds visiting the flowers during the day. It seems quite
unlikely that Anoura werckleae is solely dependent on Wercklea lutea as a
source of food, but a feeding relationship has been demonstrated and the
ranges of the two organisms may coincide. The genus Wercklea includes one
other species, W. insignis Pittier and Standley, which occurs at somewhat
lower elevations in Costa Rica (and possibly western Panama) and is found
in the same type of ecological situation as is W. lutea. It is interesting to
speculate that a relationship may also exist between Anoura cultrata and this
purplish-flowered Wercklea, although no evidence of such an association is
available in the field data which I have seen. The Costa Rican specimen of
A. cultrata comes from an area in which Wercklea insignis is locally abundant,
but more definitive information is wanting. The two species of Wercklea were
considered by Standley (1937-1938) to be Costa Rican endemics, but he also
recognized the continuity of floras through the Cordillera de Talamanca to the
Chiriqui massif. It is possible that W. insignis is not found east of the isthmian
lowlands of Panama, in which case Anoura cultrata would have a wider distri-
bution than the tree, since this bat does occur in the mountains of eastern
Darien (type locality) .

The discovery of Anoura werckleae is of particular interest because this
bat is one of the few to be found at such a high elevation in tropical Middle
America. Data for Costa Rica and Panama (LACM records and Handley,
1966) show that, as would be expected, the vast majority of species of bats
occur in the Tropical and Subtropical Zones. Between 1500 m (roughly the

1969

New ANOURA From Costa Rica

7

upper limit of the Subtropical Zone) and 1800 m the number of species drops
off rapidly. Handley (1966) gives no specific elevation above 1768 m for any
bat in Panama; Davis (1966) gives a Panamanian record for Eptesicus andinus
from 1920 m. Sturnira ludovici, the only other species of bat taken in the nets
with Anoura werckleae, has been recorded in Costa Rica from 2820 m on
Volcan Turrialba, in the Cordillera Central (Starrett and De la Torre, 1964),
and has been netted by LACM-USACR field parties at elevations up to 3050
m, on the Cerro de la Muerte massif. The only other bats to be taken above
1800 m in Costa Rica are Eptesicus andinus , up to 2743 m, and Myotis
chiloensis, 2743 m (inadvertently shown as 2590 m in Starrett and Casebeer,
1968, in which both species were reported). Elevation records for some of
these specimens were mentioned previously, in Tamsitt et al. (1964), although
they were all listed as having been taken on Volcan Turrialba.

Acknowledgments

Persons involved in the field work which provided the LACM specimens
and data referred to in this paper included, in addition to the author: Richard
S. Casebeer and Ronald B. Linsky (1961); Richard S. Casebeer, Harry N.
Coulombe, Charles L. Hogue, Anthony J. Hollister (1963); George F. Fisler
(1966). Funds supporting these field parties came from the following grants
(field years and principal investigators) : National Academy of Sciences,
Bache Fund, No. 465, and National Science Foundation, No. G- 18002 (both
1 961 -Starrett ) ; U.S. Army Medical Research and Development Command,
No.s DA-MD-49-1 83-62-G54 and DA-MD-49-1 93-63-G94 (1963, 1966-F. S.
Truxal and C. A. McLaughlin, Jr.).

I also express my appreciation to William B. Davis and Dilford C. Carter,
Texas A. and M. University (TCWC), and Charles O. Handley, Jr., U.S. Na-
tional Museum (USNM), for loans of specimens and for patience in the face
of any inconveniences caused by the length of time it has taken for this descrip-
tion to reach publication. George F. Fisler and Donald R. Patten kindly read
the manuscript and contributed useful suggestions. Figure 1 was drawn by
Mary Butler, LACM illustrator; the photograph for Figure 2 is the work of
Armando Solis, LACM photographer.

Resumen

Se describe una nueva especie de murcielago glosofaginido, con base en
dos especimenes (hembra y macho adultos) de Costa Rica, colectados en el
Bosque Pluvial Montano Bajo, a 2500 m. en el macizo del Cerro de la Muerte,
Provincia de San Jose. A esta especie, la cual esta relacionada de cerca con
Anoura cultrata Handley, se le da el nombre de Anoura werckleae en referen-
da a la asociacion que existe entre el murcielago y la planta malvacea,
Wercklea lutea Rolfe. Los dos ejemplares de esta nueva especie fueron colecta-
dos en julio (1961 y 1963) cuando las grandes flores amarillas de Wercklea
eran abundantes en los arboles que crecen a lo largo de varios arroyos presentes

8

Contributions in Science

No. 157

en la localidad tipo, 6.8 mi. al Sur del restaurante ‘La Georgina,’ en la Carretera
Interamericana. El segundo especimen fue colectado con una red de seda
colocada bajo las ramas de un arbol de Wercklea con el proposito especifico de
obtener algun otro ejemplar de este nuevo murcielago. En la piel de ambos
especimenes se encontro polen amarillo proveniente de las flores de este arbol.

Anoura werckleae se asemeja de cerca a A. cultrata, especie de la cual
se distingue por su piel pardo-anaranjada, el metacarpo III mas largo en
comparacion con el metacarpo y la primera falange del cuarto dedo, varias
caracteristicas de los dientes, del palatino posterior y del area del basiesfenoides
(Fig. 1), asimismo como en la forma del craneo (Fig. 2). A. werckleae se
distingue de las otras especies en el genero Anoura en la misma forma que
A. cultrata (vease Handley, 1960).

Anoura werckleae puede ser considerada un substituto altitudinal de
A. cultrata, la cual ha sido encontrada unicamente en Panama y Costa Rica,
entre 600 y 1600 m. de elevacion.

Es probable que A. werckleae sera encontrada a mayores elevaciones en
las masas montanosas de la Cordillera Central de Costa Rica y de la Cordillera
de Talamanca en Costa Rica y Panama, posiblemente coincidiendo con la
distribution de Wercklea lutea, aunque los datos disponibles no dejan ver
ninguna relacion de dependencia entre los dos organismos. Parece interesante
ademas, especular sobre la posibilidad de que alguna relacion similar pueda
existir entre Anoura cultrata y la otra especie del genero Wercklea endemico
de Costa Rica, W. insignis Pittier y Standley, los cuales ocurren en la misma
area general en Costa Rica.

Finalmente, se hace notar que A. werckleae es una de las pocas especies
de murcielagos encontrados en altitudes que sobrepasan los 1800 m. en Centro
America. En Costa Rica, Sturnira ludovici ha sido encontrado en elevaciones
de hasta 3050 m., Eptesicus andinus y Myotis chiloensis en altitudes de hasta
2743 m.

Literature Cited

Carter, Dilford C. 1968. A new species of Anoura (Mammalia: Chiroptera: Phyl-
lostomidae) from South America. Biol. Soc. Wash., Proc., In press.

, Ronald H. Pine, and William B. Davis. 1966. Notes on Middle American

Bats. Southw. Nat., 1 1 : 488-499.

Davis, William B. 1966. Review of South American Bats of the Genus Eptesicus.
Southw. Nat., 11: 245-274.

Fairchild, Graham B., and Charles O. Handley, Jr. 1966. Gazetteer of Collecting
Localities in Panama. Pp. 9-20, in Wenzel, R. L., and V. J. Tipton, Ectoparasites
of Panama. Field Mus. Nat. Hist., Chicago, 861 p.

Handley, Charles O., Jr. 1960. Descriptions of New Bats from Panama. U. S. Natl.
Mus.. Proc., 112: 459-479.

1969

New ANOURA From Costa Rica

9

. 1966. Checklist of the Mammals of Panama. Pp. 753-795, in Wenzel, R. L.,

and V. J. Tipton, Ectoparasites of Panama. Field Mus. Nat. Hist., Chicago,

861 p.

Holdridge, L. R., and Gerardo Budowski. 1959. Mapa ecologico de Panama. Inst.
Interam. Cien. Agricolas, San Jose, Costa Rica.

Standley, Paul C. 1937-1938. Flora of Costa Rica. Field Mus. Nat. Hist., Bot. Ser.,
18: 1-1616.

Starrett, Andrew, and Richard S. Casebeer. 1968. Records of bats from Costa
Rica. Los Angeles Co. Mus., Contr. Sci., 148:1-21.

, and Luis de la Torre. 1964. Notes on a Collection of Bats from Central

America, with the Third Record for Cyttarops alecto Thomas. Zoologica, 49:
53-63.

Tamsitt, J. R., Dario Valdivieso, and Jorge Hernandez-Camacho. 1964. Bats of
the Bogota Savanna, Colombia, with Notes on Altitudinal Distribution of Neo-
tropical Bats. Rev. Biol. Trop. 12: 107-115.

Tosi, Joseph A., Jr. 1965. Ecological map of Costa Rica, Provisional Edition. Trop.
Sci. Center, San Jose, Costa Rica.

Accepted for publication December 24, 1968.

LOS

ANGELES

COUNTY

MUSEUM

CONTRIBUTIONS
IN SCIENCE

Dumber 158

March 27, 1969

THE STATUS AND DISTRIBUTION IN UGANDA
OF THE

WHITE-BELLIED KINGFISHER,

ALCEDO LEUCOGASTER LEOPOLDI

By Herbert Friedmann

Los Angeles County Museum of Natural History • Exposition Park
Los Angeles, California 90007

CONTRIBUTIONS IN SCIENCE is a series of miscellaneous technical papers
in the fields of Biology, Geology and Anthropology, published at irregular intervals
by the Los Angeles County Museum of Natural History. 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 Robert J. Lavenberg, Los Angeles County Museum of Natural
History, 900 Exposition Boulevard, Los Angeles, California 90007.

Robert J. Lavenberg
Managing Editor

INSTRUCTIONS LOR AUTHORS

Manuscripts for the LOS ANGELES COUNTY MUSEUM, CONTRIBU-
TIONS 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. Al-
though priority will be given to manuscripts by staff members, or to papers dealing
largely with specimens in the Museum’s collections, 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 Edi-
torial Committee. They may also be subject to review by competent specialists out-
side 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.

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. Alterations or changes in the manuscript after galley proof will
be billed to the author. Unless specifically requested, page proof will not 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.

Dorothy M. Halmos
Editor

THE STATUS AND DISTRIBUTION IN UGANDA
OF THE

WHITE-BELLIED KINGFISHER,

ALCEDO LEUCOGASTER LEOPOLDI

By Herbert Friedmann1

Abstract: Current surveys of the birds of the isolated for-
ests of western Uganda have extended the known range of the
white-bellied kingfisher, Alcedo leucogaster leopoldi, 170 miles
to the east (Malabigambo Forest). The use of mist nets has re-
vealed the need for a new appraisal of the numerical status of the
bird, which is now shown to be surprisingly common, where pre-
vious collectors, without the use of such devices, had difficulty in
obtaining single examples. The present series, the first sizeable
one from any one locality, also suggests the existence of two sym-
patric color phases in this species.

The little white-bellied kingfisher, Alcedo leucogaster leopoldi, is a poorly
known and seldom collected bird of the central African forests. In the Congo,
over many years and by many collectors, it has been recorded from a good
number of localities, summarized by Schouteden (1951, 1953, 1968), but
these entail merely single or, at most, few specimens each, and with little or no
observations. East of the Congo, it has been known only from extreme western
Uganda, where it has been reported from Bwamba by Williams (1957, pp. 1 59-
160) and by Friedmann (1966, pp. 22-23), each on the basis of a single speci-
men. South to Mwinilunga, in northwestern Zambia, Benson and Irwin ( 1965,
p. 2) reported it along water courses in the Marquesia thickets. In the forests
where it lives, it appears to be closely restricted to the vicinity of small well-
shaded streams and of small wooded swampy glades, apparently seldom, if
ever, getting into areas of bright sunlight. Chapin (1932, pp. 231-232) con-
sidered this specialized microhabitat a fairly recognizable ecological entity,
and he concluded that this kingfisher was found only in such places, along with
one other related species, Alcedo quadribrachys guentheri, and a few other
kinds of birds. The other ecologically sympatric species were a heron, Tigri-
soma leucolopha\ two ibises, Lampribis rara and L. olivacea cupreipennis\ one
duck, Pteronetta hartlaubi\ three rails, Himantornis haematopus, Canirallus
oculeus, and Sarothrura pulchra centralis ; and a finfoot, Podica senegalensis.

In this microhabitat Alcedo leucogaster leopoldi appears to avoid compe-
tition with other similar predators on small fishes in shallow streams away from
the forest. In this respect it resembles the case of the tiger heron, listed above,
of which Chapin (1932, p. 423) wrote that it, . . is truly a bird of the forest,
seldom to be seen along river banks, even where heavily wooded, but keeping

1Director, Los Angeles County Museum of Natural History.

2

Contributions in Science

No. 158

Figure 1. Distribution of the white-bellied kingfisher, Alcedo leucogaster leopoldi, in
Uganda.

to the smaller streams entirely hidden within the depths of the forest, where
the river herons evidently dare not venture. Thus, it needs never to compete
with the purple heron or any of its better known relatives . . .” The other little
kingfisher found with it along these waterways, Alcedo quadribrachys
guentheri, is, however, not restricted to them, but occasionally ranges far out

1969

Uganda White-Bellied Kingfisher

3

into extensive papyrus swamps, as at Kiyuyu, on the Lualaba River, where
Chapin (1939, p. 292) was surprised to find it. The subspecies thomensis,
inhabiting the island of Sao Tome, has also been known to occur away from
wooded streams. Snow (1950, p. 585) found it on the seashore, perching on
pebbles, although he reported it in more usual locations as well.

In keeping with its adherence to the vicinity of water, this little kingfisher
feeds extensively on small minnows. The stomach contents of four of our
specimens held small fish bones and scales. It is also partly insectivorous in its
diet, as insect fragments were present in the stomachs of two examples. In
Cameroon, Serle (1950, p. 358) considered it primarily an insectivorous king-
fisher. This is a matter of some biological interest as the kingfishers as a family
are known to be both piscivorous and insectivorous, but most of their species
are usually one or the other.

Chapin (1939, p. 288) concluded that in the Congo leopoldi must be a
very rare bird everywhere, as in over five years in that country he had met with
it but a single time and had been able to obtain only one specimen (at
Avakubi) . He felt that only its general rarity could account for the fact that no
one had found it along the heavily forested banks of the Ituri or the Aruwimi
Rivers. In fact, as recently as 1922, when he first described the female plumage
from his Avakubi specimen, Chapin (1922, p. 443) referred to it as the second
known example of leopoldi ; all the other Congo records implied in the opening
sentences of the present paper being later acquisitions.

In the past two years, under a National Science Foundation grant
(GB-5107) to the Los Angeles County Museum of Natural History (LACM),
extensive collecting in the forests of western Uganda, enhanced particularly
by the use of mist nets, have caused us to change our evaluation of the numeri-
cal status of this bird, and to extend far to the east the limits of its range in
Uganda. In the Bwamba forest, in a period of less than three weeks, from
June 15 to July 3, 1967, no fewer than 11 specimens were obtained, clearly
showing that it is a fairly common bird there. The species was not obtained in
the Budongo, Bugoma, Kibale, or Impenetrable Forests, presumably because
of the higher altitudes of those areas. However, surprisingly enough, it was
met with again, and collected, in the Malabigambo Forest in the Sango Bay
area, about 40 miles south of Masaka, on the northwest coast of Lake Victoria,
some 170 miles to the east of its previously known range. Here an adult male
in nonbreeding condition was collected on January 26, 1968, and another with
slightly enlarged gonads was taken on February 4, 1968.

The two Sango Bay examples have relatively short bills, measuring 23.9
and 24 mm from the anterior end of the nostril to the tip of the maxilla, and
are matched in this respect by one of the Bwamba males. All the other Bwamba
specimens have longer bills, ranging from 25 to 27.5 mm in the males, and
from 25.1 to 25.6 mm in the females. The difference, while quite small, is
significant even though not wholly constant, with an average bill length of 26.1
in the Bwamba males as compared with an average of 23.95 mm in the two

4

Contributions in Science

No. 158

Sango Bay examples. To the eye the difference looks larger than actual mea-
surements show. Because the eastern Sango Bay population is known from two
specimens only and one of the Bwamba series has a slightly shorter bill (23.7),
it seems advisable to wait for further material before considering the advis-
ability of separating the Sango Bay birds subspecifically, although their bill
character is of interest as showing a microevolutionary trend. Small differences
in bill dimensions, both in length and in height, and in wing length, are variable
characters, which, in other segments of this species, have been found to have
geographic correlation. Thus, the nominate race, restricted to the island of
Fernando Poo, differs from batesi of the Cameroon, Gabon, and northern
Angola by only a few millimeters. According to Amadon’s findings (1953,
pp. 420-421 ), leucogaster wing lengths average from 3 to 5 mm longer than in
batesi, boxvdleri, and leopoldi. Even in this slight character there is some
altitudinal size variation in batesi, as Serle (1950, p. 358) reported that an
example of that race taken at Bamenda, Cameroon, at 6000 feet, was as large
as the Fernando Poo birds. The other subspecies of the Gulf of Guinea islands,
nais of Principe and thomensis of Sao Tome, are more distinctly different,
having much more rufescent color on the underparts of the body.

Two Malabigambo specimens (LACM) show another difference in their
bills, which are brighter, more orange-red, with less dusky brownish on the
culminal ridge of the maxilla than are the bills of the Bwamba series. In this
connection it may be recalled that Chapin (1939, p. 288) noted that Dubois’s
type of leopoldi (from Lake Leopold II) was plainly immature, with a dark
colored bill, but added that “. . . even adults of leopoldi, such as ours from
Avakubi, seem to have the bills largely dark brown, not wholly red as in batesi
and the other western races.” This much more basally restricted duskiness of
the culmen in the two Sango Bay specimens, together with the relative short-
ness of their bills, does suggest incipient divergence from typical leopoldi in
this remote and isolated population, a type of divergence seemingly present in
the subspecies batesi, according to Chapin. Whether batesi averages less dusky
culminal coloration is, however, open to question, as Rand, Friedmann, and
Traylor (1959, pp. 282-283) found that in two males and one female batesi
from Gabon the maxilla was blackish and the mandible red.

The present material is the first sizeable series from one locality of the race
leopoldi available for study, and from an examination of it, it appears that the
birds fall into what may almost be considered two color phases, as far as the
terminal bars on the feathers of the forehead, crown, and occiput are con-
cerned. About half of the Bwamba series and one of the Sango Bay birds have
these narrow bars dark, varying from Rood’s Blue to Helvetia Blue on the
forehead and fore crown, and Salvia Blue on the feathers of the hind crown
and occiput; the other half of both series have these bars much paler and
slightly more greenish, very close to Squill Blue (colors ex Ridgway’s nomen-
clature of colors) . The birds are one or the other; there are no intermediates. It
is of interest to note that the paler, more greenish color of these bars was

1969

Uganda White-Bellied Kingfisher

5

considered by Chapin ( 1939, p. 288) to be a racial character of the subspecies
leopoldi, but this conclusion was based on a very small and inadequate series
of only one example of each sex, only one of which was adult. This presumed
character was repeated as diagnostic in White’s (1965, p. 225) list, probably
on the basis of Chapin’s earlier statement, but it remains to be demonstrated
that it is a valid racial character. Our Bwamba series negates that possibility.

Of the 11 examples collected in the Bwamba Forest, only one had en-
larged gonads, a male taken on June 13; two other males, taken on June 28
and July 3, showed slight testicular swelling; all the others had small “resting”
gonads.

Because of the excessive rarity of this bird in museum collections, it
seems worthwhile to put on record the dimensions of all 14 of our specimens.
Nine males from Bwamba have wing lengths of 52,2, 54, 54.3, 54.9, 55, 55,
55.1, 55.2, and 56.1 (average 54.6); tail 22.2, 22.3, 22.7, 22.8, 23.3, 23.3,
23.7, 24.2, 24.3 (average 23.2); bill from anterior end of nostril to tip of
maxilla 23.7, 25.0, 25.3, 25.8, 26.7, 26.9, 27.2, 27.5 (average 26.1). Two
males from Malabigambo Forest, Sango Bay, have wings 55.1, 56.4 (average
55.7); tail 22.3, 24.6 (average 23.4); bill, as above, 23.9, 24.0 (average
23.95). Three females from Bwamba have wings 53.5, 54.3, 57.6 (average
55.1); tail 21.6, 22.8, 23.7 (average 22.7); bill, as above, 25.1, 25.1, 25.6
(average 25.2 mm). Males weighed from 12.5 to 15 grams (average 13.8
grams) ; females weighed from 14 to 16 (average 14.9 grams) . As may be seen
from these figures, the females tend to have slightly longer wings, and to
weigh more than the males.

Literature Cited

Amadon, Dean. 1953. Avian systematics and evolution in the Gulf of Guinea. Amer.
Mus. Nat. Hist., Bull. 100: 399-451.

Benson, C. W., and M. P. Stuart Irwin. 1965. The birds of Marquesia thickets in
northern Mwinilunga District, Zambia. Arnoldia, 1 (30) : 1-4.

Chapin, J. P. 1922. On the representatives of Corythornis leucogaster (Fraser) in the
Cameroon and the Congo. Ibis, (11) 4: 440-445.

. 1932. The birds of the Belgian Congo. Pt. I. Amer. Mus. Nat. Hist., Bull. 65 :

1-756.

. 1939. The birds of the Belgian Congo. Pt. II. Amer. Mus. Nat. Hist., Bull.

75: 1-632.

Friedmann, Herbert. 1966. A contribution to the ornithology of Uganda. Los An-
geles Co. Mus. Nat. Hist., Bull., Science, 3: 1-55.

Rand, A. L., H. Friedmann, and M. A. Traylor, Jr. 1959. Birds from Gabon and
Moyen Congo. Fieldiana, Zoo!., 41: 221-411.

Schouteden, H. 1951. Die Vogels van Belgisch Congo en van Ruanda-Urundi. Ter-
vuren, Mus. van Belgisch Congo, Ann., Zoo!.. (4) 3 (1): 1-176 (pp. 29-30).

. 1953. Nouvelles notes sur Corythornis leucogaster Leopoldi Dub. (Aves).

Rev. Zool. Bot. Africaines, 48: 297-299.

. 1968. La Faune Ornithologique du Kivu. I. Non Passereaux. Tervuren, Mus.

voor Midden-Afrika, Zool. Doc. 12: 1-168 (pp. 121-122).

Serle, William. 1950. A contribution to the ornithology of the British Cameroons.
Ibis, 92: 343-376, 602-638.

6

Contributions in Science

No. 158

Snow, D. W. 1950. The birds of Sao Tome and Principe in the Gulf of Guinea. Ibis,
92: 579-595.

White, C. M. N. 1965. A revised checklist of African non-passerine birds. Lusaka,
Government Printer, 299 pp.

Williams, J. G. 1957. Four new bird records from eastern Africa. Brit. Ornith. Club,
Bull., 77: 159-160.

Accepted for publication January 17, 1969.

LOS

ANGELES

COUNTY

MUSEUM

CONTRIBUTIONS
IN SCIENCE

UMBER 159 March 31, 1969

1

on/] 3

THORACIC CIRRIPEDIA OF THE SAN DIEGO FORMATION,

SAN DIEGO COUNTY, CALIFORNIA

'

By Victor A. Zullo

Los Angeles County Museum of Natural History • Exposition Park
Los Angeles, California 90007

CONTRIBUTIONS IN SCIENCE is a series of miscellaneous technical papers
in the fields of Biology, Geology and Anthropology, published at irregular intervals
by the Los Angeles County Museum of Natural History. 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 Robert J. Lavenberg, Los Angeles County Museum of Natural
History, 900 Exposition Boulevard, Los Angeles, California 90007.

Robert J. Lavenberg
Managing Editor

INSTRUCTIONS FOR AUTHORS

Manuscripts for the LOS ANGELES COUNTY MUSEUM, CONTRIBU-
TIONS 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. Al-
though priority will be given to manuscripts by staff members, or to papers dealing
largely with specimens in the Museum’s collections, 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 Edi-
torial Committee. They may also be subject to review by competent specialists out-
side 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.

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. Alterations or changes in the manuscript after galley proof will
be billed to the author. Unless specifically requested, page proof will not 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.

Dorothy M. Halmos
Editor

THORACIC CIRRIPEDIA OF THE SAN DIEGO FORMATION,
SAN DIEGO COUNTY, CALIFORNIA

By Victor A. Zullo1

Abstract: The cirriped fauna of the Pliocene San Diego
Formation, San Diego County, California, includes six species of
Balanus Da Costa and one each of Cetolepas gen. nov., Coronula
Lamarck, and Lepas Linnaeus. Three species, Balanus (Balanus)
kanakoffi, B. ( Megabalanus ) wilsoni, and Cetolepas hertleini are
new and known only from the San Diego Formation. The single
lepadid scutum represents the first fossil record of Lepadomorpha
from the margins of the eastern Pacific Basin. The Coronula ap-
pears to be C. barbara Darwin, known previously from the Plio-
cene and Early Pleistocene of Europe. Cetolepas hertleini af-
fords a clue to the derivation of Tubicinella Lamarck from
Coronula.

This warm temperate to subtropical fauna is composed of a
mixture of extinct and living species and is transitional between
California Miocene faunas characterized by extinct species and
Pleistocene faunas dominated by living species.

Introduction

This paper is part of a series (Hertlein and Grant, 1944, 1960) dealing
with the geology and paleontology of the San Diego Formation of San Diego
County, California, that is being developed under the aegis of Dr. Leo G.
Hertlein, California Academy of Sciences. The bulk of the material used in
this study was collected by Mr. George P. Kanakoff, former Curator of
Invertebrate Paleontology, Los Angeles County Museum of Natural History,
and the present Curator, Dr. Edward C. Wilson, and is supplemented by
specimens from the collection of the California Academy of Sciences.

Altogether, nine species, of which three are new, representing four genera,
of which one is new, are recognized. However, this account does not include
all of the species that were evidently present during the deposition of the sedi-
ments of the San Diego Formation, for a few isolated opercular valves and
disarticulated compartmental plates could not be identified with recognized
taxa.

It is difficult to characterize the cirriped fauna of the San Diego Forma-
tion, either geographically or historically, primarily because of the present lack
of data on fossil barnacle faunas of the Pacific Coast. Of the nine species
indicated by the collection, three, including one of a new genus, are new and
not known from other deposits. Two, Balanus gregarius (Conrad) and B.

Research Associate in Invertebrate Paleontology, Los Angeles County Museum of
Natural History, and Associate Curator, Department of Geology, California Acad-
emy of Sciences, San Francisco, California 94118.

1

2

Contributions in Science

No. 159

proinus Woodring, are extinct species that were apparently widespread in
shallow Pliocene seas of southern California and endemic to the Pacific Coast
of North America. A third extinct species, tentatively identified with the whale
barnacle Coronula barbara Darwin, is otherwise known only from the Pliocene
and Early Pleistocene of Europe. Two of the remaining species appear to
represent the extant Pacific Coast barnacles Balanus nubilus Darwin and
B. pacifcus Pilsbry. The Lepas sp. is similar to an extant cosmopolitan species
usually found attached to floating seaweed and debris.

Comparison of this fauna with one of similar size and age from the San
Joaquin Basin to the north (unpublished data based on UCMP collections)
indicates little similarity between them. Two species, Balanus gregarius and
B. proinus, are common to both faunas, but the remaining San Joaquin species
appear to be endemic to the basin. The Pliocene barnacles of the geographically
closer Santa Maria District are poorly known, and the two species reported
(the B. aquila Pilsby and B. hesperius proinus of Woodring, in Woodring and
Bramlette, 1950) are the same species that tenuously link the San Diego fauna
to that of the San Joaquin Basin.

In a broad sense, the barnacle fauna of the San Diego Formation suggests
warm temperate to tropical conditions, and is compositionally transitional be-
tween that of the Pacific Coast Miocene, whose species are largely extinct, and
that of the Pleistocene, which is composed primarily of extant species. On the
other hand, surprisingly little information is afforded by the fauna, with the
possible exception of the new whale barnacle, regarding the evolution of
extant Pacific Coast species.

In order to facilitate identification of the species reported herein, keys to
scuta, terga, and shells have been prepared, and are to be found immediately
following the section on systematic descriptions. Caution must be exercised in
the use of these keys, for they are applicable only to the barnacles of the San
Diego Formation, and although the common barnacles of the formation can be
recognized from the keys, note must be made of the fact that other species are
evidently present and may be better represented in future collections.

The following abbreviations are employed herein to designate those
institutions whose collections were examined: CAS, California Academy of
Sciences, San Francisco; LACM, Los Angeles County Museum of Natural
History, Los Angeles, Section of Invertebrate Paleontology; SDSNH, Natural
History Museum, San Diego; UCMP, University of California Museum of
Paleontology, Berkeley.

Age of The San Diego Formation

The San Diego Formation has been considered of Middle Pliocene
(“Etchegoin”) age in terms of Pacific Coast marine chronology. However, in
recent years the temporal extent of the Pliocene on the Pacific Coast has been
significantly reduced with the establishment of the Pliocene-Pleistocene

1969

Thoracic Cirripedia of the San Diego Formation

3

boundary at the base of the Calabrian in Italy (circa 3.5 million years Before
Present) and the recognition of the Hemphillian-Clarendonian Boundary as
the boundary between Pliocene and Miocene (circa 10 m.y. B.P.). This time
span reduction coupled with paleontological evidence presently being obtained
from a review of the Late Cenozoic of the Pacific Coast suggests that a two-
rather than threefold division of the Pliocene is more realistic. On this basis,
the San Diego Formation would be regarded as Late Pliocene in age and is
treated accordingly herein.

Description of Barnacle Localities
in the San Diego Formation

Los Angeles County Museum of Natural History

107. Clay and gravel quarry at end of Arroyo Drive, City of San Diego,
California. The 40 to 50-foot high cut contains oyster and pecten
beds about 20 to 30 feet above road level and scattered large concre-
tions throughout. G. P. Kanakoff, coll., May 12, 1947.

122. Shore bluff at the end of Loring Street, Pacific Beach, California. San
Diego Formation exposed from 0 to 20 feet above sea level is over-
lain by fossiliferous, coarse grained, red sand of Late Pleistocene age.
G. P. Kanakoff, coll.

180. On east side of 2200 block of La Jolla Boulevard at intersection with
Tiras Street, ?Pacific Beach, California. G. P. Kanakoff, coll.,
January 28, 1950.

305. 2400 feet east and 1350 feet south of the northwest corner of Section

8, T. 19 S, R. 2 W, San Bernardino Base and Meridian, San Ysidro
quadrangle, 1943 ed. (=CAS Loc. 34814). G. P. Kanakoff, coll.

305A. West side of next gulley east of LACM Loc. 305 at the same eleva-
tion. Fossils in float slump and consolidated boulders, silt, and sand-
stone, and silt in place (=CAS Loc. 36555). G. P. Kanakoff and
W. K. Emerson, coll., December, 1957.

305C. Exposure at base of hill 100 feet west and 440 feet south of the north-
east corner of Section 8, T. 19 S, R. 2 W, San Bernardino Base and
Meridian, San Ysidro quadrangle, 1953 ed., G. P. Kanakoff, coll.

319. Exactly between U.S. -Mexico boundary fence and Mr. Ericson’s (the
manager) house; 27 feet above road level on shoulder of second hill.
G. P. Kanakoff, coll.

323. Under bridge between Fifth Street and the Radio Station, about 160
feet from fence and about 350 feet from Radio Station. J. Arndt, coll.,
February, 1961.

4

Contributions in Science

No. 159

485. Marine invertebrates and shark teeth collected from an estimated 20-
foot thickness of mostly unconsolidated, yellow, medium to coarse
grained sand on 30° bulldozed slope north of Market Street and east
of Euclid Avenue, City of San Diego. Intersection of projection of
arrows on edges of northwest quarter of map marked “2.6 mi. to U.S.
101” and “0.6 mi. to U.S. 80” on 7.5' National California quad-
rangle, 1953 ed. E. C. Wilson, coll., December 10 and 24, 1967.

492. Southwest corner of intersection of Home and Fairmont Avenues,
City of San Diego. J. Arndt, coll., June, 1957.

493. East side of Wabash Canyon, below Juniper Street, City of San Diego.
C. Anderson and E. P. Chace, coll., 1961.

California Academy of Sciences

1405. Street cut 0.2 miles southwest of Alamo Drive and Center Street, City
of San Diego.

33218. Near intersection of Maple and Haller Streets, City of San Diego.
L. L. Mills, coll., January, 1951.

San Diego Museum of Natural History

L-2451. Same locality as LACM 485, E. C. Wilson, coll., July 31, 1967.

SYSTEMATIC DESCRIPTIONS

Suborder LEPADOMORPHA Pilsbry, 1916
Family LEPADIDAE Darwin, 1851
Genus Lepas Linnaeus, 1758

Lepas sp.

Figs. 1-2

Occurrence : LACM Loc. 305 A.

Range : Late Pliocene, San Diego Formation.

Remarks : The single scutum available represents the only fossil lepado-
morph to be recorded from the margins of the eastern Pacific Basin. As pre-
viously noted by Withers (1953, p. 354), the “ Lepas injudicata ” of Pilsbry
(1919, p. 188, pi. 67, fig. 5) described from the Miocene of the Panama Canal
Zone is in fact the broken half of a pelecypod shell, probably of the venerid
Chione ( Lirophora ) mactropsis (Conrad).

Extant species of Lepas are difficult enough to identify from well pre-
served material, and at present are best differentiated by the number of fila-
mentary appendages present at the base of the first cirrus. The fossil scutum is
quite distinctive, but in itself does not provide sufficient criteria for specific

1969

Thoracic Cirripedia of the San Diego Formation

5

Figures 1-10. 1-2, opercular valves of Lepas, sp., (1-2), interior and exterior of
scutum, Hypotype LACM 1196, LACM Locality 305A, height of valve 4 mm.
3-7, opercular valves of Balanus gregarius (Conrad), LACM Locality 485. (3) in-
terior of scutum, Hypotype LACM 1197, height 22 mm; (4) exterior of scutum,
Hypotype LACM 1198, height 22 mm; (5) interior of tergum showing purple color
patch on carinal side, Hypotype LACM 1199, height 18 mm; (6) exterior of tergum,
Hypotype LACM 1200, height 21 mm; (7) interior of tergum, Hypotype LACM
1201, height 23.5 mm. 8-10, opercular valves of Balanus nubilus Darwin, LACM
Locality 305. (8-9) interior and exterior of scutum, Hypotype LACM 1203, height
27 mm; (10) exterior of tergum, Hypotype LACM 1204, height 14.5 mm.

6

Contributions in Science

No. 159

identification. Among extant species that possess strong radial sculpture, it
approaches the scutum of L. pectinata Spengler, especially in the narrowness
of the margin on the occludent side of the ridge extending from umbo to apex.

Suborder BALANOMORPH A Pilsbry, 1916
Family BALANIDAE Leach, 1817
Genus Balanus Da Costa, 1778
Subgenus Balanus
Balanus gregarius (Conrad, 1856)

Figs. 3-7, 45

Tamiosoma gregaria Conrad, 1856, p. 315; 1857a, p. 72, pi. 4, fig. 18; Gabb,
1869, p. 61, pi. 18, figs. 22a-d; Dali, 1902, p. 5.

Balanus estrellanus Conrad, 1857b, p. 195, pi. 8, fig. 1; 1877, p. 156.

Radiolites gregaria Conrad, 1864, p. 214.

Balanus H. estrellanus Conrad, 1876, p. 273.

Balanus gregarius (Conrad). Pilsbry, 1916, p. 126, pi. 28, figs. 1-3, pi. 29;
Zullo, 1964, p. 360; Durham and Addicott, 1965, p. 14, pi. 1, figs. 2, 3, 6, 8
(not pi. 2, figs. 4, 7).

Balanus ( Tamiosoma ) cf. B. (T.) gregarius (Conrad). Woodring, in Wood-
ring, Stewart, and Richards, 1940, p. 96, pi. 36, figs. 2-5, 8, 9.

Balanus concavus concavus Bronn. Ross, 1962, p. 14, figs. 6, 7.

Occurrence : LACM Locs. 107, 305, 305C, 319, 323, 485, 492, 493;
CAS Locs. 1405, 33218; SDSNH Loc. L-2451.

Range : Early Miocene through Late Pliocene, central and southern Cali-
fornia; Pliocene, Baja California.

Remarks : Balanus gregarius is the most widespread and abundant barna-
cle in collections from the San Diego Formation. Opercular valves of this
species, and especially scuta, are common at LACM Locs. 305 and 485, to-
gether with shells and isolated compartmental plates. Several complete speci-
ments with opercular valves in life position were obtained from LACM Loc.
485. Scuta were also found at LACM Locs. 319 and 323 and CAS Loc. 1405.
Large shells with the distinctive vesiculose basis of B. gregarius were collected
at LACM Locs. 107, 492, and 493, and at CAS Loc. 33218.

Balanus gregarius, in the broad sense, is a common fossil encountered in
Miocene and Pliocene deposits of the San Francisco Bay Area, Salinas Valley,
and San Joaquin Valley in California, and in Pliocene rocks at Rosario in
Baja California. As indicated by the preceding synonymy, there has been
considerable confusion regarding its identification and affinities. Conrad

1969

Thoracic Cirripedia of the San Diego Formation

7

(1856, 1857a, 1864), who originally described B. gregarius from its distinctive
vesiculose basis, considered it to be a rudistid pelecypod. Later, Conrad (1876)
and Dali (1902) recognized its relationship to the genus Balanus, but did not
have the opercular valves available to them. The valves were first described by
Woodring, in Woodring et al. (1940), but because the delicate beaks of the
terga were missing in the specimens examined, Woodring (op. cit.) was led
to conclude that B. gregarius was closely related to B. concavus Bronn. The
marked similarity of the scuta of B. gregarius to those of various subspecies of
B. concavus was also responsible for Ross’ (1962) record of B. concavus
concavus on the basis of a scutum from the Pliocene of Rosario, Baja Cali-
fornia. The reassignment of this scutum to B. gregarius is supported by the
occurrence in the same deposit (UCMP Loc. 4300) of undoubted B. gregarius
whose scuta are identical in form to that figured by Ross (op. cit.).

When well preserved, intact opercular valves of B. gregarius are avail-
able, the marked resemblance of this species to the extant Pacific Coast species
B. aquila Pilsbry becomes immediately apparent. The resemblance is so great
as to suggest that the two species are conspecific (Zullo, 1964), but a more
detailed study of fossil populations, their variation, and their comparison with
living B. aquila is needed before any conclusions can be drawn.

Balanus kanakoffi sp. n.

Figs. 11-22

Diagnosis: Balanus s.s. with conic, strongly ribbed or plicate shell and
small, untoothed, subtriangular orifice; sutures between carinolaterals and
carina linear and often obscured externally in older specimens; scutum with
prominent growth ridges cut by deep longitudinal striae, giving a marked
cancellate appearance to the exterior; beaked tergum with broad, basal ly
truncate spur and open spur furrow.

Description: Shell conic, with small, subtriangular, un toothed orifice;
parietes strongly and regularly ribbed or plicate, often preserving light, ir-
regularly spaced, transverse color bands corresponding to growth increments
against a darker background; parietal tubes large, rectangular, without trans-
verse septa, but solidly filled in upper half; one or two secondary septa occa-
sionally present on outer lamella between major parietal septa; radii narrow,
solid, thick, with slightly oblique summits; sutural edges of radii coarsely
denticulate; alae thin, with horizontal summits and sharp, non -denticulate
sutural edges; sheath one-half height of shell, lower edge dependent; interior
of parietes below sheath with conspicuous, square ribs corresponding to
parietal septa; basis thick with large, radiating, non septate tubes.

Scutum thick, slightly concave externally between apex and base; exterior
ornamented by high, sharp, regularly spaced growth ridges crossed by broad,
deeply incised radial striae, forming a lattice structure with nodes on the

8

Contributions in Science

No. 159

growth ridges and thin vertical pillars in the inter-ridge areas between striae;
tergal border reflexed 90°; adductor ridge absent, or at best defined as the
raised lower lip of the large, deep, oval adductor muscle pit located in the
upper half of the valve; articular ridge confined to upper half of tergal margin,
reflexed over shallow articular furrow; depressor muscle pit lenticular, deep,
extending from basal margin to base of articular ridge.

Tergum thin, narrow, beaked; width of basal margin equal to one-half
height of valve; spur long, broad, situated close to basiscutal angle, and occupy-
ing about three-fifths the width of basal margin and one-fourth the total height
of valve; juncture of spur with basal margin angular on scutal side, broadly
curved on carinal side; articular ridge low, inconspicuous, erect; articular
furrow broad, shallow; depressor muscle crests, numerous, conspicuous.

Occurrence and Type Disposition : LACM Locs. 305 (type locality),
305A, 305C; Holotype LACM 1209; Paratypes LACM 1205-1211; Paratypes
CAS 13162-13165.

Range : Late Pliocene, San Diego Formation.

Remarks : Balanus kanakoffi differs most markedly from other species of
the subgenus Balanus in the linear and often obscure nature of the suture be-
tween the carina and carinolaterals. It has its greatest affinities with the B.
calidus Pilsbry — B. spongicola Brown — B. trigonus Darwin group. Although
sharing certain characteristics with each of the above mentioned species,
B. kanakoffi is readily distinguished from them in total aspect, which combines
a strongly ribbed or plicate shell with a markedly cancellate scutum and a
beaked tergum having a broad, basally truncate spur.

The scutum resembles that of B. spongicola in external sculpture, but
differs in the shorter articular ridge and the larger depressor muscle pit. The
tergum of B. kanakoffi is similar to that of B. trigonus, but can be recognized
by the greater development of the beak and the longer, narrower spur whose
juncture with the basal margin on the carinal side is not angular.

Although the shells are easily separated, the opercular valves of B.
kanakoffi are somewhat similar to those of the extant Gulf of California
species, B. eyerdami Henry. However, the latter differs in having a channel
entering the base of the prominent beak on the inner surface, a more elongate
and basally pointed tergal spur, and a larger, but less defined pit for the de-
pressor muscle on the scutum.

This species is named in honor of Mr. George P. Kanakoff.

Balanus nubilus Darwin, 1854
Figs. 8-10

Balanus nubilus Darwin, 1854a, p. 253, pi. 6, figs. 2a-c; Ross, 1962, p. 24.

Balanus nubilis (sic) Darwin. Pilsbry, 1916, p. 131, pi. 30, figs. 1-4, pi. 31,
figs, 3, 3a, 4, 5.

1969

Thoracic Cirripedia of the San Diego Formation

9

Figures 11-18. B alarms kanakoffi sp. n., LACM Locality 305; (11) interior of scu-
tum, Paratype LACM 1205, height 8.5 mm; (12) shells on cidaroid spine, Paratype
lot LACM 1206, length of group 16 mm; (13) interior of scutum, Paratype CAS
13162, height 6 mm; (14) exterior of scutum, Paratype CAS 13163. height 6 mm;
(15) exterior of scutum, Paratype LACM 1207, height 7 mm; (16-17) exterior and
interior of tergum, Paratype LACM 1208, height 8.5 mm; (18) interior of tergum,
Paratype CAS 13164, height 7 mm.

10

Contributions in Science

No. 159

Occurrence : LACM Loc. 305.

Range: Late Pliocene, California and Baja California; Pleistocene, Ore-
gon to Baja California; Recent, southern Alaska to San Quintin Bay, Baja
California.

Remarks : Fragments of two terga and one scutum are referred to this
species. Balanus nubilus is abundant in the cool water Late Pliocene deposit at
Moonstone Beach, California ( e.g ., UCMP Loc. B-7346; Allison, Durham,
and Zullo, 1961), and is found at various Pleistocene localities along the
Pacific Coast from Cape Blanco, Oregon (e.g., UCMP Loc. B-7372) south to
Baja California (Ross, 1962, p. 27).

Balanus sp., cf. B. pacificus Pilsbry, 1916
Figs. 23-25

Occurrence : LACM Locs. 107, 305.

Range : (?) Late Pliocene, San Diego Formation; Pleistocene, central and
southern California, Baja California; Recent, San Francisco, California, to
northern Peru.

Remarks'. Three shells that may be referrable to this species were obtained
from LACM Loc. 107. In addition, ten scuta from LACM Loc. 305 also appear
to represent B. pacificus. The internal structure of these scuta is like that of
Pleistocene and extant specimens, but the longitudinal striae are not as well
developed. This difference may be the result of pre-depositional wear.

Balanus pacificus is a commonly encountered fossil in Late Pleistocene
deposits at Tomales Bay (UCMP Loc. B 6354), Newport Beach (UCMP
Locs. A-2509, A-3101, A-3102, A-3103), and in southern San Diego County
(UCMP Loc. A-9005), California, and at Santa Ynez (UCMP Loc. A-3582)
and San Quintin (UCMP Locs. A-8677, A-9586, B-3069) Bays in Baja
California. This species has not previously been reported from Pre-Pleistocene
rocks, but a possible ancestor occurs in the Late Miocene Santa Margarita
Formation of the Nipomo quadrangle, California (UCMP Loc. A-1456).

Ross (1964, p. 489) separated this eastern Pacific species from the
Tethyan B. concavus Bronn complex on the basis of the nearly horizontal
rather than oblique summits of the radii and the absence of transverse septa in
the parietal tubes.

Subgenus Megahalanus Hoek, 1913

Balanus wilsoni sp. n.

Figs. 33-46

Diagnosis'. Cylindric megabalanid with large, subtriangular, untoothed
orifice; parietes smooth with narrow, reddish stripes on white background;
radii broad with nearly horizontal summits and small radial pores; radial septa

1969

Thoracic Cirripedia of the San Diego Formation

11

Figures 19-25. 19-22, shells of Balanus kanakoffi sp. n., LACM Locality 305; (19)
dorsal view of shell, Holotype LACM 1209, carinorostral length 12 mm; (20) inter-
ior of shell showing internal ribbing, sutural edges, and tubes in basis, Paratype
LACM 1210, basal width 9 mm; (21) dorsal view of shell, Paratype CAS 13165,
carinorostral length 12 mm; (22) interior of rostrum showing sutural edges of radii
and parietal tubes, Paratype LACM 1211, height 10 mm. 23-25, opercular valves of
Balanus sp., cf. B. pacificus Pilsbry, LACM Locality 305; (23) exterior of scutum.
Hypotype LACM 1212, height 10 mm; (24) interior of scutum, Hypotype CAS 206,
height 7 mm; (25) interior of scutum. Hypotype LACM 1213, height 10 mm.

12

Contributions in Science

No. 159

Figures 26-32. Balanus proinus Woodring: (26-27) exterior and interior of rostrum,
Hypotype LACM 1214, LACM Locality 305C, height 6 mm; (28) group of shells,
Hypotype LACM 1215, LACM Locality 305C, width of group 15 mm; (29) interior
of scutum, Hypotype CAS 13166, LACM Locality 323, height 4 mm; (30-32) ex-
terior of scutum, exterior and interior of tergum, Hypotype LACM 1216, LACM
Locality 305C, height of scutum 3 mm; height of terga 3.2 mm.

1969

Thoracic Cirripedia of the San Diego Formation

13

denticulate on lower sides only; sheath dark red; scutum higher than broad,
with narrow, reflexed tergal margin; exterior of scutum sometimes bearing
single longitudinal row of square pits; adductor ridge absent; tergum broad,
spur furrow closed on scutal side only in young specimens, and completely
closed in older individuals; tergal spur narrow, elongate; depressor muscle
crests present.

Description : Shell cylindric; orifice large, subtriangular, untoothed;
parietes smooth, with narrow, reddish stripes on white background; parietal
tubes large, rectangular, open throughout length; one or two secondary septa
usually present on interior of outer lamella between primary parietal septa;
radii broad with nearly horizontal summits; sutural edges of radii with sec-
ondary septa on lower sides of primaries only; tubes of radii usually small; alae
broad with oblique summits; sheath dark red, one-third to one-half height of
shell, with dependent lower edge; interior of parietes sharply ribbed below
sheath; basis with single row of small, non-septate tubes.

Scutum reddish externally, especially on tergal half, white with pink cast
internally; ornamented externally by low, widely spaced major growth ridges,
between which are fine, closely spaced minor growth lines; a few scuta with
single longitudinal row of large, rectangular pits similar to those of B. trigonus
Darwin on tergal side of valve center; basitergal corner of scutum truncate
parallel to occludent margin; narrow strip of tergal margin sharply reflexed;
prominent articular ridge about three-fourths length of straight part of tergal
margin, slightly reflexed over deep, broad articular furrow; adductor ridge
absent; adductor pit large, oval, deep; apical portion of interior faintly to
markedly striate longitudinally; pit for depressor muscle deep, small, elongate.

Tergum white with pink cast, broad, flat, thin, with slightly sinuous basal
margin; width of basal margin equal to about two-thirds height of valve; spur
narrow, with nearly parallel sides, placed about its own width from basiscutal
angle; spur short, length about one and one-half times its own width; base of
spur pointed on scutal side; spur furrow closed, except in early stages where
carinal side not folded over furrow; articular ridge prominent, erect; articular
furrow broad; five or more moderately developed crests for lateral depressor
muscles.

Occurrence and Type Disposition : LACM Locs. 107, 122, 485 (type
locality); SDSNH Loc. L-2451; Holotype LACM 1224; Paratypes LACM
1217-1223; Paratypes CAS 13167-13168, 13170; Paratypes SDSNH 04258-
04259.

Range : Late Pliocene, San Diego Formation.

Remarks : This barnacle most closely resembles austral species of the
subgenus Megabalanus, rather than those of the tropics and the Northern
Hemisphere ( e.g Balanus tintinnabulum complex), in its possession of sec-
ondary denticulae on the lower sides only of the primary septa of the radii.
Young specimens of Balanus wilsoni resemble B. algicola Pilsbry from South

14

Contributions in Science

No. 159

Figures 33-41. Opercular valves of Balanus wilsoni sp. n., LACM Locality 485; (33)
interior of scutum, Paratype LACM 1217, height 10 mm; (34) interior of scutum,
Paratype LACM 1218, height 9 mm; (35) interior of scutum, Paratype CAS 13167,
height 9 mm; (36) exterior of scutum showing color banding, Paratype LACM 1219,
height 10 mm; (37) exterior of scutum with row of pits, Paratype CAS 13168, height
7.5 mm; (38) exterior of scutum with row of pits, Paratype LACM 1220, height 12
mm; (39) interior of tergum, Paratype LACM 1221, height 9 mm; (40-41) interior
and exterior of tergum, Paratype CAS 13170, height 14 mm.

1969

Thoracic Cirripedia of the San Diego Formation

15

Figures 42-46. Balanus wilsoni sp. n. ; (42) group of shells, SDSNH Paratype 04258,
SOSNH Locality L-2451, carinorostral length of orifice of left specimen 12 mm;
(43) exterior of tergum of young individual showing partially open spur furrow,
Paratype LACM 1222, LACM Locality 122, height 4 mm; (44) exterior of tergum
of older individual showing partially open spur furrow in apical half, Paratype
LACM 1223, LACM Locality 485, height 10 mm; (45) Holotype LACM 1224 (des-
ignated by “H”) associated with group of Balanus gregarius (Conrad), Hypotype lot
LACM 1202, LACM Locality 485, length of group 95 mm; (46) sutural edge of
radius of lateral plate (basal edge to left), Paratype SDSNH 04259, SDSNH Local-
ity L-2451, height 17 mm.

16

Contributions in Science

No. 159

Africa in having the spur furrow closed only on the scutal side, but older
specimens can be distinguished from this and other austral megabalanids with
unbeaked terga by their closed spur furrows.

Aside from the major difference observed in the structure of the sutural
edges of the radii, B. wilsoni can be distinguished from the extant Californian
and Mexican megabalanids B. tintinnabulum calif ornicus Pilsbry, B. tintinna-
bulum coccopoma Darwin, and B. tintinnabulum peninsularis Pilsbry by its
smooth, red striped parietes, the absence of an adductor ridge on the scutum,
and the presence of well defined depressor muscle crests on the tergum.

Large specimens of B. wilsoni were obtained only at the type locality (and
at SDSNH L-2451 which represents the same locality), where they were
attached to B. gregarius. Young adults attributed to this species were collected
at LACM Locs. 107 and 122. These differ from larger individuals in having an
uncolored sheath and a tergal spur furrow closed on the scutal side only. The
half closed nature of the spur furrow can be seen in apical portions of well
preserved terga of large specimens from LACM Loc. 485 (Figs. 43-44).

This species is named in honor of Dr. Edward C. Wilson.

Subgenus Hesperibalanus Pilsbry, 1 9 1 61
Balanus proinus Woodring, 1950

Figs. 26-32

Balanus hesperius Pilsbry, var. Woodring, in Woodring, Stewart, and Richards,
1940, pp. 30,97.

Balanus hesperius proinus Woodring, in Woodring and Bramlette, 1950, p. 92,
pi. 14, figs. 11, 15, pi. 16, figs. 1-3, 8-12.

Occurrence : LACM Locs. 305, 305C, 323.

Range : Pliocene, southern California.

Remarks : Balanus proinus is represented at the above mentioned locali-
ties by complete shells with contained opercular valves and several disarticu-
lated compartmental plates and isolated scuta. Fossil hesperibalanids are
common in Cenozoic deposits of the Pacific Coast of North America (Zullo,
1966), and are best distinguished at the specific level on characters of their
terga. Balanus proinus was originally described as a subspecies of B. hesperius
Pilsbry, but differs from it and its extant subspecies in the short, broad tergal
spur, and is here considered as a distinct species. The scuta of B. proinus also
differ from those of B. hesperius in having a longer basal margin and a less well
developed internal callus.

xHenry and McLaughlin (1967) have synonymized Hesperibalanus with Solidoba-
lanus Hoek, 1913.

1969

Thoracic Cirripedia of the San Diego Formation

17

Besides the type locality and the localities described herein, B. proinus
has been found in the Early Pliocene Pancho Rico Formation of Salinas Valley,
California (UCMP Locs. A-4947, A-7570, unpublished data), the Middle
Pliocene Foxen mudstone of the Santa Maria District, and the Late Pliocene
San Joaquin Formation of the Kettleman Hills, California (Woodring, in
Woodring and Bramlette, 1950, p. 92).

Genus Cetolepas gen. n.

Diagnosis'. Coronulines with short, nearly cylindric body chamber; parie-
tes bearing radial buttresses terminating in T-shaped flanges forming outer
wall; buttresses reduced or absent in older individuals; sheath decidedly, but
not strongly, grooved transversely, shorter than inner wall; inner surface of
radial plate strongly grooved; sutural edges of radii narrow, linear; short hori-
zontal sutural septa extend inwardly and outwardly from central vertical
septum of sutural edge; no specialized alar plate or open space between radii
and sheath; sutural edge of ala blunt, coarsely denticulate.

Type Species : Cetolepas hertleini sp. n.

Remarks : Cetolepas is similar to Cryptolepas Dali, but differs in the
fainter grooving of the sheath and, most significantly, in the narrow, linear
nature of the sutural edge of the radius, which consists of a single, vertical
septum from which short, horizontal septa branch alternately on either side.
In Cryptolepas rachianecti Dali and the fossil C. murata Zullo, the grooves of
the sheath are deeply incised and the sutural edges of the radii are much
broader and consist of a single series of septa that originate on the inner lamella
of the radial plate and branch profusely as they pass to the outer half of the
sutural edge in a complex pattern of crenulations.

Cetolepas differs from Cetopirus Ranzani in the possession of a grooved
sheath and in the less complex sutural edge of the radius. In Cetopirus the
completely branched sutural septa originate on the outer lamella of the radial
plate and extend basally.

Cetolepas hertleini sp. n.

Figs. 47-72

Diagnosis'. Same as for genus.

Description: Shell in younger stages of growth low conic, composed of
nearly cylindric body chamber from which T-shaped buttresses radiate to
form outer wall surrounding large basal cavities; shell conic in progressively
eroded older specimens; buttresses in older specimens reduced to low folds,
giving shell the appearance of a short Tubicinella Lamarck; body chamber
formed entirely by sheath; basal opening of body chamber nearly as large as

18

Contributions in Science

No. 159

orifice; inner margins of radial buttresses usually not projecting below sheath
in young conic forms; buttresses bifurcate towards base, ornamented externally
by conspicuous, regularly spaced, horizontal ridges cut by longitudinal striae
into fine beads; adjacent external flanges of buttresses not articulated or closely
appressed; in older stages, flanges more distantly spaced, leaving basal cavities
open to exterior; surface of large internal cavities longitudinally ribbed; walls
of buttresses composed of an outer lamella with single row of small longi-
tudinal tubes and an inner solid lamella; radii broad in conic forms, narrower
and more linear in cylindric specimens; sutural edges of radii linear, having a
single vertical septum from which horizontal septa alternately diverge in-
wardly and outwardly; spaces between outer row of horizontal septa unfilled,
forming single row of tubes in radii; inner row of horizontal septa greatly
reduced or wanting in older cylindric examples; no special alar plate and
cavity present inward of radii; alae rest directly on inner surface of radii;
sutural edges of alae blunt, coarsely denticulate.

Opercular valves unknown.

Occurrence and Type Disposition : LACM Locs. 305 (type locality),
305A, 305C, 319, 485; Holotype LACM 1230; Paratypes LACM 1225-1229;
Paratypes CAS 13171-13176.

Range : Late Pliocene, San Diego Formation.

Remarks : Cetolepas hertleini is readily distinguished from species of
Coronula Lamarck and Cryptolepas by the simple, linear nature of the sutural
edges of the radii. Older specimens resemble Tubicinella in general shell con-
figuration and in the form of the sutural edge, but differ in the plicate, rather
than smooth, basal edge, and the blunt and crenulate, rather than sharp and
smooth, edges of the alae.

As previously noted (Zullo, 1961 ) it would appear that the extant species
of Cryptolepas evolved from a coronulid ancestor similar to Coronula ( Cetopi -
rus) complanata (Morch) through a form resembling the Late Pleistocene
species Cryptolepas murata. In several respects, Cetolepas hertleini gives the
appearance of a form that might have been intermediate between Cetopirus
and Cryptolepas murata. However, the radically different sutural edge of the
new species indicates that this is not the case. It seems likely that the sutural
edge of Cetolepas hertleini is homologous with the outer, unbranched part of

Figures 47-58. Exteriors of compartmental plates of Cetolepas hertleini gen. and sp.
n., arranged from top left to lower right in increasing age and wear, LACM Locality
305; (47) rostrum, Paratype LACM 1225, height 16 mm; (48) lateral, Paratype
CAS 13171, height 10 mm; (49) rostrum, Holotype LACM 1230, height 11 mm;
(50) carina, Paratype LACM 1227, height 14 mm; (51) rostrum, Paratype CAS
13172, height 13 mm; (52) lateral, Paratype CAS 13173, height 11 mm; (53) lateral,
Paratype LACM 1228, height 12 mm; (54) rostrum, Paratype CAS 13174, height 16
mm; (55) lateral, Paratype LACM 1229, height 11 mm; (56) lateral, Paratype CAS
13175, height 8 mm; (57) lateral, Paratype CAS 13176, height 10 mm; (58) lateral,
Paratype LACM 1226, height 10 mm.

1969

Thoracic Cirripedia of the San Diego Formation

19

20

Contributions in Science

No. 159

Figures 59-72. Interiors of compartmental plates of Cetolepas hertleini gen. and sp.
n., as shown and arranged in figures 47-58; (59) Paratype LACM 1225; (60) Para-
type CAS 13171; (61) Holotype LACM 1230; (62) radial edge of Holotype LACM
1230; (63) Paratype LACM 1227; (64) Paratype CAS 13172; (65) Paratype CAS
13173; (66) Paratype LACM 1228; (67) Paratype CAS 13174; (68) Paratype
LACM 1229; (69) Paratype CAS 13175; (70) Paratype CAS 13176; (71) radial
edge of Paratype LACM 1226; (72) Paratype LACM 1226.

1969

Thoracic Cirripedia of the San Diego Formation

21

the sutural edge of Coronula complanata, thus suggesting derivation from the
Cetopirus stock, but independently of the lineage that gave rise to Cryptolepas.

The most striking evolutionary clues are those provided by what are here
interpreted as the older, eroded stages of growth in Cetolepas hertleini. In
these stages the shell is reduced to a simple, transversely ringed cylinder with a
broadly plicate basal margin. The inner row of horizontal septa are greatly
reduced or lacking, leaving only the outer row of septa to form the sutural edge.
The spaces between septa are unfilled, and the septa tend to branch towards
the outer lamella.

Thus, the morphology of these older stages is that of a Tubicinella, differ-
ing only in the basal plications and the crenulations of the alar sutural edges.
If the earlier stages of growth and especially the development of the radial
buttresses were arrested, it would appear a simple matter to derive the shell
of Tubicinella from a form similar to Cetolepas hertleini.

The derivation of Tubicinella from Cetolepas rather than from Cryptole-
pas is supported by a comparison with the growth stages seen in C. rachi-
anecti. Pilsbry (1916, p. 280) recognized two growth stages: 1) a young stage
with a thick disk, rounded periphery, and simple buttresses; and 2) an older
stage in which the buttresses are longer and freely branched. If these are
indeed growth stages of a single species, then their ages relative to each other
should be reversed. In the freely branched stage the radial buttresses converge
at the apex, are nearly parallel when viewed from the base, and the shell
exhibits little or no wear dorsally. In the cylindric stage the short, simple
buttresses are nearly parallel both dorsally and ventrally, and the dorsal surface
of the shell is clearly eroded. Thus, as in Cetolepas hertleini, the more complex
growth form with convergent buttresses represents the younger stages in the
development of the shell, which, as it grows and becomes progressively worn
away dorsally, assumes a cylindric form in which the buttresses cease to
converge. Except for the reduction of the buttresses, these older stages retain
the characteristics of the younger growth form, and exhibit no tendency to
assume the features characteristic of Tubicinella.

The generic name is derived from a combination of Cetopirus and
Cryptolepas. The species name honors Dr. Leo G. Hertlein.

Genus Coronula Lamarck, 1802

Coronula barbara Darwin, 1854 (?)

Figs. 73-77

Coronula barbara Darwin, 1854a, p. 421, pi. 15, fig. 6; 1854b, p. 38, pi. 2,
figs. 8a-e; Alessandri, 1895, p. 72, pi. 3, figs. 8a-b; 1906, p. 317, pi. 18, fig. 12.

Occurrence: LACM Loc. 305.

Range: Pliocene and Early Pleistocene, Europe; (?) Late Pliocene,
southern California.

22

Contributions in Science

No. 159

Figures 73-77. Coronula bcirbarci Darwin (?); (73-75) exterior, alar edge, and radial
edge, Hypotype LACM 1231, LACM Locality 305, height (73-74) 36 mm, (75) 30
mm; (76-77) exterior and sculpture, Hypotype USNM 651308, U. S. Geological Sur-
vey locality M-2090, width (76) 36 mm, height (77) 35 mm.

Remarks'. The single large specimen agrees with Darwin’s (1854a, b)
description of C. barbara in having prominent transverse ridges on the ex-
terior of the shell and the interior surfaces of the transverse flanges, and in the
zig-zag pattern of the teeth that serve to interlock adjacent flanges of the outer
shell wall. The space between the radius and the special alar plate is not as
completely filled in the San Diego specimen as it is in the examples figured by
Darwin (1854a, b) and Alessandri (1906), and more closely approximates
the condition seen in C. diadema (Linnaeus). For this reason the identification
of the San Diego specimen with C. barbara is questioned. Another specimen
that is specifically identical to the San Diego Coronula was collected by Dr. J.
Vedder of the U.S. Geological Survey, Menlo Park, California, from the Late

1969

Thoracic Cirripedia of the San Diego Formation

23

Pliocene upper part of the Fernando Formation of the Tustin quadrangle,
Orange County, California (U.S. Geological Survey Loc. M-2090, 1650 ft.
SE and 900 ft. SW of N corner Irvine Block 57 ) .

Keys to the Balanomorph Cirripedia
of the San Diego Formation

Scuta

1. Exterior longitudinally striate 2.

Exterior lacking longitudinal striae 4.

2. Adductor ridge clearly present 3.

Adductor ridge absent Balanus kanakofji

3. Adductor ridge bifurcate Balanus gregarius

Adductor ridge single Balanus pacificus

4. Adductor ridge present 5.

Adductor ridge absent Balanus wilsoni

5. Ridged callus between articular and adductor ridges Balanus proinus

No such callus; vertical rib present

in depressor muscle pit Balanus nubilus

Terga

1. Spur furrow open 2.

Spur furrow at least partially closed 4.

2. Base of spur truncate 3.

Base of spur broadly rounded Balanus proinus

3. Juncture of spur with basal margin

gently arched on both sides Balanus nubilus

Juncture angular on scutal side,

gently arched on carinal side Balanus kanakoffi

4. Exterior growth lines crossed by longitudinal striae at

least on scutal side Balanus gregarius

No longitudinal striae externally 5.

5. Base of articular ridge confluent with scutal margin Balanus pacificus

Base of articular ridge widely

separated from scutal margin Balanus wilsoni

Shells

1 . Compartmental plates solid throughout Balanus proinus

Compartmental plates, when viewed from base, with rows

of pores or large enclosed cavities 2.

2. Radii, as well as parietes, porose 3.

Radii without pores 4.

3. Shell smooth externally, with color stripes Balanus wilsoni

Shell with conspicuous horizontal and/or vertical ridges

forming distinct external sculpture Cetolepas hertleini

24

Contributions in Science

No. 159

4. Parietal “septa” as seen from base in form of T-shaped
buttresses enclosing large basal cavities; base of shell

wall as thick as diameter of body cavity Coronula spp.

Parietal septa not T-shaped; shell wall much thinner

than diameter of body cavity 5.

5. Parietal tubes with transverse septa; basis usually
elongate and filled at least in part with

vesicular material Balanus gregarius

Parietal tubes without transverse septa;

basis flat and not vesiculose 6.

6. No radii visible externally between

carina and carinolaterals Balanus kanakoffi

Radii clearly present between carina and carinolaterals 7.

7. Shell externally roughened and plicate;

without color pattern Balanus nubilus

Shell smooth externally; usually

showing color stripes Balanus pacificus

Literature Cited

Alessandri, G. de. 1895. Contribuzione alio studio dei Cirripedi fossili d’ltalia.
Societa Geol. Ital., Boll., 13: 234-314, text-figs. 1-2, pis. 3-5.

. 1906. Studi monographici sui Cirripedi fossili d’ltalia. Palaeontogr. Ital. 12:

207-324, text-figs. 1-9, pis. 13-18.

Allison, E. C., J. W. Durham, and V. A. Zullo. 1961. Cold-water late Cenozoic
faunas of northern California and Oregon. Geol. Soc. Amer., Spec. Paper, 68: 2
(abstr.).

Conrad, T. A. 1856. Descriptions of three new genera; twenty-three new species
Middle Tertiary Fossils from California, and one from Texas. Acad. Nat. Sci.
Philadelphia, Proc., 8: 312-316.

. 1857a. Description of the Tertiary fossils collected on the Survey. In Pacific

Railroad Reports, 6 (Geol. Rpt., 2) :69-73, pis. 2-5.

„ 1857b. Report on the palaeontology of the Survey. Ibid., 7: 189-196, pis.

1-10.

. 1864. Notes on Shells, with descriptions of new fossil Genera and Species.

Acad. Nat. Sci. Philadelphia, Proc., 16: 211-214.

. 1876. Note on a Cirripede of the California Miocene, with remarks on fossil

shells. Acad. Nat. Sci. Philadelphia, Proc., 28: 273-275.

. 1877. Note on the relations of Balanus estrellanus of the Californian Mio-
cene. Amer. Jour. Sci. (3)13: 156-157.

Dall, W. H. 1902. On the true nature of Tamiosoma. Science (n.s.) 15: 5-7.

Darwin, Charles. 1854a. A monograph on the sub-class Cirripedia 2. Balanidae,
Verrucidae. Ray Soc., London, viii + 684 pp., 1 1 text-figs., 30 pis.

. 1854b. A monograph on the fossil Balanidae and Verrucidae of Great Brit-
ain. Palaeontogr. Soc., London, 44 pp., 6 text-figs., 2 pis.

Durham, D. L., and W. O. Addicott. 1965. Pancho Rico Formation Salinas Valley,
California U.S. Geol. Surv. Prof. Paper 524A: i-iii, 1-22, text figs. 1-7, pis. 1-5.

Gabb, W. M. 1869. Cretaceous and Tertiary fossils. California Geol. Surv., Paleon-
tology, 2: i-xiv, 1-299, pis. 1-36.

1969

Thoracic Cirripedia of the San Diego Formation

25

Henry, D. P., and P. A. McLaughlin. 1967. A revision of the subgenus Solidoba-
lanus Hoek (Cirripedia Thoracica) including a description of a new species with
complemental males. Crustaceana 12: 43-58, text-figs. 1-3, pi. 1.

Hertlein, L. G., and U. S. Grant IV. 1944. The geology and paleontology of the
marine Pliocene of San Diego, California. Pt. 1, Geology. San Diego Soc. Nat.
Hist., Mem., 2: 1-72, text-figs. 1-6, pis. 1-18.

. 1960. The geology and paleontology of the marine Pliocene of San Diego,

California. Part 2a, Paleontology. San Diego Soc. Nat. Hist., Mem., 2: 73-133,
pis. 19-26.

Pilsbry, H. A. 1916. The sessile barnacles (Cirripedia) contained in the collections
of the U. S. National Museum, including a monograph of the American species.
U. S. Natl. Mus., Bull. 93: i-xi, 1-366, text-figs. 1-99, pis. 1-76.

. 1919. Cirripedia from the Panama Canal Zone. U. S. Natl. Mus., Bull. 103:

185-188, pi. 67.

Ross, Arnold. 1962. Results of the Puritan-American Museum of Natural History
Expedition to Western Mexico. 15. The littoral balanomorph Cirripedia. Amer.
Mus. Novitates 2084: 1-44, figs. 1-24.

. 1964. Cirripedia from the Yorktown Formation (Miocene) of Virginia.

Jour. Paleontol., 38: 483-491, text-figs. 1-2, pis. 71-72.

Withers, T. H. 1953. Catalogue of fossil Cirripedia in the Department of Geology,
vol. III. Tertiary. Brit. Mus. (Nat. Hist.), London, xv -f 396 pp., 105 text-figs.,
64 pis.

Woodring, W. P., and M. N. Bramlette. 1950. Geology and paleontology of the
Santa Maria District, California. U. S. Geol. Surv., Prof. Paper 222: i-iv, 1-185.
text-figs. 1-9, pis. 1-23.

. R. Stewart, and R. W. Richards. 1940. Geology of the Kettleman Hills oil

field, California. U. S. Geol. Surv., Prof. Paper 195: 1-170, text-figs. 1-15, pis.
1-57.

Zullo, V. A. 1961. A new whale barnacle from Late Pleistocene deposits at San
Quintfn Bay, Baja California. Veliger 4: 13-14, pi. 3.

. 1964. Re-evaluation of the late Cenozoic cirriped “Tamiosoma” Conrad.

Biol. Bull. 127(2) : 360 (abstr.).

•. 1966. A new species of Balanus (Cirripedia, Thoracica) from the Late

Eocene Cowlitz Formation of southern Washington, U.S.A. Crustaceana 1 1 :
198-204, text-figs. 1-2.

Accepted for publication January 15, 1969.

LOS

ANGELES

CONTRIBUTIONS
IN SCIENCE

COUNTY

MUSEUM

lUMBER 160

March 27, 1969

^7/ 7^

2-/

OBSERVATIONS ON THE BEHAVIOUR OF TWO SPECIES
OF HONEY-GUIDES

Indicator variegatus (Lesson) and Indicator exilis (Cassin)

By A. L. Archer and Robert M. Glen

Los Angeles County Museum of Natural History • Exposition Park
Los Angeles, California 90007

CONTRIBUTIONS IN SCIENCE is a series of miscellaneous technical papers
in the fields of Biology, Geology and Anthropology, published at irregular intervals
by the Los Angeles County Museum of Natural History. 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 Robert J. Lavenberg, Los Angeles County Museum of Natural
History, 900 Exposition Boulevard, Los Angeles, California 90007.

Robert J. Lavenberg
Managing Editor

INSTRUCTIONS LOR AUTHORS

Manuscripts for the LOS ANGELES COUNTY MUSEUM, CONTRIBU-
TIONS 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. Al-
though priority will be given to manuscripts by staff members, or to papers dealing
largely with specimens in the Museum’s collections, 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 Edi-
torial Committee. They may also be subject to review by competent specialists out-
side 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.

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. Alterations or changes in the manuscript after galley proof will
be billed to the author. Unless specifically requested, page proof will not 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.

Dorothy M. Halmos
Editor

OBSERVATIONS ON THE BEHAVIOUR OF TWO SPECIES
OF HONEY-GUIDES

Indicator variegatus (Lesson) and Indicator exilis (Cassin)

By A. L. Archer and Robert M. Glen1

Abstract: Twenty-two specimens of Indicator exilis and
eight specimens of Indicator variegatus were collected either at
or in the close vicinity of three wild bees’ nests situated in one
part of the Malabigambo Forest. The high density of both species,
particularly I. exilis, at the beehives and not elsewhere, suggests
that the birds had pre-located the hives and visit them periodical-
ly. The possibility of olfactory attraction as an aid to beehive lo-
cation is discussed, together with other locating aids. A method of
capturing Honey-guides is described.

Between 2 February and 19 February, 1968, the authors of this paper
made records of possible numbers of Honey-guides visiting three wild bee-
hives. The short study took place in Malabigambo Forest, 0° 57' S, 31° 36' E,
southern Uganda, at an elevation of 3,000 feet. Twenty two specimens of
/. exilis and eight specimens of /. variegatus were secured and preserved as
study skins. The seemingly large series of /. exilis obtained was deemed
necessary due to the great similarity of five sibling Honey-guide species,
/. exilis, I. pumilio, /. willcocksi, /. meliphilus and /. minor. The three wild
beehives located were all situated in similar presently dry seasonal swamp
forest, and were easily accessible for checking a minimum of three times a day
and frequently four or five times. Experience showed that the hours most
favoured by Honey-guides for hive visitations were between 10 am and 5:30
pm, which coincided with the normal peak periods of sun during the day.

For the purpose of this paper the three hives are identified as A, B and C.
All three hives were located within 40 yards of a straight abandoned rail track
cut by a logging company into the heart of Malabigambo Forest. Flive C lay
some 1,600 yards to the north of hive B, whilst hive B was situated approxi-
mately 750 yards north of hive A. The greatest distance between the two
outside beehives was therefore 2,350 yards in a straight line. It must be stressed
however that the distances between the hives were not accurately measured
but are only approximate.

The intention of this paper is to demonstrate that it seems probable that
the opportunity to collect numbers of each of these two species of Honey-
guides was a result, not of their constant density within the limited area, but
rather of their habit of making periodical visits to all wild beehives known to
them, possibly over a larger area of the forest. It should be pointed out that a
total of twenty five days were spent collecting in Malabigambo and despite
there being an aggregate total of over 2,500 yards of mist nets utilized within

’Field representatives, Los Angeles County Museum of Natural History.

1

?

Contributions in Science

No. 160

Table I (Hive A)
Indicator exilis

No.

Date

Netted
at hive

Shot
at hive

Netted
adjacent
to hive

SB. 278

2.2.68

male

SB. 301

3.2.68

female

—

— -

SB. 413

6.2.68

—

—

male

SB. 439

7.2.68

—

—

male

SB. 440

7.2.68

—

—

female

SB. 623

14.2.68

male

— .

—

SB. 635

14.2.68

male

—

—

SB. 636

14.2.68

female

—

—

SB. 656

15.2.68

—

male

— -

SB. 657

15.2.68

—

female

—

SB. 760

19.2.68

—

male

—

SB. 279

2.2.68

Indicator variegatus

male

SB. 351

5.2.68

—

—

female

this period, not a single Honey-guide was netted except in the immediate
vicinity of the beehives. In addition, neither of us observed /. exilis away from
a hive in the field and on only one occasion was /. variegatus seen. One sped
men of /. exilis was netted in Namalala Forest in a line of nets set at random.

Details follow for the three wild beehives, giving the number of Honey-
guides secured at each and the time intervals involved.

Hive A : This hive was located on 1 February 1968. The natural bees’
nest was some ten feet up in the hollow of a living tree. It was chopped out the
same day and contained mainly empty comb with a little honey and few larvae.
After the bees had settled down, Archer erected a mist net at the base of the
tree and with the aid of three poles was able to completely circumpose the
trunk. Pieces of comb, some with honey and larvae, were used as bait and
were either wedged into bark on the trunk or impaled on twigs at a level about
half way up the net, between the tree trunk and surrounding mist net. The
bottom of the net was at ground level. Pieces of comb were burnt over the
embers of a small fire used to keep the bees away when the hive was being cut
out. The resulting aroma of melting beeswax could be detected by a human at
over twenty yards distance.

The first Honey-guide, an adult male Indicator exilis, was netted the
following day. During the next seventeen days a total of four more /. exilis

1969

Behaviour of Honey-Guides

3

Table II (Hive B)
Indicator exilis

No.

Date

Netted
at hive

SB. 327

4.2.68

male

SB. 333

5.2.68

male

SB. 334

5.2.68

male

SB. 519

11.2.68

male

SB. 520

11.2.68

male

SB. 572

12.2.68

female

SB. 637

14.2.68

female

SB. 740

18.2.68

male

SB. 741

18.2.68

male

SB. 332

Indicator variegatus
5.2.68

male

SB. 476

9.2.68

female

were netted at the hive; in addition three others were shot while feeding at the
cut-out bees’ nest and three more were caught in mist nets set up within 50
yards of the hive. (See Table I for dates specimens were collected.)

No specimens of Indicator variegatus were captured at hive A; however,
one specimen of this large Honey-guide was shot within a few yards of the hive
and a second was netted in the line of mist nets set up close by.

It is of importance to note that on no occasion was fresh comb added to
the original bait used at hive A. The last specimen of Indicator exilis was
collected at this hive eighteen days after it was first chopped out, a bird shot by
Glen while it perched at the mouth of the abandoned bees’ nest.

Hive B: Located on 3 February 1968; Glen on the same day erected a
mist net after the bees’ nest had been opened up. This hive was contained in a
dead hollow tree which had fallen but remained propped up at a 45 degree
angle, the entrance being approximately 12 feet from the ground. The net was
set up below the hive and baited in much the same way as at hive A; it en-
compassed all the bait but a gap of about 18 inches was left between the
ground and the bottom of the net. This method proved to be more successful
than that used at hive A, possibly because the birds could enter the net from
below as well as from above.

Between 4 and 18 February a total of nine Indicator exilis and two
Indicator variegatus were caught at hive B. (See Table II for dates.)

As in the case of hive A, no fresh bait was added to the original honey
comb. This comb, after exposure to a considerable amount of rain and damp-

4

Contributions in Science

No. 160

Table III (Hive C)
Indicator exilis

No.

Date

Netted
at hive

SB. 559
SB. 680

12.2.68

17.2.68

male

male

Indicator variegatus

SB.460
SB. 517
SB. 649
SB. 717

8.2.68

10.2.68

15.2.68

18.2.68

male

male

male

female

ness, developed a coating of mould and although pieces were knocked to the
ground it did not appear that any wax was eaten.

Hive C : This hive possibly produced the most interesting results. Unlike
hives A and B, hive C was not cut out because the bees’ nest was situated in the
trunk of a dead tree which had been smothered by a large parasitic fig, Ficus.
The mechanics of chopping out the hive were made impossible by the abund-
ance of latex issuing from the Ficus on every axe blow, the latex being par-
ticularly aggravating to the eye. After more than two hours the attempt was
given up. The entrance to the hive was some 15 feet from the ground and on
this occasion our mammal trapper, Mwangi Kariuki, set up a mist net at the
base of the tree on the same principle used at hive B. Pieces of honey comb
from elsewhere were used as bait. It should be noted that although the bees’
nest itself was untouched, the usual procedure had been adopted, with a fire
being lit and highly audible chopping taking place.

The initial attempt to cut out the hive and the setting of the net took place
on 7 February. On 8 February a specimen of Indicator variegatus was caught.
This was followed, during the next ten days, by a further catch of three more
specimens of I. variegatus and two of Indicator exilis. (See Table III for dates.)

Once again no fresh honey comb was added to the bait and the bees
remained unmolested.

Discussion : The data shown above gives rise to two interesting possi-
bilities. In the first instance it has been shown that no specimens of Indicator
exilis were either netted or observed at a distance greater than 50 yards from a
known natural bees’ nest in Malabigambo Forest. The three hives under dis-
cussion were situated in a straight line with the greatest distance between the
two end hives being the relatively short space of approximately 2,350 yards.
Despite extensive netting undertaken along the track which linked the three
hives, all the Honey-guides collected were obtained at these hives except for

1969

Behaviour of Honey-Guides

5

the three caught in the nets within 50 yards of hive A. These factors give
weight to the presumption that the three hives were all known to the Honey-
guides secured and the fact that the birds were caught over a period of nearly
three weeks supports this hypothesis. The conclusions from these results is the
strong possibility that wild bees’ nests throughout a large area of forest are
known to Honey-guides frequenting it and are visited periodically by the birds
in the hope of obtaining beeswax and other dietary items. A number of natural
hives have sufficiently large entrances to allow a Honey-guide to enter and
obtain beeswax without the hive being opened by human symbionts. It must
be assumed that Honey-guides cover a large area of forest if specimens can be
consistently secured at a single hive over a period of two weeks or more.

The second problem lies in how the natural hives are originally located by
Honey-guides. It seems unreasonable to assume that in dense forest a Honey-
guide would be capable of following an insect as fast in its movements as a bee.
A possible alternative is that Honey-guides are vested with a highly developed
sense of smell, as indicated by Stager (1967). Both of us, in other Uganda
forests, have had experience of trapping Honey-guides, well away from any
known bees’ nest, by simply burning honey comb and setting up mist nets
designed to trap any bird investigating the smell. In Bwamba Forest of western
Uganda, at least three specimens of /. maculatus were thus secured and in
Budongo Forest a single I. conirostris fell victim to this ruse. It would seem
probable that the location of many bees’ nests is assisted through olfactory
attraction by Honey-guides, although undoubtedly a number must inevitably
be discovered by chance. It is possible that the sight or sound of bees swarming
at a hive could attract a Honey-guide to a nest. This possibility must not be
overlooked and in all probability does apply in some instances. It must be
borne in mind, with regard to the three hives under discussion, that the bees
had vacated two of the three nests within 48 hours of their being opened up. In
spite of this, Honey-guides were captured for up to nearly two weeks after the
bees had disappeared from hives A and B.

An interesting exercise over a long period would be the banding and
releasing of Honey-guides netted at known hives in a large given area of forest.
The results would undoubtedly bring a closer understanding of the behaviour
patterns regarding this complex avian family.

Acknowledgments

We must thank Dr. Herbert Friedmann for much valuable advice and
helpful suggestions regarding this paper. Mr. I. S. C. Parker also offered helpful
criticisms of the manuscript.

At the time the observations were made, we were employed by the Los
Angeles County Museum of Natural History, under National Science Founda-
tion Grant GB-5107, and we are grateful to Dr. Herbert Friedmann for
permission to publish this paper.

6

Contributions in Science

No. 160

Literature Cited

Friedmann, Herbert. 1955. The Honey-guides. U. S. Natl. Mus. Bull., 208: 1-292.
Stager, Kenneth E. 1967. Avian Olfaction. Amer. Zoologist, 7: 415-420.

Accepted for publication January 24, 1969.

Lfbg

LOS

ANGELES

COUNTY

MUSEUM

CONTRIBUTIONS

IN SCIENCE

II

UMBER 161

April 3, 1969

FURTHER NOTES ON THE AFRICAN MOLOSSID BAT
TAD ARID A ALOYSI1SABA UDIAE

By R. L. Petersgn

Los Angeles County Museum of Natural History • Exposition Park
Los Angeles, California 90007

CONTRIBUTIONS IN SCIENCE is a series of miscellaneous technical papers
in the fields of Biology, Geology and Anthropology, published at irregular intervals
by the Los Angeles County Museum of Natural History. 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 Robert J. Lavenberg, Los Angeles County Museum of Natural
History, 900 Exposition Boulevard, Los Angeles, California 90007.

Robert J. Lavenberg
Managing Editor

INSTRUCTIONS FOR AUTHORS

Manuscripts for the LOS ANGELES COUNTY MUSEUM, CONTRIBU-
TIONS 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. Al-
though priority will be given to manuscripts by staff members, or to papers dealing
largely with specimens in the Museum’s collections, 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 Edi-
torial Committee. They may also be subject to review by competent specialists out-
side 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.

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. Alterations or changes in the manuscript after galley proof will
be billed to the author. Unless specifically requested, page proof will not 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.

Dorothy M. Halmos
Editor

FURTHER NOTES ON THE AFRICAN MOLOSSID BAT
TAD ARID A ALOYS1ISABA UDIAE

By R. L. Peterson1

Abstract: Four additional specimens of Tadarida aloysii-
sabaudiae are recorded, bringing the total known number to
seven. Two are fully adult males, the first known. Sexual di-
morphism is described and illustrated in the relative sizes of the
anterior premolars. The conformations of the basisphenoidal pits
are described and their taxonomic value discussed. Tadarida rus-
sata is shown to be the closest known living relative, differing pri-
marily in relative size (shorter skull and forearm length), and its
only known locality is mapped along with those of T. aloysiisa-
baudiae.

Only three specimens of the central African bat Tadarida aloysiisabaudiae
have been previously reported (Peterson, 1967). These consisted of two adult
females and a subadult male. Two additional specimens have since been
received by the Royal Ontario Museum ( a second adult female and a subadult
male), and two specimens have now been identified in the collections of the
Los Angeles County Museum of Natural History (the first known adult
males).

The holotype was taken from the Toro district east of Ruwenzori, West-
ern Province, Uganda (Fesfa, 1907, 1909; see Fig. 1, No. 1), whereas the
second specimen (a subadult male) was collected at Avakubi, Oriental
Province, Congo (Allen et al. 1917; Koopman, 1965; see Fig. 1, No. 2). The
original and two subsequent specimens in the Royal Ontario Museum came
from near Masindi in the Budongo Forest, Uganda, and were collected by
John G. Williams or his associates (two in 1966, one in 1968; see Fig. 1,
No. 3). The two adult males in the Los Angeles County Museum of Natural
History collection (Cat. Nos. 27478 and 27479) were collected in the Kibale
Forest, 0° 50' N, 31° 06' E (approximately 35 miles east of the type locality)
on November 9, 1966, by Robert Glen and Andrew Williams (see Fig. 1,
No. 4), under NSF grant GB-5107.

A review of the measurements of the seven known specimens (Table 1 )
shows that the adult males tend to be larger than the adult females in the
following cranial characters: palatal length, mastoid breadth, lachrymal
breadth of rostrum, interorbital breadth, and in the breadth and the height of
the braincase. As in several other molossid bats, the anterior lower premolar
is relatively heavier in the males. In the occlusal view, the anterior premolar is
approximately equal to the posterior premolar, whereas in the females it is
noticeably smaller than the posterior one (see Fig. 2).

department of Mammalogy, Royal Ontario Museum, Toronto, Ontario, Canada.

1

2

CONT IIBUTIONS IN SCIENCE

No. 161

In a preliminary review of a number of the species in the family
Molossidae, it has been found that the conformation of the basisphenoidal pits
offers an important systematic character which is highly specific in distinguish-
ing species that appear superficially to be closely related. Moreover, there

7° 2

i° 2

?° 3i

D° 31° 3

2° 33° 34° 35

v.

-A

S

/

/

/

i

\ — •

UA

/

<

-v.

If

V — /

i

i

k

A

\

:

1

K

V.

yrfLAlb

ert

\,

t

f

\

02

o~

o

Jy

i ^

^03

u

GAN

D A

/

J

r'

>

f

/

«i

d4

t'

>

O

/

dward

K’-\ - ' ‘

' V
1

o

\

jL^

Kivu

/ S
\

}

: Lake V

ictor

RUAN

\URUN

)

j

3A4
PI w..

itm§

)

pAL.Tan

i

C>

1

i

ga/fyika

TAN C

i A NY

KA

OT

f

/

/

TYPE LOCALITIES
• — T. aloysiisabaudiae
A — T. russata

Figure 1. Map of central Africa showing the known distribution of Tadarida aloysii-
sabaudiae and the type locality of T. russata. 1. Toro district, Western Province,
Uganda. 2. Avakubi, Oriental Province, Congo. 3. Budongo Forest near Masindi,
Uganda. 4. Kibale Forest, Uganda.

1969

Notes on the African Molossid Bat

3

Figure 2. Sexual dimorphism in the relative size of the anterior lower premolar and in
the basisphenoidal pits. Left: LACM 27478 $ from Kibale Forest. Right: ROM
41768 $ from Budongo Forest, Uganda.

appears to be a general correlation between the development of the pits with
the size and complexity of the ears, with simple-eared forms such as Platymops
having the least developed basisphenoidal pits, and species with larger and
more complex ears having the highest developed pits, as seen in the genus
Otomops. It is postulated that these two major characters are probably closely
correlated with the echolocation mechanism. With this possibility in mind, a
detailed study of this variation in basisphenoidal pits throughout the family is
underway. In general, there appear to be some slight variations that may be
correlated with age and, in some cases, with sex. In the case of the seven
known examples of T. aloysiisabaudiae, the adult males appear to have slightly
larger pits, with a narrower dividing septum between them, than in the case of
the subadult males and adult females (Fig. 2). The detailed conformation of
these pits appears to be distinctive from all other African molossids examined
to date, except perhaps T. russata, which has been compared only by means of
photographs of the holotype and an additional pair of paratypes. Even here
there appear to be minor differences between the two taxa, but detailed studies
of T. russata are still needed to make adequate comparisons. The medial walls
of the pits are expanded beneath the central septum in T. aloysiisabaudiae, and
in adults the distance between becomes less than .2 mm. The breadth of the
dividing septum at its narrowest point ranges from .4 in young to .2 mm or
less in adult males.

In general, there seems little doubt that T. russata, known only by the
original type series (27 specimens) , is the closest known relative of T. aloysiisa-
baudiae. The basic shape of the skull is strikingly similar but differs in being
shorter (greatest length, 17.8-19.4 (holotype, 18.7) as compared to 20.4-22.1
in the known examples of the latter). The condylobasal length of the holo-

Measurements of the known specimens of Tadarida aloysiisabaudiae AMNH (American Museum of Natural History),
LACM (Los Angeles County Museum of Natural History), ROM (Royal Ontario Museum)

4

Contributions in Science

No. 161

§ O'

I?

Is

3“

£ §

^ co

r- i-h

<n

OT

q

H

q

v™H

o

00

d

O

d

t-h OT

r-H

<N

in

IT)

cs

<N

in

in o

(N oT

q

ON

q

q

c4

r"H

q

On

—i of

(Nl

oT

in

CN

<N

oT

o

no

o

o

hH

n

o

(N

h-H

o

q

o 6

00

t-H

CO

ri

•n

>n

CM

oT

°o

On

00

<N (N

00

O

CO

^H (N

or

in

q

q

oo

oT

00

or co

CO

<N

r-H

q

ON

o

T“ H
t-H

or t-(

(N

in

in

<N

in

q

q

q

q

(N

co ri

<N

CO

ri

q

<N

r-H

or

(N

>n

•n

(N

n

•n

O O <N

rH (N

00

cc

ON

o

r-H OT

n in

>n <N

r-H

in

o

q

q

vd

— x
& a

Q-i oi

c3

43

C3

c3 2
O 43
C3 Oh
O hh 'g

HH 1/3 G

2 ^ <N

^ X
5 0
Oh c3

S-H

§ a

W)

a

2nd phalanx 10.7 12.3 11.8 — 10.2 11.1 10.8

Fifth Digit: Metacarpal 28 31.3 31.5 — 30.8 31.1 30.5

1st phalanx 14.2 15.2 14.7 — 14.4 14.4 14.6

2nd phalanx 3.9 4.1 4.5 — 3.6 4.5 4.5

Cranial Measurements

1969

Notes on the African Molossid Bat

5

in

00

N;

N;

on

©

N;

N;

m

00

r-

in

in

NO

(N

ON

oo

ri

N"

o

00

in

xj-

00*

NO

o\

N;

NO

NO

ON

00

m

in

NO

in

NO

in

l>

o\

00

ON*

i— H

NO

d

00*

in

00*

r)

r-H

<N On rn
00 1 I iri rt

O) »o n t-h

o\ I 6 in oo

in a\ «n
oo h d

Xt- 00
^ NO

O N; in NO >n

O oo h in d

+1

Os m on

ri o oo <N —

N; o no

NO N"

<N N; NO N; G\ oo
o oo 6 oo

ro Tj- <N NO m ON VO

-d C\ 00* <N 1-H NO NO

^ o h h oo iri n t
t ^doohiritoo

in O ro ri O oo ro oo

t-h On 00 (N* rH N> I <d 6 d I l>

<N ^ -H I I

OX)

G

o

_ <L>

5 .>

ox) ‘35

o

G

O

70

G

O

U

G

O

W> S

N S

£ §

c/3 O

O _

03

<D

NO

<u

<Z)

40

<*>

o3

’T3

S-H

oS

CD

03

o

0)

G

<D

NO

on

O

.s

*c3

S-H

’S3

t-H

40

*c3

M— 1

NO

t+H

o

mh

o

3

Ih

NO

O

NO

O

r3

3

d

i-i

c3

#op

03

Q

G

CD

)H

PQ

*<d

E

<U

I-H

PQ

£ o

6 6

6

Contributions in Science

No. 161

type is 16.6, as compared to 18.5 for a subadult and 19.4-19.9 in the five
known adults of T. aloysiisabaudiae. In zygomatic width, Allen et al. (1917)
show that 23 specimens of T. russata range from 10.5 to 11.4, whereas in six
of the known T. aloysiisabaudiae, one subadult is 11.1 and the adults range
from 1 1.3 to 1 1.7. In forearm length, 24 of the original T. russata ranged from
42 to 46 as compared to 48.8 and 51.0 in the two subadults, and 49.5 to 52.7
in five adults of T. aloysiisabaudiae. In general external characters the two
taxa are strikingly similar, particularly in coloration, where they are
essentially indistinguishable.

The type locality of T. russata at Meje, Congo (see Fig. 1 ), is only about
80 miles distant from the nearest locality of T. aloysiisabaudiae at Avakubi,
Congo. Although it is not yet possible to demonstrate the true relationships of
these two forms, their allopatric distribution and the fact that they seem to
differ only in size, provide a strong suggestion that T. russata might ultimately
be regarded as only a well-marked geographical race of T. aloysiisabaudiae.

Acknowledgments

I am grateful to Donald R. Patten and the Los Angeles County Museum
of Natural History for the opportunity to study the African molossid speci-
mens in their collections and for other cooperation extended. I am also
indebted to R. G. Van Gelder and Karl F. Koopman for their cooperation in
permitting me the opportunity to study holotypes and other specimens in the
American Museum of Natural History. John G. Williams, Robert Glen, Terry
Shortt and Paul Geraghty all made valuable contributions in the securing of
specimens. Mrs. Sophie Poray prepared the illustrations. The assistance and
cooperation of the ROM Department of Mammalogy staff is gratefully
acknowledged. This paper is a product of a larger bat research program which
has had the financial assistance of the Canadian National Sportsmen’s Show
and the National Research Council of Canada.

Literature Cited

Allen, J. A., H. Laing, and J. P. Chapin. 1917. The American Museum Congo ex-
pedition collection of bats. Am. Mus. Nat. Hist., Bull., 37: 405-563.

Festa, E. 1907. Nyctinomus Aloysii-Sabaudiae, nov. spec. Turin, Univ., Musei Zool.
Anat. Compar., Boll., 22 (546) : 1-2.

. 1909. Chirotteri ed insettivora. In Luigi, duke of Abruzzi, II Ruwenzori,

Relazioni Scientifiche. Ulrico Hoepli, Milan. 1: 75-88.

Koopman, Karl F. 1965. Status of forms described or recorded by J. A. Allen in
“The American Museum Congo expedition collection of bats.” Am. Mus. Novi-
tates, 2219: 1-34.

Lanza, B., and D. L. Harrison. 1963. A new description of the type specimen of
Nyctinomus aloysiisabaudias Festa, 1907. Zeitschriftf. Saugertierkunde, 28: 102-
107.

Peterson, Randolph L. 1967. A new record for the African molossid bat Tadarida
aloysiisabaudiae. Can. J. Zool., 45: 449-452.

Accepted for publication January 28, 1969.

MUSEUM

ANGELES

LOS

COUNTY

CONTRIBUTIONS
IN SCIENCE

lUMBER 162

April 24, 1969

Bol-V

CiLm

THE BIRDS OF THE SANGO BAY FORESTS,
BUDDU COUNTY, MASAKA DISTRICT, UGANDA

By Herbert Friedmann and John G. Williams

Los Angeles County Museum of Natural History • Exposition Park
Los Angeles, California 90007

CONTRIBUTIONS IN SCIENCE is a series of miscellaneous technical papers
in the fields of Biology, Geology and Anthropology, published at irregular intervals
by the Los Angeles County Museum of Natural History. 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 Robert J. Lavenberg, Los Angeles County Museum of Natural
History, 900 Exposition Boulevard, Los Angeles, California 90007.

Robert J. Lavenberg
Managing Editor

INSTRUCTIONS FOR AUTHORS

Manuscripts for the LOS ANGELES COUNTY MUSEUM, CONTRIBU-
TIONS 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. Al-
though priority will be given to manuscripts by staff members, or to papers dealing
largely with specimens in the Museum’s collections, 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 Edi-
torial Committee. They may also be subject to review by competent specialists out-
side 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.

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. Alterations or changes in the manuscript after galley proof will
be billed to the author. Unless specifically requested, page proof will not 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.

Dorothy M. Halmos
Editor

THE BIRDS OF THE SANGO BAY FORESTS,
BUDDU COUNTY, MASAKA DISTRICT, UGANDA

By Herbert Friedmann1 and John G. Williams2

Abstract: The bird life of the forests of Sango Bay, Buddu
County, Masaka District, Uganda, is here described for the first
time. Hitherto there were a few haphazard specimen records for a
small part of the avifauna of Buddu widely scattered in the pub-
lished data, and almost none of these came from the Sango Bay
area. The collection reported herein, made in 1968 in the Mala-
bigambo and Namalala Forests, comprises 760 examples of 138
species of birds. While no new forms have been found among
them, there are indications of incipient micro-evolutionary varia-
tional trends in a few species. The collection reveals numerous
extensions of previously known ranges of some birds. The most
important, and the greatest, in a geographic sense, of these refer
to such little known birds as Alcedo leucogaster leopoldi, Phyl-
lastrephus icterinus tricolor, Malimbus malimbicus malimbicus,
and N esocharis ansorgei ansorgei.

The fact that in the wet season these forests are actually
flooded has resulted in the terrestrial nesting sylvan birds accom-
modating their breeding time to the early part of the dry season,
and also in the development of incipient arboreal habits in even
so ground haunting a species as the little rail, Sarothrura pulchra
centralis.

Introduction

Exploration of the part of southern Uganda immediately to the west of
Lake Victoria may be said to have begun with Speke’s famous expedition of
1862, when he first traversed eastern Ankole, a short distance to the west of
the Sango Bay area, the region of our present interest. In 1875 Stanley
circumnavigated Lake Victoria and undoubtedly passed close to and saw the
forests where they reached the shores of Sango Bay, but he did not collect or
report any birds on that trip. Captain Speke did make a small bird collection,
published on in 1864 by P. L, Sclater, but none of his specimens came from
the area of our present concern. Similarly, no birds were obtained by Villiers,
who made a diagonal crossing from southeast to northwest of southern Buddu
and eastern Ankole in 1893. The first ornithological collector to visit the
region was apparently Franz Stuhlmann, who, with Emin Pascha, traversed
Karagwe, immediately to the south of the Sango Bay area, and crossed the
Kagera River, but they did little collecting on that stage of their joint expedi-
tion. Stuhlmann obtained only nine species of birds in this area, each repre-
sented by a single specimen, and none of them were forest dwellers. Stuhl-

1Director, Los Angeles County Museum of Natural History.

2Co-investigator, Uganda Forest Survey.

1

7

Contributions in Science

No. 162

of forests.

mann’s entire collection, made on a long trip of which the Buddu section was
only a small part, was published on by Reichenow (1892, 1893).

Knowledge of the birds of that portion of Uganda on the west shore of
Lake Victoria from Masaka south to the Tanzanian border, the region called
Buddu, was not added to until the work of Walter G. Doggett, naturalist at-
tached by Sir Harry Johnston to the Anglo-German Boundary Commission
under Lt. Col. C. Delme-Radcliffe. Doggett, whose work was prematurely
terminated by his death from drowning when his canoe capsized in the Kagera
River, collected in Buddu, Msozi, Mulema, and Ketoma, all relatively near the
Sango Bay area, from December, 1902, to May, 1903. His collections were
reported on by Ogilvie-Grant (1905), and were later incorporated in Reich-
enow’s (1911) account of the birds of the lake region of central Africa. This
paper was based largely on the material collected by the German Central
African Expedition of 1907-1908 under Archduke Adolf Friedrich of Meck-
lenburg, but included all previous records as well. This expedition entered the
Minziro Forest, in Tanzania just south of the Uganda border, but collected
very little there. The Minziro Forest is the southern continuation of the

1969

Birds of Sango Bay Forests in Uganda

3

Malabigambo Forest, north of the Tanzania-Uganda border, and it is in the
Malabigambo that the collection was made that has occasioned this present
paper. In Reichenow’s 1911 survey paper, the small number of species re-
corded from “Buddu” or from “Buddu Wald” reveals how slight Doggett’s
sampling of the local avifauna was. Since his pioneer efforts, only sporadic and
minor additions have been made to the published information, based pri-
marily on specimens obtained by, or for, F. J. Jackson; Neave, who collected
a few birds in the Tero Forest; Goudey, who also collected in that same forest;
and later V. G. L. van Someren. These records are merely listed as from Buddu,
Kagera, and Tero Forest by Jackson (1938) and by van Someren (1922,
1932). They involve some 68 kinds of birds, but only seven are forest species
( Mesopicos xantholophus, Bleda syndactyla, Bleda eximia, Stizorhina fraseri,
Platysteira castanea, Trochocercus nigromitratus, and Terpsiphone rufiventer) ,
all of which are represented in our present collection.

The lack of ornithological data from Buddu prior to Doggett’s work may
be seen from the blanks on Reichenow’s map (1902, Atlas, map B), revealing
no localities from which specimens had been reported for all of Buddu from
the Katonga River in the north, south to the Kagera River, on the Tanzanian
border, although “Buddu” is listed in the gazeteer as a locality from which
both Neumann and Stuhlmann had collected birds. Reichenow’s atlas volume
was published before the report on Doggett’s collection was issued, and his
localities are therefore not mentioned.

Stuhlmann we have already discussed. Neumann (1899, pp. 68, 69)
recorded examples of two species of plantain-eaters, Musophaga rossae and
Chizaerhis ( —Crinifer ) zonurus , from the Buddu coast, and further cited, as
evidence of a west African element in the avifauna there, the presence of such
birds as Corytheola cristata, Agapornis pullaria ( ugandae ), Campephaga
phoenicea (probably Campephaga quiscalina martini ), Cinnyris superbus
( = Nectarinia superba buvuma), Cyanomitra viridisplendens (= Nectarinia
verticalis viridisplendens) , and Cyanomitra cyanolaema {—Nectarinia cyano-
laema octaviae) . All of the species listed by Neumann, except the last one, are
represented in the present Malabigambo collection, and this species was not
actually obtained by Neumann in Buddu. His only specimen was taken at Kwa
Mtessa, Ssingo Province, northern Uganda. It is not clear, therefore, why he
included it in his list of west African species found in Buddu.

Doggett collected 14 species in Buddu near the western shore of Lake
Victoria. All of them were non-forest birds, a gull, two plovers, a sandpiper,
three herons, a stork, an ibis, a pygmy goose, whiteheaded eagle, osprey, and
two kingfishers. At Msozi, also near the lake shore, he obtained 19 species,
again non-forest birds, and at the mouth of the Kagera River he collected a
single species, the whitefaced tree duck. His Mulema collection was more
extensive but less immediately pertinent to our present study, as that locality
is a considerable distance farther to the west and is about 1200 feet higher in

4

Contributions in Science

No. 162

altitude than the Sango Bay forests. Most of the Mulema birds were non-sylvan
species. At Ketoma he collected only two birds, a pipit and a parrot.

The great majority of the records listed by Jackson (1938) and by van
Someren (1922; 1932) are reported merely as Buddu, and probably did not
come from Sango Bay. Apparently no important or extensive collections have
been made in that exact area, but we do know that Captain C. R. S. Pitman
spent a little time there in August, 1929, whence he recorded the yellow-crested
woodpecker and the Congo forest akalat (1931a,b), both of which are included
in our 1968 Malabigambo collection.

Captain Pitman (in litt., July 1968) informs me that he also collected the
following four species there — Circaetus cinerascens Muller, Halcyon malim-
bica malimbica (Shaw), Pycnonotus virens holochlorus (van Someren),
Nectarinia chloropygia orphogaster (Reichenow).

The first of these is a bird of the open country, not of the forest, and was,
in fact, collected outside of, but close to, the edge of the Malabigambo. The
other three are forest birds and all are represented in our present collections.
To judge from his letter, Captain Pitman considered the Sango Bay forests to
show some ecological and faunal affinity with the forest regions east of Lake
Kivu.

While odd records of Buddu bird specimens may well be found scattered
through the literature or hidden unpublished in museum collections, the chief
compilations hitherto have been those of Reichenow (1911), van Someren
(1922; 1932) and of Jackson (1938). These combined lists include Buddu
records for 68 species of birds, 48 of which are not represented in our present
collection, which contains 138 species, making a combined total of 186 kinds
of birds from Buddu County. The 49 species recorded many years ago from
Buddu but not represented in our Sango Bay material are almost all birds of
open bushveldt, not of true forest, and this fact makes it appear that the present
collection is the first sizeable one made in the interior of the Sango Bay forests.
Even in the present collection there are some birds that were obtained just
outside of the forest or in open glades within it, and are not typically sylvan
species. Aside from these few species, the collection does give us our first fairly
comprehensive indication of the avifauna of the wet woodlands of the area,
and in particular of the Malabigambo Forest, the largest of the Sango Bay
forests.

The 48 species of birds from these Buddu lists not represented in the
present collection are:

Nycticorax nycticorax nycticorax (Linnaeus)

Butorides striatus atricapillus ( Afzelius)

Butorides rufiventris (Sundevall)

Scopus umbretta umbretta Gmelin
Anastomus lamelligerus lamelligerus Temminck

1969

Birds of Sango Bay Forests in Uganda

5

Dendrocygna bicolor (Vieillot)

Dendrocygna viduata (Linnaeus)

Nettapus auritus (Boddaert)

Anas sparsa leucostigma Ruppell
Terathopius ecaudatus (Daudin)

Aquila wahlbergi Sundevall
Pandion haliaetus haliaeius (Linnaeus)
Francolinus sephaena grantii Hartlaub
Burhinus vermiculatus vermiculatus (Cabanis)
Vanellus spinosus (Linnaeus)

Vanellus senegallus lateralis Smith
Charadrius pecuarius pecuarius Temminck
Charadrius tricollaris tricollaris Vieillot
Tringa glareola Linnaeus
Calidris alba (Pallas)

Glareola nuchalis nuchalis Gray
Streptopelia senegalensis senegalensis (Linnaeus)
Treron calva salvadorii Hartert and Goodson
Corythaixoides personata leopoldi (Shelley)
Chrysococcyx caprius (Boddaert)

Bubo africanus cinerascens Guerin
Ceryle maxima (Pallas)

A pus niansae (Reichenow)

Eurystomus glaucurus afer (Latham)

Tockus nasutus nasutus (Linnaeus)

Macronyx croceus (Vieillot)

Turdoides jardinei emini (Neumann)

Parisoma plumb eum plumbeum (Hartlaub)
Saxicola torquata axillaris (Shelley)

Cossypha heuglini heuglini Hartlaub
Cossypha niveicapilla melanonota (Cabanis)
Erythropygia hartlaubi Reichenow
Muscicapa minima pumila (Reichenow)

Cisticola galactotes nyansae Neumann
Parus niger purpurascens van Someren
Dicrurus adsimilis divaricatus (Lichtenstein)
Nectarinia venusta igneiventris Reichenow
Nectarinia mariquensis suahelica (Reichenow)
Lamprotornis chloropterus elisabeth Stresemann
Passer griseus griseus (Vieillot)

Ploceus baglafecht stuhlmanni (Reichenow)
Ploceus intermedius intermedius Ruppell
Vidua hypocherina Verreaux

6

Contributions in Science

No. 162

The Los Angeles County Museum Survey of 1968

As may be deduced from the foregoing comments, the extensive forests
of the Sango Bay area on the western side of Lake Victoria, have been very
slightly studied faunally in the past. In 1968, as part of a zoological survey of
the isolated forests of western Uganda by the Los Angeles County Museum of
Natural History, sponsored by National Science Foundation Grant GB-5107,
it was deemed advisable to send a collecting party to the Sango Bay area even
though it is considerably to the east of the forests of extreme western Uganda
that were our primary concern.

The field party consisted of A. L. Archer, who devoted his attention
largely to small mammals, but also did some of the bird collecting, both by
gun and by use of mist nets, and made examinations and reports of all bird
stomachs; R. Glen, who took charge of all herpetological collecting, but did
much of the netting and shooting of birds and wrapped all the bird specimens;
and Mrs. E. Archer, who did all the labelling of birds and mammals, and took
charge of the butterfly and moth collecting. The party also included eleven
trained African skinners, trappers, driver and camp cook, and employed a
small number of local laborers for camp chores and for cutting lines for nets
through the bush. The visit to the Sango Bay forests began on January 25 and
terminated on February 19, 1968. During this period very sizeable collections
were made, including 760 birds of 138 species, 256 mammals of over 40
species, 484 reptiles and amphibians, 470 butterflies, and over 1800 moths.
The present paper deals only with the birds collected, but all of these collec-
tions now are in the Los Angeles County Museum of Natural History.

Almost all of the bird collecting was done in the Malabigambo Forest,
about ten miles west of Lake Victoria, just north of the Tanzanian border, at
an altitude of 3800 feet. The Malabigambo, comprising about 25,600 acres, is
the largest of the four forests of the Sango Bay area, the others being the
Namalala, Kaiso, and Tero Forests. Because of its large size, collecting was
limited to the Malabigambo until the last few days of the stay, when lack of
additional species in the daily “take” made it seem that the local fauna had
been fairly well covered. This left time for a little work in only one of the
other forests, the Namalala, but there is little reason to think that these forest
faunas differ appreciably among themselves. The Namalala Forest extends
almost to the shore line of Lake Victoria, while Tero is virtually a pair of small
forest islands surrounded by marsh and year-round swamp. Kaiso is described
as being very similar to Malabigambo and is subject to similar drainage, and
in neither does a waterlogged swamp forest condition exist for much of the
year. All four forests were being exploited for timber, especially trees of a
species of Podocarpus, and the regeneration of these sylvan elements has been
negligible. The forests were leased until 1966 to a private lumbering company,
during which time there was little control of their cutting activities. These
woodlands were the first in Uganda to be exploited for sawmilling, and it is

1969

Birds of Sango Bay Forests in Uganda

7

fortunate that enough of their sylvan character has been preserved to enable
our party to make a representative, if somewhat belated, survey of their
terrestrial vertebrates.

The following basic data on these forests is taken from a mimeographed
report of April, 1955, by the staff of the Chief Conservator of Forests of the
Uganda Forest Department. Information on the botanical aspects of these
forests in their original condition are fortunately available in a report by Dawe
(1906) and in an unpublished one by Dawe and Stevenson (1908), entitled
“Report on a visit to the forests between the Kagera and Bokova Rivers.”
Contemporaneous with these is an account of the Minziro Forest of the
immediately adjacent portion of Tanzania, by Holtz (1906-1911), while an
unpublished report listing the trees, shrubs, herbs, and grasses of the Sango
Bay area by Philip (1953), entitled “Report on a joint safari with A.C.F.
Lake Province, Tanganyika to the Minziro Forest and Sango Bay Estates”; is
on file in the Entebbe office of the Uganda Forest Department. All these
unpublished accounts are briefly summarized in the 1955 mimeographed
report mentioned above.

The Namalala and Tero Forests are separate blocks situated along the
Bukora and Nakafunja Rivers, respectively, and are about two miles from
Sango Bay. The Kaiso and Malabigambo Forests are, in effect, the northern
arms of a larger block of woodlands, the Minziro Forest, most of which is in
Tanzania, south of the Uganda border. The Namalala has an estimated area of
5760 acres, Tero has 1907, and Kaiso has 4020, all three together comprising
less than half the acreage of Malabigambo. As all four of these forests lack any
major drainage channels, during the wet season the whole area is covered by
water and even during the dry season some of it is waterlogged a few feet from
the surface. The Namalala, which lies near the shore of Lake Victoria, is said
to be better drained than the rest, but even so the 1955 report describes it as
having the characters of “hog wallows,” some of which join up to form wind-
ing ribbonlike depressions over twenty meters long.

It appears that area represents an old flood plain of the Kagera River
formed soon after the rise of Lake Victoria to its present level, and it is
assumed that the present course of the Kagera in this area is a relatively recent
development and that the plain on either side is a submerged, and later raised,
plain of the Kagera or its tributaries, among which are the Bukora and
Nakafunja Rivers. The larger streams are papyrus -choked and their rate of
flow is slow, except after heavy rains, when a temporary head of water allows
for a relatively steep gradient across the plain to Lake Victoria.

The Sango Bay forests are, thus, moister, more apt to be covered with
ground water at times, than any of the forests of extreme western Uganda;
but strangely enough, the bird collections herein reported do not reveal any of
the usual effects of increased humidity, such as greater melanism or more
intense pigmentation. Two explanations present themselves — either the forests

8

Contributions in Science

No. 162

are relatively too young to have given time for such trends to begin to show
themselves in their avian inhabitants, or the populations of the resident birds
are sporadically added to by colonizers of their own kind. That the latter is
more probable than the former is suggested by the following statement in the
1955 forestry report, already referred to “. . . The sharp edges of the forest
indicates that it is not spreading but neither are there signs that it is decreasing
in size ... It is known that the climate of Uganda and particularly the Lake
shore has varied considerably during Historical and Recent Geological times.
It has been suggested that the forest at present is out of sympathy with its site
conditions and, if not changing in make up, is a ‘fossil’ forest of the past. To
support this theory, it is considered that the climatic conditions of the past
may have been considerably drier (this is certainly established) and the whole
site better drained with a lower lake level . . .” This is made even more explicit
in Dale’s (1954) report, in which he attempted to correlate climatic changes
with the increase and decrease of the extent of forested areas in Uganda. He
stated that after the “Nakuru” wet phase, about 850 B.C., eastern Africa
became drier. The fourteenth, fifteenth, and sixteenth centuries were wetter
again, but drier conditions prevailed from then to the middle of the nineteenth
century, since when the climate has once more become wetter.

In the following annotated account of the specimens collected, many
species are reported as being found chiefly, if not wholly, in the undergrowth
of the seasonal, then dry, swamp forests. This raises the question as to where
these birds go in the wet season when the forest floor is apt to be uncom-
fortably wet, and, at times, under standing water. The collectors actually
provide a suggestion of an answer in their field notes on one of the little sylvan
rails, Sarothrura pulchra centralis, when they comment that even so terrestrial
a bird as this has learned of necessity to be partly arboreal to an extent not
recorded, or possibly not necessary, in some other parts of its range. It would
seem that those species of birds that nest on, or very close to, the ground might
be forced to breed during the dry season. However, this is not borne out with
any definiteness by the gonadal condition of the specimens herein reported.
Certainly the birds of these species taken during the dry season in Malabigambo
are by no means all, or even largely, individuals with enlarged reproductive
organs. For example, of the rail mentioned above, not one of the 6 specimens
obtained had gonads that were more than slightly enlarged. Other terrestrial
nesters in the present collection that showed little or no gonadal swelling in
January and February are Francolinus squamatus , Guttera edouardi, Saro-
thrura elegans, Alcedo leucogaster, Ceyx picta, Halcyon malimbica, Aletha
diademata, and Alethe poliocephala. However, our series of Erithacus erythro-
thorax revealed gonadal enlargement in six out of a total of 11 individuals
collected. Other species that showed a similar range of gonadal condition are
Trichastoma rufipenne, Trichastoma fulvescens, and Erithacus cyornithopsis.
It may be that the two Trichastomas nest high enough off the ground, although

1969

Birds of Sango Bay Forests in Uganda

9

still in the low undergrowth, to be able to accommodate to a degree of fairly
shallow inundation of the forest floor.

We learn from the latest compilation of the rainfall statistics of the Sango
Bay area shown in the Atlas of Uganda, issued in 1962 by the Department of
Lands and Surveys of Uganda, that the average precipitation is 3 inches a
month in January and February, the time our collectors were there, and rises to
6 inches in March, to 10 inches in April and May, and then subsides to 2 inches
in June, July, August, and September, climbing to 3 inches in October and to
4 inches in November and December. This implies that the time when the
forest floor is critically flooded is limited to the period from late March to mid
June, which gives the ground haunting birds of the forest a good part of the
year free from excessive ground water. It also suggests that the season when
our present collection was made was toward the end of the dry period, and
this, in turn, suggests that while some of our specimens still had enlarged
gonads most of them were already through breeding. The fact that few of them
are in abraded plumage also bears out the thought that they are probably post-
breeding rather than pre-breeding birds, as most birds molt soon after nesting
and then are in fresh plumage. This, in turn, indicates that they must have done
their breeding in the early part of the dry season.

Acknowledgments

The material presented and discussed in this paper was collected for the
Los Angeles County Museum of Natural History, under a grant GB-5107 from
the National Science Foundation. For the loan of specimens needed for com-
parison in a few species thanks are hereby expressed to the American Museum
of Natural History and the Field Museum, and to their responsible curators.
Captain C. R. S. Pitman, formerly Game Warden of Uganda, now retired,
kindly wrote us about a few birds he collected in the Sango Bay region in 1929.

Annotated List of Specimens

Except where otherwise noted, all specimens were collected in, or at the
edge of, the Malabigambo Forest, 3800 feet above sea-level, in January and
February, 1968. The nomenclature used in this paper follows that of White’s
check-list, except where otherwise explained, even though it does seem that in
some cases he has “lumped” races that may prove to be distinct. The order of
the passerine families follows that of Mackworth-Praed and Grant (1955,
1957) ; the sequence of the species within each family is that of White (1960-
1965) except in the case of groups more recently covered in the volumes of
the continuation of Peters’s Check-list. All the specimens are in the collection
of the Los Angeles County Museum of Natural History, referred to in a few
instances as LACM.

10

Contributions in Science

No. 162

THRESKIORNITHIDAE
Bostrychia hagedash brevirostris (Reichenow)

One adult male, testes slightly enlarged, February 17; bill length (chord)
174.6 mm; weight 1262 grams; maxilla with basal half of culmen dark crim-
son; iris white; feet black, with scales on toes and part way up the tarsus pink.
The specimen was collected in a native cultivation area at the edge of the
forest, where it was feeding on the ground. It had a number of beetles and
caterpillars in its stomach.

The long bill of this specimen agrees with White’s figures (1965, p. 30)
for Ethiopian and Sudanese birds, 152-180 mm, as compared with 122-163 for
coastal east African examples, and 130-165 for west African ones. If the race
nilotica where to be recognized, our present specimen would fit with that
subspecies in its bill length, as well as in its geographical source.

ACCIPITRIDAE
Polyboroides radiatus typus Smith

One adult male, in non-breeding state, was collected in the seasonal
swamp forest, as it landed on a low tree at the edge of a path on January 26.
The bird is in worn plumage, weighed 652 grams; bill black, with the soft parts
white; iris dark brown; bare skin around the eye yellow; feet yellow. The bird
had a partly digested nestling of an unidentified bird in its stomach.

Lophaetus occipitalis (Daudin)

One adult male, testes not enlarged, was taken on February 1 1, in a forest
glade in dry seasonal swamp forest, where it was seen perching on the top of a
dead tree. It had the remains of an unidentified rodent in its stomach. Maxilla
pale blue gray, with black tip; mandible at base, and cere grayish yellow; gape
yellow; iris rich yellow; feet pale yellow; weight, 931 grams.

PHASIANIDAE

Francolinus squamatus schuetti Cabanis

Three scaly francolins were obtained: an adult male, testes not enlarged,
January 29; adult male, testes slightly enlarged, February 5; adult female,
ovary slightly enlarged, February 7. The first one was taken along a vehicle
track in dry seasonal swamp forest; the other two in dense undergrowth in the
same habitat. The two males weighed 519 and 553 grams, respectively, the
female 398 grams. Notes on the soft parts: bill deep orange to reddish orange
in the males, bright red in the female; iris dark brown to dark gray brown; feet
orange to reddish orange; stomach contents, grit and small seeds.

We follow White (1965, p. 89 )in considering zappeyi as synonymous
with schuetti , although our present birds are slightly darker in the median
areas of the feathers of the breast and upper back than are others from extreme

1969

Birds of Sango Bay Forests in Uganda

11

western Uganda, the westernmost examples available for comparison. How-
ever, Hall (1963, p. 137) found that all the discontinuous, scattered popula-
tions of this species east of the lower Guinea forest are sufficiently alike to be
united in the race schuetti. Within this large area, from the Sudan and Ethiopia,
south through Uganda and Kenya to western Tanzania, there is considerable
local and individual variation, but no consistent variational trends that may be
correlated with geography, in spite of the diverse types of habitats in which
discrete populations are found.

Guttera edouardi sethsmithi Neumann
An adult female, ovary slightly enlarged, was taken on February 8, in a
snare set by a mammal trapper in currently dry seasonal swamp forest. It
agrees in coloration with other examples from western Uganda (Budongo
Forest). It weighed 963 grams; stomach contents, small fruit seeds and grit.
Soft parts: basal portion of maxilla pale greenish gray, rest of maxilla more
grayish and paler, becoming dirty white at the tip; mandible bluish gray shad-
ing to pale gray terminally; iris dark brown; feet and tarsi dark brown, tinged
with gray; bare skin around eye blackish, shading to deep violet blue on the
occiput and nape; chin and throat red.

RALLIDAE

Sarothrura elegans (Smith)

Two examples of the buff-spotted rail were netted in dense undergrowth
in seasonal swamp forest, February 5 and 12 respectively. Both were males
with slight testicular enlargement. The February 12th bird has the buff spots
on the upperparts slightly paler and less rufous than the other. The latter
agrees with others from western Uganda (Bugoma and Bwamba). The soft
parts were noted as follows: maxilla dark horn gray to black; mandible with
pink to pinkish white edges and with dark gray tips; iris dark brown; feet dull
grayish brown; weight, 40 and 43 grams; stomach contents, insect fragments
and grit.

Sarothrura pulchra centralis Neumann
Six of these little rails were obtained, January 27 to February 6, two adult
males, three adult females (one of which was wrongly labeled as a male, but
corrected in the field data slip) , one subadult female. The last specimen has the
chestnut breast obscurely banded with fuscous posteriorly and the abdominal
bars slightly less contrasting (the black bars more fuscous) than the adults.
These specimens agree with a series from western Uganda (Kibale, Budongo,
Bugoma, Bwamba). All were in non-breeding condition; all had insects and
small seeds in their stomachs, and weighed 33, 44, 46 and 48 grams (females),
46, and 49 grams (males) ; bills dark horn gray to black; feet dull gray brown;

12

Contributions in Science

No. 162

iris dark brown in males, reddish brown to light brown in females.

These rails get up from the forest floor more often than has been sus-
pected hitherto. Thus, one was collected about four feet off the ground as it
was clambering about on a horizontal branch, waving its outstretched wings,
apparently as a balancing device. Another was eight feet up in a bush when
collected. It may be that these usually terrestrial birds have learned to climb
up into the bushes and trees during the annual wet season when these swamp
forests are flooded, and that this behavior carries over into the dry season.
In the Budongo Forest Williams also found these rails 6 to 8 feet off the
ground. A low chuck-chuck note was heard from two of the birds collected.
One specimen, possibly a sick bird, was picked up by hand along a path in
the forest.

COLUMBIDAE

Columba unicincta Cassin

This pigeon is fairly common in the Malabigambo Forest. Three adult
males were collected; one with enlarged testes, January 30, “appeared to be
building a nest” according to the collectors; the others with slightly enlarged
gonads, February 5; bill pale gray, with the soft parts darker and more
bluish; iris red; the bare skin around the eye dull red; feet pale gray; stomach
contents, orange fruit seeds and large grit; weight, 390, 440 and 490 grams.
The birds were found well up in the tall trees in the seasonal (dry) swamp
forest.

Through the kindness of the authorities of the American Museum of
Natural History and of the Field Museum, it has been possible to assemble
a long series (24 specimens) of the Afep pigeon, including two practically
topotypical birds from Gabon, five from the Congo, and 17 from western and
central Uganda. These showed a general agreement in coloration and in
dimensions. The study was initiated because smaller material at first suggested
that birds from extreme western Uganda seemed to have the white area on the
lower middle abdomen more extensive than did the present Sango Bay ex-
amples. It was found that this apparent difference was partly due to the prep-
aration of the specimens and was not correlated with geography. No signifi-
cant consistent difference between Gabon birds and a good number of eastern
specimens (Sezibwa River and Mabira Forest, Uganda) could be demon-
strated. The Sango Bay specimens are less extensively white on the posterior
underparts than examples from Bugoma, Budongo, Mabira, and Sezibwa, but
are matched by others from still farther west, from Beni and Niangara in the
Congo.

Streptopelia semitorquata (Ruppell)

One adult male, testes enlarged, was taken at Katera, at the edge of the

1969

Birds of Sango Bay Forests in Uganda

13

Malabigambo Forest, on February 13; bill black; iris brick red brown; tarsi
and toes maroon; stomach contents, broken seeds and grit; weight, 213 grams.

Turtur tympanistria (Temminck)

Three specimens of the tambourine dove were collected: an adult female,
January 29; adult female, February 6; adult male, February 8. All were found
in the seasonal (then dry) swamp forest; all had seeds, fibrous matter (seed
husks?) and grit in their stomachs; one female had a small ovary, the other
two birds showed slight gonadal enlargement. This is a widely distributed bird
of the wooded areas of much of tropical Africa. The present examples weighed
60, 65, and 67 grams, respectively, the two extremes being the females.

Treron australis gibberifrons (Madarasz)

Two adult males were taken on January 29, one with enlarged testes, the
other with only slight testicular swelling. Both had been feeding on wild figs
and other small fruits; cere orange red; apical third of bill bluish white; iris
aquamarine; feet bright orange red; weight, 230, 233 grams. We follow White
in considering granviki a synonym of gibberifrons.

PSITTACIDAE

Agapornis pullaria ugandae Neumann

One adult male red-headed lovebird, testes slightly enlarged, was collected
in open undergrowth at Mugera, in the Malabigambo Forest on February 19,
on the last day in the field; maxilla pale opaque orange; mandible whitish gray;
iris dark brown; feet olivaceous gray; bare skin around eye dark gray; weight,
43 grams.

MUSOPHAGIDAE
Tauraco schutti emini (Reichenow)

On January 28, an adult male of this tauraco was collected in seasonal
swamp forest; testes slightly enlarged; maxilla black; basal third of mandible
red; iris soft brown; feet black; bare skin around eye red; stomach contained
fruit pulp and pits; weight, 259 grams. The bird came close to the collector,
apparently attracted by the noise of two brown-eared woodpeckers that were
being removed from the mist net. Our specimen agrees with others from
Bwamba and Kibale Forests, western Uganda.

Musophaga rossae Gould

This tauraco is represented in the collection by an adult male, testes
slightly enlarged, taken on February 13; maxilla chrome yellow, brownish
orange at the edge; edge of frontal shield orange; mandible dull orange, with
the tip yellow; iris dark brown; bare eye skin chrome yellow; feet black;

14

Contributions in Science

No. 162

stomach contents, whole small fruits; weight, 444 grams. The bird was one of
two (a pair?) in a low tree where they were observed uttering a raucous
warbling call accompanied by bobbing of their heads. This species had been
recorded earlier from the southern Buddu coastal area by Neumann (1899,

p. 68).

Crinifer zonurus (Ruppell)

One adult male, testes enlarged, was shot in a tree in a cultivated area near
the edge of the Namalala Forest, February 18; bill pale greenish yellow; iris
dark brown; feet blackish brown; stomach contents, whole large purple fruits;
weight, 501 grams. When wounded, the bird’s cries brought in three other
individuals of its kind, flying about within a few feet of the collector, scolding
the while. This specimen is in worn plumage, but agrees closely with another
from Bwamba valley, western Uganda. Neumann (1899, p. 68) recorded this
bird from southern Buddu coast many years ago.

Corytheola cristata (Vieillot)

An adult female, ovary slightly enlarged, was shot in a tall tree in dry
seasonal swamp forest, on February 16; bill chrome yellow; tips from nostril
orange red; iris dark red; feet black; stomach contents, whole fruits and
unidentifiable vegetable matter; weight, 992 grams. This species was found
roosting in the upper branches of the trees during the mid-day heat. Many of
the small feathers of the neck of our specimen still show the sheaths, indicating
very recent ecdysis.

CUCULIDAE

Cercococcyx mechowi Cabanis

The dusky long-tailed cuckoo reaches its eastern limits on the west and
north shores of Lake Victoria, the easternmost record being from a little to the
west of Jinja. The Sango Bay area would seem to be its southern limit on the
west shore of the lake. The species must be fairly numerous in the Mala-
bigambo Forest, as no less than three examples were obtained within eight
days, a male with slightly enlarged testes on February 6, a female with slight
ovarian swelling on February 11, and another female, in a similar state, on
February 14; maxilla dark horn brown to black; mandible yellow green-gray,
browner at the tip; iris dark brown; feet pale yellow to pale orange yellow;
stomach contents, insects, including hairy caterpillars, and, in one case, one
seed; weight, 60 (male), 50, 56 grams (females).

One of the females has the dark cross bars on the throat and breast much
finer than the other; both are less heavily barred than the male. These three
examples agree with a series from western Uganda: Kibale, Budongo, Bugoma,
and Bwamba.

1969

Birds of Sango Bay Forests in Uganda

15

The collectors noted that this forest cuckoo, which spends the day in
dense undergrowth and is wary in the extreme, in the late evening rises high
above the undergrowth and calls, often from an exposed perch, and appears
to lose some of its daytime shyness. The call note was described as “ quick
quick quick sometimes a melodious warble as well.

Chrysococcyx cupreus cupreus (Shaw)

One adult female emerald cuckoo, in non-breeding state, was taken from
a lichen covered tree on January 29; maxilla dull black; mandible pale blue
with dark gray patches; iris dark brown; feet blue gray; stomach contents,
insect remains; weight, 35 grams. The bird appeared to be feeding on insects
among the lichens on the bark of the tree.

Ceuthmochares aereus aereus (Vieillot)

Two males, testes slightly enlarged, were collected on January 29 and
February 12, respectively; bill deep yellow, with a black spot on culmen ridge
at base; iris dark red to deep red brown; bare skin around eye turquoise blue
with a greenish wash; feet black; stomach contents, beetles, a locust, and a
caterpillar; weight, 57, 58 grams.

Chapin (1939, pp. 204-205) considers Uganda birds australis, but these
two, as well as a good series of western Uganda specimens, are typical aereus,
with no ochraceous tinge on the underparts. White (1965, p. 188) considers
aereus the race of northeastern Uganda and australis that of the rest of that
country. All Uganda birds available for direct comparison are aereus, very
different from Tanzanian examples of australis.

One of our specimens was found skulking about in a dense mass of
creepers that covered a tree in the dry seasonal swamp forest.

CAPRIMULGIDAE
Caprimulgus fossii fossii Hartlaub

One male, with enlarged testes, was collected in a cultivated area at the
edge of the Malabigambo Forest on February 16; bill black, nostrils and gape
brown; iris blackish brown; eye ring buff; feet dull brown; stomach contents,
beetles; weight, 43 grams. This specimen comes from the area where nominate
fossii and apatelius are said to intergrade, but it agrees very closely with east
Kenyan fossii.

APODIDAE

Apus apus apus (Linnaeus)

One female, in non-breeding state, of this Palaearctic swift was shot out
of a mixed flock of swifts and swallows hawking over an open glade in the

16

Contributions in Science

No. 162

forest on the evening of February 11; bill blackish; iris dark brown; feet
medium brown; stomach contents, winged ants; weight, 40 grams.

COLIIDAE

Colius striatus kiwuensis Reichenow

One female, with enlarged ovary, was taken on February 12; basal half
of culmen whitish, shading to blackish toward the tip; mandible whitish, with
blackish basal area; feet reddish pink; stomach contents, entirely fruit remains;
weight 78, grams; plumage very abraded. The speckled mousebird was fre-
quently encountered along the periphery of the forest, but was never seen to
actually enter the forest itself.

TROGONIDAE

Apaloderma narina narina (Stephens)

An adult male, testes slightly enlarged, was collected on February 15;
maxilla pale green horn, the basal third orange yellow; mandible orange
yellow on basal half, the rest pale green; iris deep chocolate red; bare eye skin
pale blue green above the eye, green to yellow green below the eye and at base
of bill; feet flesh color; stomach contents, one large green grasshopper, other
insect remains; weight, 67 grams. This trogon is evidently not a plentiful bird
in the Malabigambo, as this was the first one seen by either collector, and the
species was very seldom heard calling.

ALCEDINIDAE

Alcedo leucogaster leopoldi (Dubois)

Two specimens, both males, one in non-breeding state, January 26, the
other with slight gonadal enlargement, February 4, extend the known range
of this kingfisher about 170 miles farther east than previous data indicated.
These two, and a series from the Bwamba Forest, extreme western Uganda,
have already been discussed (Friedmann, 1969) in detail. One of them was
netted in fairly open undergrowth in presently dry seasonal swamp forest in a
place where no permanent water was known for at least a mile. The stomach
of one bird contained a few insect remains; the other was empty; weight,
12.5 grams.

Ceyx picta picta (Boddaert)

Eight pigmy kingfishers, collected January 27 to February 18, attest to
to the frequency of this bird in the Malabigambo; all were males, one with
slight testicular enlargement, the others with resting gonads; maxilla varying
with age from dull horn black to orange red, blackish at base; mandible red;
iris dark brown; feet orange red to pinkish orange; stomach contents, grass-
hoppers and other insect remains; weight, 12, 12, 13.5, 14, 14, 15, 15, and 15.9

1969

Birds of Sango Bay Forests in Uganda

17

grams, respectively. The younger birds not only have dusky maxillae but also
have the feathers of the back lighter, more bluish, less violet than the older
ones.

Halcyon malimbica malimbica (Shaw)

A common kingfisher in the Malabigambo, represented by five specimens,
taken on January 31, February 3, 6, and 7; two males with no gonadal swell-
ing, one female with enlarged ovary and two with slightly swollen ovaries;
maxilla red, with black at edges and base; mandible blackish; iris dark brown;
feet orange to orange red, with brownish scales; stomach contents, beetles,
other insect remains, one large milliped; weight, 84 to 93 grams. This series
agrees with others from western Uganda.

Many years ago Captain Pitman {in lift., July 1968) found this king-
fisher to be common in the Malabigambo and collected two males and one
female, now in the British Museum. Also it may be noted that Goudey collected
this kingfisher in the nearby Tero Forest years earlier (Jackson, 1938, p. 571 ) .

MEROPIDAE
Merops albicollis Vieillot

The white-throated bee-eater is not a true forest bird, but occurs in the
open glades in the Malabigambo and, at least in the dry season, in the seasonal
swamp forest. Three examples were obtained, one of each sex with small
gonads, and one whose sex was unrecorded, on January 30, February 6 and
10, respectively; bill black; iris crimson to dark red; feet pale olivaceous yellow
to olivaceous brown; stomach contents, insect remains; weight, 23, 26, 27
grams.

These birds were seen perching on bare branches of the trees from which
they flew out and caught insects on the wing.

Merops pusillus meridionalis (Sharpe)

One specimen, a female in non-breeding state, was collected in an open
glade in the Malabigambo on February 9; bill black; iris red; feet dark gray
brown; stomach contents, insect remains; weight, 15 grams.

Merops variegatus loringi (Mearns)

Like the preceding species, this one was encountered in an open glade in
the forest, where one immature male, in non-breeding state, was taken on
February 13; bill black; iris reddish brown; feet dark gray; stomach contents,
one grasshopper and other insect remains; weight, 21.5 grams.

BUCEROTIDAE

Tockus alboterminatus geloensis (Neumann)

One example of the crowned hornbill, a female in non-breeding state, and

18

Contributions in Science

No. 162

very worn plumage, was collected while feeding on wild figs, on January 29, in
seasonal swamp forest; bill dark brownish orange, the base narrowly ivory
yellow; iris pale yellow orange; feet blackish; stomach contents, wild figs,
insect remains; weight, 249 grams.

Bycanistes subcylindricus subquadratus Cabanis
On February 3 a pair of these hornbills was seen feeding on wild figs in
the dry, seasonal swamp forest, and one of them was collected, an adult male
with slightly enlarged testes; bill dull brown, the basal half of the casque gray
white; iris dark brown; bare eye skin black; feet scaly black; stomach contents,
large fruit stones, large green beetle; weight, 1347 grams. The black-and-white
casqued hornbill seems to occur in almost every forest from Cameroon to
Uganda, and western Kenya.

CAPITONIDAE

Lybius hirsutus ansorgii (Shelley)

This barbet is represented by four specimens, three adult males, all in
non-breeding state, taken on February 5 and 17, one adult female with slightly
enlarged ovary, February 1; bill black; iris dark red; feet black; stomach
contents, whole greenish fruits and fruit husks; weight, 49 (female), 56, 58,
and 62 grams (males). The female has the breast and abdomen much richer
yellow-green than the males, and also has the yellow spots on the upperparts
more golden, those of the males being pale lemon yellow. One of the males
was observed repeatedly flying from its perch, picking one fruit or berry while
hovering, and then returning to the perch each time.

Buccanodon duchaillui duchaillui (Cassin)

The yellow spotted barbet is common in the Malabigambo, where seven
specimens were collected, three adult males, with no or slight testicular en-
largement, and four adult females varying in their ovarian state from small to
large, January 26, 27, 29, 30, 31, and February 11; bill black; iris dark brown;
feet dark grayish brown; stomach contents, wild figs, fruit seeds; weight,
(males) 38, 41, 43, (females) 30, 38, 39, and 45 grams. The collectors noted
that this barbet makes a call note similar to that of a Smithornis, a “ proop ”
sound.

Pogoniulus bilineatus mfumbiri Ogilvie-Grant
That this little tinker bird is extremely common in the Malabigambo is
shown by a series of 14 specimens, six males, eight females, with varying
degrees of gonadal enlargement, from none at all to very much, taken between
January 27 and February 17. This abundance undoubtedly helps to account
for the unexpectedly large population of its parasites, Indicator exilis, and

1969

Birds of Sango Bay Forests in Uganda

19

I. variegatus, in the area. This series agrees very closely with a still larger one
from extreme western Uganda, and fits with the race mfumbiri. The Buddu
area and, on the north shore of Lake Victoria, the Entebbe district, mark the
eastern limits of the range of this subspecies. The soft parts were recorded as:
bill black; iris dark brown; eye skin dark gray; feet dark gray to dark brownish
gray; weight, 12.5 to 14 grams in males, 1 1 to 14.5 grams in females; stomach
contents, small fruits, seeds, and insects. Many of the specimens were caught
in the mist nets along a path through the dry seasonal swamp forest. Some
came in answer to an imitation of their monotonous toop-toop-toop call. A
female, collected on February 10, was one of two immature birds seen chasing
about in the bush. It resembles the adults in plumage except for the presence
of faint greenish-yellow narrow tips to the feathers of the upper back.

INDICATORIDAE

Indicator variegatus Lesson

The scaly-throated honey-guide must be unusually numerous in the sea-
sonal swamp forest of the Malabigambo, possibly more so there than has yet
been recorded anywhere else in its very extensive range, if published data may
be taken as a reliable index. In the personal field experience of the senior
author, this honey-guide was nowhere a numerous species and it was a great
surprise to him to find that in 16 days between February 2 and 18 no fewer
than eight of these ordinarily non-abundant birds were collected in a small
area; five males, three with small gonads, one with some testicular swelling,
and one with large testes. During the same time three females, one with large,
and two with somewhat enlarged ovaries, were obtained. If, as seems most
probable, these honey-guides are not locally migratory, it follows that one
might expect that the species of barbets and woodpeckers, now known to occur
in good numbers in the Malabigambo, would be the most likely local hosts of
these brood parasites. Yet to date the little tinker-bird, Pogoniulus bilineatus,
is the only one of these potential fosterers for which there are any records of
such parasitism, and this one must be heavily parasitized by Indicator exilis,
probably even more often than by I. variegatus.

The soft parts were noted as follows: maxilla dull brownish gray horn
color, shading to pinkish basally; mandible similar but somewhat paler; iris
dark brown; feet with olive scales on dark gray skin; stomach contents, insect
fragments and bees’ wax; weight, 48 to 56 grams in the males, 52.5 to 60.5
grams in the females.

Indicator exilis pachyrhynchus (Heuglin)

A remarkable series of 23 specimens, 16 males and 7 females, of this
little honey-guide were obtained in the 26 days from January 27 to February
19. This is by far the largest number of individuals ever taken from one

20

Contributions in Science

No. 162

restricted area (the extreme points were less than two and a half miles apart)
in so short a time span, and is evidence of the unexpectedly high numerical
status of this species in the Malabigambo Forest. Inasmuch as these birds are
not migratory and are parasitic on small barbets, especially species of Pogo-
niulus, it is significant that the local population of the latter is so high as to be
able to stand the inroads of so numerous a brood parasite. The collectors, A. L.
Archer and R. Glen, are planning a separate report giving their interpretations
of their observations on the occurrence and movements of this honey-guide.

Of the 16 males collected two showed no testicular swelling, 13 showed
some, and one had large testes; of the seven females, four showed some
ovarian development and three had large ovaries; one of the latter, taken on
February 14, had one yolked egg 7 mm long in the ovary. The birds weighed
18 to 23 (average 21.4) grams (males); 16 to 19 (average 17.6) grams (fe-
males). Males, soft parts: maxilla black; mandible pinkish gray basally, black
terminally; iris dark brown; eye skin gray; feet with olive scales on dark gray
skin. Females, soft parts: maxilla black; mandible gray to blackish; iris dark
brown; eye skin dark gray; feet gray with an olive wash. The stomachs of these
individuals contained masses of beeswax, insect remains, and, in one case, ants.

Two of the females and one of the males have black streaks on the upper
throat, an occasional plumage variation in the eastern population of I. exilis.
The male is somewhat melanistic and has the forehead, lores, sides of the head,
the chin, and throat dusky, the color of the throat almost obscuring the
blackish streaks. In light of the tendency of the genus Indicator to produce
sympatric sibling species, this specimen is of more interest than an ordinary
slight color deviant would be in other birds. With only one such example, the
most plausible treatment of it is to consider it a partial melanism.

This is one instance where we do not follow White’s check-list. He con-
siders pachyrhynchus as the same as nominate exilis, but the eastern birds are
darker grayish, less olivaceous below, and on average larger than Cameroon
(topotypical exilis ), Gabon, and central Congo examples. Chapin (1962, p.
42) also found pachyrhynchus to be a recognizable race.

PICIDAE

Campethera cailliautii cailliautii (Malherbe)

The little spotted woodpecker is represented by one female, ovary slightly
enlarged, taken in the Namalala Forest, on February 13, where it was one of a
pair seen feeding low in the undergrowth of the dense, wet swamp forest;
mixilla blackish gray; mandible pale gray with dark tip; iris dark brown; bare
skin around eye olive; feet pale olive; stomach contents, ants; weight, 45 grams.

Campethera caroli (Malherbe)

Six examples, four males, two unsexed, were taken in the Malabigambo

1969

Birds of Sango Bay Forests in Uganda

21

Forest, January 28 to February 14, all in non-breeding state; maxilla dull
black; mandible the same; iris dark brown to maroon red; eye skin dark olive;
feet olive; stomach contents, largely ants but also some other insect fragments;
weight, 57 to 63 grams. These specimens, together with others now available
from western Uganda, make it seem inadvisable to separate budongoensis\ in
this we agree with White’s binomial treatment of the brown-eared woodpecker.

Campethera nivosa herberti (Alexander)

The buff-spotted woodpecker is a very common bird in the undergrowth
in the seasonal swamp forest of both the Malabigambo and Namalala Forests.
Ten specimens, seven from Malabigambo, January 27 to February 19, and
three from Namalala, February 13 to 19, were obtained, five females, three
males, and two birds whose sex was not recorded. All the birds had either no
or only slight gonadal development; maxilla dark horn gray to blackish gray;
mandible paler at base and along the edges; iris dark red to dark brown; feet
olive; stomach contents, mainly ants; weight, 30 to 36 grams in the males,
31 to 35 grams in the females.

Mesopicos xantholophus (Hargitt)

The yellow-crested woodpecker was first recorded from the Sango Bay
forests by Pitman in 1931 (1931b) and was found there again by our party,
who collected one adult male, with small gonads, on January 29; maxilla
blackish horn; mandible and tomia lighter in shade; iris dark brown; feet
brownish gray; stomach contents, insect larvae; weight, 67 grams. This speci-
men agrees in coloration and in dimensions with others from extreme western
Uganda (Budongo and Bugoma).

HIRUNDINIDAE

Hirundo rustica rustica Linnaeus

The European barn swallow has an extensive winter range in Africa, but
never goes into forested areas. Three examples were obtaind flying about in a
large open glade in the Malabigambo, on February 2, 7, and 11.

Hirundo angolensis Bocage

The Angola swallow was fairly common in clearings and open glades
in the Malabigambo; five specimens were obtained on February 6 and 7, two
males and three females. Both males had enlarged gonads, the females did not;
bill black; iris dark brown; feet dark brown; stomach contents, flying ants;
weight, 16 to 18 grams. One of the males and one of the females have more
white in the middle of the abdomen, as in the supposed Uganda race arcti-
cincta, but the others have the underparts more generally suffused with dusky

22

Contributions in Science

No. 162

earth brown. We therefore agree with White (1961, p. 54) in considering
arcticincta synonymous with angolensis.

Hirundo abyssinica unitatis Sclater and Praed
One adult male, in non-breeding state, was taken on February 1 1 ; maxilla
blackish; mandible slightly paler; iris dark brown; feet dark brown; stomach
contents, winged ants; weight, 17.5 grams. The bird was collected from a mixed
flock of swallows and swifts feeding in the air over an open glade in the forest.

Psalidoprocne albiceps albiceps Sclater
The white-headed swallow was found in open glades in both the Mala-
bigambo and the Namalala Forests. In the former, one adult male was taken
on February 7; in the latter, a young bird was collected on February 17; bills
black; iris dark brown; feet dark brown; stomach contents of adult, winged
ants; of young bird, empty; weight, 10.5 grams in young bird, 12.5 grams in
adult.

TIMALIIDAE

Trichastoma albipectus barakae (Jackson)

Very common in the undergrowth in the seasonal swamp forest of the
Malabigambo. A series of ten specimens was collected; six males, one with
enlarged testes, the others with small or only slightly enlarged gonads, three
females, all in non-breeding state, and one bird whose sex was unrecorded,
January 29 to February 11; maxilla black; mandible variable — gray, brownish
gray, to blackish with yellowish basal area; iris brown to dark brown; feet
silvery gray, sometimes with a pinkish flesh tinge; stomach contents, insect
remains and insect eggs; weight, females; 26, 29, 31.5 grams; males; 22.5,
28, 32, 33.5, 34, 34 grams.

We follow Deignan (1964, in Peter’s Check-list, vol. 10, pp. 240-427) in
including Malacocincla in Trichastoma. This accounts for the divergence of
our nomenclature from that of White’s list.

Trichastoma rufipenne rufipenne Sharpe
This is another very common bird in the same habitat as the preceding
species. The ecological sympatry of these very similar birds, and the next one,
can only be made possible by highly critical behavioral, perhaps auditory,
isolating mechanisms not apparent in dried museum specimens. At the same
time one cannot help but wonder whether these barriers to random crossing
may occasionally weaken, as there are specimens that are difficult to place in
one or another category and that raise the possibility of hybrids.

The abundance of this species in the Malabigambo is evidenced by the
fact that 20 specimens were obtained from January 27 to February 19, nine

1969

Birds of Sango Bay Forests in Uganda

23

of each sex, and two whose sex was undetermined. The females were all in
non-breeding state or with only slight ovarian enlargement; the males varied
from no gonadal swelling to enlarged testes; maxilla black; mandible gray;
iris light yellowish brown to dark brown to chestnut; feet blackish brown, the
toes grayer; stomach contents, insect fragments, caterpillar skin, small beetles;
weight of females ranged from 20 to 25 grams (one tagged as 9.5 grams,
probably an error for 19.5) and males between 22 to 29 grams (one tagged
as 46 grams, probably in error for 26).

Trichastoma fulvescens ugandae (van Someren)

The Uganda brown thrush-babbler is common in the Malabigambo, where
six specimens, three males, two females, one unsexed, were taken, January 30
to February 9. The gonads were enlarged in two of the males, slightly so in
the other male and in one female, not enlarged in the other. The unsexed
individual is probably an immature bird, as it has a smaller bill and has the
mandible, gape, and the tomial edge of the maxilla more yellowish, and is
shorter winged (64 mm) as well. The soft parts were recorded for these speci-
mens as follows: maxilla blackish horn, in one case with pale grayish white
edges; mandible pale gray, darker basally; iris medium brown to bright red
brown; feet dark gray, with or without a violet tinge, to dark blackish brown;
stomach contents, insect remains, one caterpillar skin; weight, 28 to 33.5 grams,
23 grams in the immature bird.

The five adults average very slightly paler, a little more grayish tawny,
less rufescent, on the underparts than a long series from Bwamba, but agree
with an equally extensive series from Budongo and Kibale Forests of western
Uganda in this respect. The Bwamba birds also average slightly more rufescent
on the back than the Budongo, Kibale, and Sango Bay birds. The last named
agree very closely with one from the Kakamega Forest in extreme western
Kenya. The Kakamega-Malabigambo specimens are geographically closer to
topotypical ugandae, described from the Sezibwa River, north of Entebbe,
which suggests that the Bwamba population and the Congo birds immediately
to the west of that area may show some intermediacy in coloration between
typical ugandae and nominate fulvescens, the race of Gabon and Cameroon,
but that the populations inhabiting the Kibale and Budongo Forests do not.

Turdoides melanops sharpei (Reichenow)

Four specimens taken out of a flock on February 12, two of each sex,
with little or no gonadal enlargement; bill black; iris white; feet blackish brown;
stomach contents, insect remains; weight, 79, 80, 86, 91 grams, respectively.
The birds were collected in open grassland near an isolated section of damp
forest.

24

Contributions in Science

No. 162

PYCNONOTIDAE

Pycnonotus barbatus tricolor (Hartlaub)

This ubiquitous bulbul was found in the then dry seasonal swamp forest,
especially the more open portions, when five examples were collected, Febru-
ary 1 to 12; one male with enlarged testes, and four females with small or only
slightly enlarged ovaries; bill black; iris dark brown; feet blackish; stomach
contents, bramble seeds and fruit skins (3 of the 5 had nothing in their
stomachs); weight of male 37, and females 37, 38, 38, and 41 grams. These
specimens are in worn plumage. They are tricolor, agreeing with others of that
race from Budongo and Kibale, and not with minor of Mt. Moroto.

Pycnonotus virens holochlorus (van Someren)

A common forest bird, represented by 1 1 specimens taken January 29 to
February 15, five adult males, some with large and some with small testes; one
immature male; four adult females, one with large, the others with small or
slightly swollen ovaries; one unsexed bird; maxilla horn blackish; mandible
lighter, more brownish; iris grayish brown to dark brown; feet pale yellowish
brown; stomach contents, fruit seeds and insect remains; weight, males 22 to
26 grams, females 23 to 27 grams. These specimens agree with long series
from the forests of western Uganda.

Pitman {in litt., July, 1968) found this bulbul to be numerous and noisy
during his visit to the Malabigambo some decades ago, and he collected two
females, now in the British Museum.

Pycnonotus gracilis ugandae (van Someren)

A common bulbul in the Malabigambo, this species is represented by nine
specimens taken January 30 to February 17; six males with gonads varying
from small to large; three females with some to much ovarian swelling; bill
black; iris brownish gray to dark brown; feet grayish olive; stomach contents,
small fruits and insect fragments; weight, 19 to 24 grams. The birds were
found chiefly in low fruiting trees in the seasonal (dry) swamp forest. These
specimens agree in coloration with a series from the forests of extreme
western Uganda.

Pycnonotus curvirostris curvirostris (Cassin)

This greenbul was found to be very common in the Malabigambo Forest,
and nine specimens were procured January 30 to February 11, six adult males
with enlarged testes, and three females with small or slightly swollen ovaries;
bill dark gray brown, with paler tomial edges; iris red brown; feet olive to
pale brownish; stomach contents, small fruits; weight, 22 to 36 grams (one
tagged as 53 grams, probably an error for 33 grams) .

1969

Birds of Sango Bay Forests in Uganda

25

Pycnonotus latirostris eugenius (Reichenow)

This is another very common bulbul in the Malabigambo Forest, where
15 specimens were taken January 27 to February 11; nine adult males, six
adult females, with gonads ranging from small to much enlarged; bill horn
black, tomial edges paler, to yellowish; iris brownish gray; feet yellowish
brown; stomach contents, fruit seeds, insect remains in six instances, the
remainder with empty stomachs; weight, 24 to 27 grams in females, 23 to 31
grams in the males. This fine series is nearly topotypical, the race eugenius
being described from Bukoba, a short distance to the south of Sango Bay.

Pycnonotus gracilirostris congensis (Reichenow)

Another common denizen of the forest, represented by eight examples
collected January 29 to February 12; five males and three females, all but
one of the males with large testes; none of the females with more than slightly
enlarged ovaries; bill black; iris dark red; feet black; stomach contents, insect
fragments and small fruits; weight, 27 to 35 grams. These examples agree with
long series from the forests of western Uganda. According to White (1962b,
p. 78), chagwensis is not separable from congensis, but Rand (1960, p. 256)
recognizes the former and assigns to it all the Uganda forest populations. If
the two are to be kept separate, all the present birds would be chagwensis, but
in light of Chapin’s findings (1953, p. 117) it seems doubtful that the Uganda
birds are sufficiently different from congensis, and sufficiently constant in the
difference, to warrant using the distinct name chagwensis for them. Rand
(1958, p. 181), on the other hand, concluded that the characters of chagwensis
are not simply those of intergradation between congensis and the Kenya race
percivali, and for this reason he recognized it. In the absence of topotypical
congensis for comparison, and in the light especially of Chapin’s conclusions,
we go along with White.

Ixonotus guttatus bugoma Rand

Eight examples of the spotted greenbul, taken January 27 to February 14,
reflect the numerical status of this bird in the Malabigambo Forest; seven
males and one female, all in non-breeding condition; maxilla dark dull gray
brown; mandible paler, gray to grayish horn; iris dark brown to deep red-
brown; feet gray to dark gray; stomach contents, wild figs and other small
fruits; weight, 31 to 38 grams. These specimens are in fairly worn plumage.
The birds were found in small parties feeding in the upper parts of tall trees
in the seasonal (dry) swamp forest, occasionally together with yellow-spotted
barbets, Buccanodon duchaillui.

Chlorocichla flavicollis pallidigula (Sharpe)

This bulbul is less numerous than the previous half dozen species. It was

26

Contributions in Science

No. 162

not found in the Malabigambo, but on February 18 one was taken in the
Namalala Forest — an adult male with testes much enlarged; bill black; iris
brownish cream; feet black; stomach contents, a few insect remains; weight,
51 grams. The absence of a specimen from Malabigambo can hardly be taken
to demonstrate that the species does not occur there as well as in Namalala.

Phyllastrephus hypochloris (Jackson)

This olive greenbul is a very common bird in the Malabigambo Forest, as
is evidenced by the collection there of ten examples within a fortnight, Febru-
ary 2 to 16; six adult males with gonads varying from small to enlarged, three
adult females in non-breeding state, and one bird whose sex was unrecorded;
maxilla dark brown horn to horn black, with yellowish tomial edges; mandible
paler, grayer, with the tip duskier; iris recorded in some specimens as gray, in
others as grayish brick red to medium red or pale chestnut; feet pale greenish
gray to pale bluish gray; stomach contents, largely insect fragments and some
small fruit seeds; weight, females 19 to 23 grams, males 20 to 30 grams.

These specimens agree in coloration and in size with others from western
Uganda: Bwamba, Kibale, and Kayonza (Impenetrable) Forests. This close
agreement justifies Rand’s (1960, p. 266) conclusion that kagerensis Reiche-
now, described from Buddu Forest, close to Sango Bay, is a synonym of
hypochloris. The collectors obtained most of the specimens in mist nets in the
undergrowth of the seasonal (dry) swamp forest. The birds go in small family
(?) parties, and when one is caught the others are apt to come in response to
its cries.

Phyllastrephus icterinus tricolor (Cassin)

The Malabigambo Forest is the most eastern point from which this
greenbul has yet been recorded. It must be abundant there as 12 specimens
were taken between January 28 and February 16; nine males, testes slightly
to much enlarged; three females with small ovaries; maxilla blackish horn to
dark horn brown, with tomium and tip yellowish; mandible paler, grayish with
yellowish tomial edges; iris yellowish gray to creamy brown; feet gray, nails
yellow; stomach contents, fine insect fragments, one caterpillar skin, few
beetles; weight, females, 16, 17, 21 grams, males, 18 to 21.5 grams. One of
the females, judging by its dimensions, may be wrongly sexed. The birds were
found chiefly in the undergrowth in the seasonal (dry) swamp forest.

This remarkable series of what has hitherto been a relatively rare bird in
collections, together with others in the LACM study material from Bwamba
and Budongo, has been reported on in detail (Friedmann 1968a). As pointed
out in that paper, icterinus and its larger, but extremely similar, congener
xavieri seem to occur sympatrically throughout many of the isolated forests
of Uganda, over a much greater area than was known to Chapin (1944) when
he first elucidated the distinction and relationship of the two sibling species.

1969

Birds of Sango Bay Forests in Uganda

27

Phyllastrephus xavieri xavier i (Oustalet)

This greenbul occurs commonly and ecologically sympatrically with the
smaller, but exceedingly similar P. icterinus in the Malabigambo Forest. Dur-
ing two weeks, January 30 to February 14, eight specimens of xavieri were
obtained, four of each sex, with gonads varying from small to much enlarged;
maxilla horn blackish, with whitish yellow tomial edges; mandible gray brown,
yellowish below; iris gray to pale yellowish gray; feet silvery gray to brownish
gray; stomach contents, insect fragments and one caterpillar skin; weight,
females 17 to 21 grams, males 23 to 26 grams. These birds were found chiefly
in the undergrowth of the seasonal (dry) swamp forest.

The Malabigambo Forest is the easternmost locality where both this
species and icterinus are known to occur together, but xavieri was already
known from other forests of about the same longitude, Butambala and Masindi.

Bleda syndactyla woosnami Ogilvie-Grant
This large bulbul occurs in all the forests of Uganda and beyond, to the
west, east, and south of that country. Its occurrence in Malabigambo is there-
fore not unexpected; eight specimens were collected, three males and five
females, with gonads ranging from small to large, January 30 to February 11;
maxilla dark horn to blackish; mandible paler, more grayish; iris dark red
(gray brown in one individual); feet pale gray with a pinkish or a yellowish
tinge; stomach contents, insects, including a green beetle, bramble (?) seeds,
one milliped; weight, males 40 to 47.5 grams, females 40 to 44 grams. One of
the males is immature and has both the maxilla and the mandible yellowish
terminally. The bristle-bill was found chiefly in the undergrowth in the sea-
sonal (dry) swamp forest.

Many years ago Neave (Jackson, 1938, p. 857) recorded this bulbul from
the nearby Tero Forest, Sango Bay.

Bleda eximia ugandae van Someren

This yellow-tailed bristle-bill occurs sympatrically with the preceding
species. While their habits and requirements are probably very similar, their
appearance is so markedly different caudally as to suggest that they carry their
isolating mechanisms posteriorly, and one would suspect that tail flashing and
spreading probably plays a major role in their courtship antics. The difference
in eye color, yellow in this species and red in syndactyla, probably also plays a
decisive part in preventing random crossings. In the fortnight from January 29
to February 13, four specimens were collected, two males, one female, one
bird of unrecorded sex, one of the males with enlarged testes, the other and
the female with only slight gonadal swelling; maxilla black to dark gray horn,
with pale horn to whitish cutting edges; mandible pale bluish gray; iris rich

28

Contributions in Science

No. 162

lemon yellow, darker at outer edge; feet gray; stomach contents, ants and other
insects; weight, 38 to 42.5 grams.

These four specimens agree with a series of Bwamba birds.

Jackson (1938, p. 858) noted a specimen recorded from the nearby Tero
Forest, taken by Neave.

Nicator chloris (Valenciennes)

This nicator occurs across Uganda in forested areas and its presence in
Malabigambo causes no surprise. It is apparently not abundant there as only
three examples were collected January 29 to February 14, two males and one
female, all in non-breeding state; bill black; iris dark brown; feet gray; stomach
contents, grasshoppers, one large caterpillar, and other insect remains; weight,
51 grams in each case. The collectors found these birds in the dense high under-
growth of the seasonal (dry) swamp forest; they noted a loud peeuck note
repeatedly given by one of these birds.

TURDIDAE

Erithacus cyornithopsis lopezi (Alexander)

The little Congo akalat is an abundant bird in the Malabigambo Forest,
where 17 specimens were collected January 27 to February 14, chiefly in the
dense undergrowth of the seasonal (dry) swamp forest. The collection of such
a series could not have been accomplished without the use of mist nets. Thus,
(Chapin (1953, 503), who found that in the Ituri Forest, in the Congo, it was
“unusually elusive in the extreme . . . Occasionally one may appear at the
edge of a forest road . . .,” was able to procure only three examples in as many
years. Of our 17 specimens, eight are males with gonads ranging from small to
large, and nine are females, all in non-breeding state; one male and two
females are juvenile, the others are adults; maxilla black to dark brownish
horn; mandible similar but yellow below and at base; iris dark brown; feet
pale gray, with or without a pinkish tinge; stomach contents, insect fragments,
one beetle; weight, males 16 to 20 grams, females 13 to 17 grams.

This species was reported from the Sango Bay forests long ago by
Pitman (1931a).

This akalat is a bird of the undergrowth in the relatively lowland forest,
whereas the related aequatorialis dwells in the highlands from 4500 to 6500
feet. The extensive collections now at LACM hold large series of aequatorialis
from the Kibale and Impenetrable Forests, and of the present bird from
Malabigambo. Strangely enough neither are represented in the large collections
from Bwamba, Budongo, or Bugoma. The two species are closely related and,
indeed, were combined as conspecific races by Chapin (cit. supra). We follow
Ripley (1964, p. 34) in keeping them as distinct species.

1969

Birds of Sango Bay Forests in Uganda

29

Erithacus erythrothorax mabirae (Jackson)

This little forest robin is another common bird of the Malabigambo, and,
indeed, of the forests from the eastern half of the Congo forest east to Buddu
and Mabira. Between January 28 and February 17, 11 specimens were col-
lected, eight males, six of which had enlarged testes and two did not, and three
females in non-breeding state, one of which is a young bird; bill black; iris
dark brown; feet flesh color to pinkish brown, tinged with gray; stomach con-
tents, ants and other insect fragments; weight, males 14 to 16 grams, females
14 to 15 grams. Most of the specimens were netted in the undergrowth of the
seasonal (dry) swamp forest.

This series has been compared with a long series from Bugoma, Budongo,
and Bwamba Forests and agrees fairly well with them. The extent of the
olivaceous tinge on the upper parts is variable; most, but not all, of the
Malabigambo birds have a little less of it than the Bugoma series, but the
Budongo and Bwamba birds are mostly like those from Malabigambo. The
yellowish wash on the abdomen in the Malabigambo and Bugoma specimens is
considerably more pronounced than in Budongo birds, but Bwamba specimens
appear to include both extremes. It seems better, therefore, to consider all the
Uganda birds as one race, although the degree of yellowishness or whiteness on
the abdomen is a geographically variable character. Chapin (1953, p. 507)
noted that typical erythrothorax has the lower breast and belly whitish in
Upper Guinea but that in Cameroon and middle Congo it has a yellowish tinge
on these parts.

Cossypha cyanocampter bartteloti Shelley

The blue-shouldered robin-chat is very numerous in the Malabigambo.
Between January 28 and February 18 the collectors obtained 13 specimens
there, eight males, two females, and three birds whose sex was not recorded;
two are juvenile, one in molt from juvenile to adult plumage; gonads ranged
from small to enlarged; bill black; iris dark brown; feet variable from gray with
a pink tinge, to dirty white with a brown tinge, to dark brown; stomach con-
tents, insect fragments, including one caterpillar; weight, males 28 to 31
grams, females 27 to 29 grams. These specimens agree with a long series from
western Uganda: Bwamba, Kibale, Budongo, Bugoma. Most of them were
collected in nets in the undergrowth of the seasonal (dry) swamp forest.

Alethe diademata woosnami Ogilvie-Grant
The fire-crested alethe occurs in the forests of Uganda, east at least as far
as Buddu and Mabira. In the Malabigambo the collectors found it to be
common and procured 11 specimens between January 28 and February 15.
These included two males and nine females, all in non-breeding condition;
bill black; iris dark brown to deep red brown; feet gray with a bluish or a

30

Contributions in Science

No. 162

pinkish tinge; stomach contents, insect fragments, one beetle; weight, males 29
to 33 grams, females 22 to 37 grams. The birds were found in the open as well
as the dense undergrowth of the seasonal (dry) swamp forest. The collectors
noted that in almost all cases where imperfect or unwanted specimens were
taken from the mist nets, they flew off about 10 feet and alighted on the ground,
where they apparently spend most of their time, not in the bushes or trees.

Alethe poliocephcda carruthersi Ogilvie-Grant
The brown-chested alethe is very common in the Malabigambo Forest,
where our collectors obtained 15 specimens within a single week, January 27
to February 4; five adult and two young males, one of the adults with enlarged
testes, the others not, and five adult and three young females, all in non-
breeding state; bill black; iris chestnut; feet pale pinkish gray to pale flesh color;
stomach contents, beetles and other insect fragments. This series agrees with
very extensive material from the forests of western Uganda.

Stizorhina fraseri vulpina Reichenow
The rufous flycatcher-thrush occurs in most of the forests of Uganda and
its presence in Malabigambo is not surprising. Between January 28 and Febru-
ary 17, nine specimens were collected, suggesting the frequency of this bird
in that area, six males and three females, with gonads ranging from small to
enlarged; bill black, the base of the mandible pale yellowish in one female; iris
dark brown; feet pale brown-whitish brown to medium purplish brown;
stomach contents, mostly ants but including some green beetle fragments;
weight, males 34 to 37 gram,s females 34 to 38 grams.

As noted by Jackson (1938, p. 918), Neave reported this species from the
nearby Tero Forest many years ago.

Neocossyphus poensis praepectoralis Jackson
The white-tailed ant-thrush is common in the Malabigambo where ten
specimens were obtained January 27 to February 16, five adult males, four
adult females, and one juvenile female. The males showed gonadal variation
from small to large, the females were all in non-breeding state; bill dull black,
in one case with paler edges; iris dark brown; feet varying from dusky flesh
color to pale pinkish brown to grayish olive; stomach contents, ants in every
case, other insect remains as well in one, and a single fruit in another; weight,
males 47 to 52 grams, females 48 to 55 grams. Most of these birds were netted
in open undergrowth in the seasonal (dry) swamp forest. These specimens
agree with very long series from western Uganda. This thrush appears not to
have been recorded east of western Uganda, judged by Ripley’s statement of
range (1964, p. 96), but the map in Mackworth-Praed and Grant (1955, p.
255) suggests Buddu as the southeastern limit of its distribution. Thus the

1969

Birds of Sango Bay Forests in Uganda

31

present examples are probably of this peripheral population of the subspecies.

Mackworth-Praed and Grant state that the juvenal plumage of this bird
is unknown. It may, therefore, be of interest to note that our young bird, not
fully grown, resembles the adults except that the rufous brown of the abdomen
and lower breast extends farther forward, and even tints the grayer chin and
throat, and the feathers of the top of the head, back, and upper wing coverts
are tipped with rufous brown. It is in post-juvenal molt and may have lost
some of the other characters of its first pennaceous feathering.

Saxicola torquata axillaris (Shelley)

The stone chat is not a true forest bird, but was found in a small open
grassy glade in the Malabigambo, where two adult males and one female (by
plumage, sex not recorded) were obtained February 5, 11, and 12; bill black;
iris dark brown; feet black; stomach contents, insect fragments, insect eggs,
seeds; weight, 15 to 16 grams.

Myrmecocichla nigra (Vieillot)

The sooty chat is not a forest bird, but two adult males were taken in
tussock grass at the edge of the Malabigambo on February 13. Both were in
non-breeding state; bill black; iris dark brown; feet black; stomach contents,
ants; weight, 41 and 42.5 grams.

Only one other male, from near Fort Ternan, western Kenya, has been
available for comparison. It has a slightly longer bill (11.8 mm from anterior
end of nostril to the tip, as opposed to 9.5 and 10.4 mm in the Malabigambo
birds) and a shorter wing (94.1 mm as opposed to 97 and 99.5 mm in the
Sango Bay specimens) ; the tail length is similar in all three, 60.1 to 62.8 mm.

MUSCICAPIDAE

Muscicapa caerulescens brevicauda Ogilvie-Grant

One adult female, in non-breeding condition, was collected in the seasonal
(dry) swamp forest, on February 14; maxilla blackish horn; mandible bluish
gray; iris very dark brown; feet brownish gray; stomach contents, insect frag-
ments and one whole beetle; weight, 17 grams. The bird was observed hawking
for insects high up under the canopy of the dense forest.

Muscicapa griseigularis (Jackson)

The gray-throated flycatcher is numerous in the Malabigambo forest and
was found in the Namalala as well. In the former, five specimens were obtained
February 1 to 14; in the latter one non-breeding female was taken February 18.
The Malabigambo birds comprised two of each sex and one of unrecorded sex,
all in non-breeding state; maxilla black; mandible paler, grayer, with blackish
edges; iris dark brown; feet bluish gray; stomach contents, insect remains and

32

Contributions in Science

No. 162

insect eggs; weight, 10 to 13 grams. These examples agree wtih a long series
from the forests of western Uganda (Bwamba, Bugoma, Kibale) .

Megabyas fiammulata aequatorialis Jackson
The shrike-flycatcher occurs in most, if not all, of the forests of Uganda,
and its presence in the Malabigambo is therefore not surprising. Between Jan-
uary 26 and February 15, four examples were obtained, three males and one
unsexed bird in female plumage; one of the males had enlarged testes, the
other did not; bill black; iris red; feet dull pinkish brown to purplish brown;
weight, 30 to 34 grams.

Bias musicus musicus (Vieillot)

This crested, black and white flycatcher was found in both the Mala-
bigambo and the Namalala Forests, but apparently not in numbers. A single
specimen was taken in each forest, an unsexed bird in female plumage in the
Malabigambo on February 12, and a male with enlarged testes in the Namalala
on February 18; bill black; iris bright chrome yellow in the male, pale yellow
with the outer edge green-gray in the female (?) ; feet pale yellow in the male,
pale gray with brownish wash in the other; stomach contents, insect fragments;
weight, 21 to 23 grams.

Platysteira cyanea ny ansae Neumann
The wattle-eye was not met with in Malabigambo, but two examples, one
of each sex, were taken in the Namalala Forest on February 17 and 18, the
male with large, the female with small, gonads; bill black; iris deep gray with a
violet tinge in the female, dark gray with a silver inner ring in the male; feet
dark brownish gray; eye wattle in the male orange red on upper half, dull gray
below, in female similar but with a leaden olive tinge on lower part; stomach
contents, insect remains; weight, 16 and 17 grams. These birds agree with
others from western Uganda. The birds were found in the tops of small trees in
mixed swamp forest and tall swamp grass.

Platysteira castanea castanea Fraser

Common in the Malabigambo Forest, where seven specimens were col-
lected February 2 to 16; five males, four of which had enlarged testes, two
females with small ovaries; bill black; iris dark brown; feet dull maroon; wat-
tles dull maroon, some with paler edges; stomach contents, insect fragments,
small beetles; weight, 13 to 15 grams. The birds were netted in the dense under-
growth of the dry swamp forest. Many years ago Neave collected this species
in the Tero Forest, Sango Bay (Jackson, 1938, p. 933).

Platysteira blissetti jamesoni (Sharpe)

This wattle-eye is very abundant in Malabigambo Forest, as is shown by

1969

Birds of Sango Bay Forests in Uganda

33

the collection there of 14 specimens between January 27 and February 9; five
males, six females, three birds of unrecorded sex, the sexed birds with gonads
ranging from small to large; bill black, with a whitish tip in one example; iris
gray brown to dark red brown; feet dull gray maroon; wattles pale turquoise;
stomach contents, insect fragments and insect eggs; weight, 10 to 14.9 grams.
Most of the specimens were netted in the open undergrowth of the seasonal
(dry) swamp forest. The present specimens agree with long series from the
forests of western Uganda.

Trochocercus nigromitratus (Reichenow)

The dusky crested flycatcher is very abundant in the Malabigambo
Forest; 15 specimens were taken there January 28 to February 6; ten males,
four females, one unsexed, all but one with small gonads; bill black; iris dark
brown; feet dark gray; stomach contents, insect remains; weight, 8.9 to 10.5
grams. Long series from the forests of western Uganda agree with these
specimens in coloration and in size. In the absence of material from elsewhere,
we accept White’s decision (1963, p. 36) that this species does not break up
into recognizable races. If toroensis Jackson were to be reinstated, our present
series would belong to it. Most of the Malabigambo specimens were netted in
the open undergrowth of the seasonal (dry) swamp forest, where they were
observed searching for food among the low branches. This flycatcher was
reported from the Tero Forest, Sango Bay, many years ago by Neave
(Jackson, 1938, p. 941 ).

Terpsiphone rufiventer somereni Chapin
The black-headed paradise flycatcher was met with in both the Mala-
bigambo and the Namalala Forests; four specimens were taken in the former,
January 27 to February 16, one in the latter on February 18; two males, one
with small and one with large testes, three females, two with small and one with
large, ovaries; bill blue gray to pale metallic cobalt, with darker tips and edges;
iris dark brown; eye ring cobalt; feet blue gray; stomach contents, small insect
fragments; weight, 13 to 19 grams.

SYLVIIDAE

Schoenicola platyura alexinae (Heuglin)

This broad-tailed warbler is not a forest bird, but three examples were
collected in patches of open grassland in the seasonal (dry) swamp forest
February 9, 11, and 15; all were males with enlarged testes; maxilla black;
mandible gray to pale gray; iris pale to dark brown; feet horn brown to gray
brown; stomach contents, insect remains; weight, 14, 15, 16 grams.

Phylloscopus sibilatrix (Bechstein)

A winter visitor to central Africa from Eurasia, this warbler, characterized

34

Contributions in Science

No. 162

by its extremely short outermost remex, is represented in the collection by an
adult male in non-breeding state, taken on February 13. It was shot in a tall
forest tree, where it was seen actively hunting for insects among the foliage;
maxilla horn brown, with paler tomial edges; mandible pale horn; iris dark
brown; feet pale yellowish brown; stomach contents, insect fragments; weight,
8.5 grams.

Cisticola erythrops sylvia Reichenow
This grass warbler was found in tall grass in a clearing next to a wet
swamp in the Namalala Forest on February 18, when an adult female, with
slight ovarian enlargement, was taken; maxilla black; mandible whitish gray;
iris hazel brown; feet pale brown; stomach empty; weight, 14 grams.

Cisticola hunteri chubbi Sharpe

One specimen, adult male with slightly enlarged testes, was taken in an
open glade in the Malabigambo, February 15; bill black, with a gray patch on
the base of the mandible; iris reddish brown; feet whitish brown; stomach
empty; weight, 19 grams.

Cisticola robusta nuchalis Reichenow
On February 18, one female in non-breeding state, was collected in
tussock grass in an open glade of the Malabigambo Forest; maxilla black;
mandible gray, somewhat darker at the tip; iris pale brown; feet whitish brown;
stomach contents, insect remains; weight, 14 grams.

Cisticola natalensis strangei (Fraser)

Two specimens, one of each sex, both with slight gonadal enlargement,
were taken in tall grass near the edge of the Malabigambo on February 1 1 ;
maxilla dark horn, paler on the distal half; mandible paler horn; iris pale buff
brown; feet flesh color with an orange tinge; stomach contents, insect remains
and one caterpillar skin; weight, 14 grams (female), 21 grams (male).

Cisticola brachyptera brachyptera (Sharpe)

This grass warbler was found in tussock grass in an open glade in the
seasonal (dry) swamp forest, on February 10, when one male, with slightly
enlarged testes, was collected; maxilla black with buff edges; mandible buff
with grayish tip; iris pale yellowish brown; feet pale yellowish brown; stomach
contents, insect fragments; weight, 9 grams. White (1962a, p. 676) states that
this race becomes more heavily streaked above in Uganda south to Mabira and
Entebbe, but our present specimen is, if anything, less streaked than one from
Bwamba Valley.

1969

Birds of Sango Bay Forests in Uganda

35

Prinia leucopogon reichenowi (Hartlaub)

The white-chinned prinia has a wide-spread range across Uganda and to
the lands beyond to the west, east, and south, so its occurrence in the open
areas of the Namalala Forest is not surprising. It is more surprising, however,
that it was not met with in similar areas in the Malabigambo. Four specimens
were obtained in the former forest on February 18, one male with enlarged
testes, and three females with ovaries varying from small to large; bill black;
iris chestnut; feet dull brownish pink; stomach contents, insect fragments and
insect eggs; weight, 12 to 14.5 grams (females), 14 grams (male). These
specimens agree closely with others from western Uganda.

Apalis jacksoni jacksoni Sharpe

The black-throated apalis is common in the Malabigambo; ten specimens
were collected January 29 to February 12; six males and four females all in
non-breeding state; bill black; iris gray brown to dark brown; feet light brown
to flesh with an orange tinge; stomach contents, insect fragments; weight, 8 to
9 grams in the females, 9 to 10 grams in the males. These examples agree
with a series from the Impenetrable Forest, extreme southwestern Uganda.

The collectors found this warbler in small (family ?) parties of three or
four individuals, foraging among the leaves of the medium-sized and smaller
trees, where they were relatively tame, or, at least, easy to approach.

Apalis binotata binotata Reichenow

The masked apalis was found in the dense undergrowth of the seasonal
(dry) forest, where four males, two adult and two subadult, and one female
were collected, February 2 to 16; two of the males with enlarged testes, the
other specimens with little gonadal swelling; bill black, the edges paler and
the gape yellowish in the two subadult males; iris pale brown to hazel brown;
feet dark gray brown to dark brown, the toes paler; stomach contents, insect
fragments and small beetles; weight, 7 to 9 grams.

One of the adult males has the top of the head more blackish than the
other and agrees in this respect with the race personata, of which a good series
from the Impenetrable Forest has been available for comparison; the other
specimens have the gray crown and occiput as in typical binotata from the
Kibale Forest. Yet the population in the Malabigambo can hardly include both
races, and since four of the specimens are gray crowned they must be called
binotata. The fact that the fifth example is like personata can only be inter-
preted as showing that in occasional individuals the coronal melanism of that
race crops up in the nominate form. As pointed out in an earlier paper (Fried-
mann, 1966, p. 39) the ranges of the two subspecies are difficult to visualize,
but peronata is more of a highland race, chiefly over 5000 feet elevation, while
binotata reaches 5000 feet at least in the Kibale Forest.

36

Contributions in Science

No. 162

The two male examples that are listed above as subadult individuals have
the chin and the sides of the upper throat whitish with the black chiefly on the
middle and posterior portions of the throat. They resemble the description of
the female in Jackson (1938, p. 1059), which leads one to think that the sub-
adult male plumage may be like that of the female.

The collectors found this warbler chiefly in the dense undergrowth of
the seasonal (dry) swamp forest.

Apalis rufogularis nigrescens (Jackson)

The red-breasted forest warbler is apparently not uncommon in the
Malabigambo, where three specimens were collected on February 14, in tall
tree tops in the seasonal (dry) swamp forest, one of each sex and one whose
sex was not recorded (male by plumage characters); maxilla blackish; man-
dible gray to whitish gray; iris light brown to brick red; feet brown, toes grayer;
stomach contents, insect fragments; weight, 8 to 9 grams.

Bathmocercus cerviniventris vulpinus Reichenow
The black-faced rufous warbler is a very common bird in the Mala-
bigambo Forest, where nine males, five females, and one unsexed bird were
collected, January 27 to February 9, with gonads ranging from small to much
enlarged; bill black; iris dark brown; feet dark brownish gray to blue gray;
bare skin on throat pinkish to pale blue; stomach contents, insect fragments
and insect eggs; weight, 15.5 to 19.5 grams in males, 13 to 16 grams in females.

These birds were found feeding close to the ground in the seasonal swamp
forest. They have a high pitched, very ringing, almost “anvil-like” whistle,
repeated about every 30 seconds.

In an earlier report (Friedmann, 1966, p. 41) western Uganda birds were
referred to the race jacksoni, but since then additional material has served to
support Chapin’s conclusion (1953, p. 418) that jacksoni could not be sep-
arated from vulpinus, and it is reassuring to find that White has also considered
all eastern Congo, Uganda, and Kenya specimens as vulpinus.

Camaroptera chloronota toroensis (Jackson)

A very common bird in both the Malabigambo and Namalala Forests; 16
specimens were collected January 27 to February 18; seven males, five females,
and four whose sex was unrecorded; mostly in non-breeding condition, but
one with enlarged testes (February 5); maxilla dark horn to blackish, with
paler edges; mandible pale pinkish gray to yellowish gray, darker at the tip;
iris pinkish brown to hazel brown and to dark brown; feet pale brown to dark
brown, with the toes paler; stomach contents, insect fragments; weight, 10 to
13 grams in males, 9 to 10.5 grams in females. These birds were found chiefly
in the dense undergrowth in the seasonal swamp forest, then dry.

1969

Birds of Sango Bay Forests in Uganda

37

Camaroptera brachyura tincta (Cassin)

Like the preceding species, this one was found in both the Malabigambo
and the Namalala Forests, but smaller numbers of individuals were collected.
Between January 26 and February 17, seven specimens were taken, three of
each sex and one of unrecorded sex, one with enlarged gonads, the others small;
bill black, the distal part of the tomium and the tip paler; iris hazel brown to
pale buff; feet pink or brownish white with a brownish wash; stomach contents,
insect fragments; weight, 10 to 12.5 grams.

Sylvietta virens baraka Sharpe

This short-tailed warbler was found in both the Malabigambo and the
Namalala Forests; two females were taken in the former on February 2 and
16; one of each sex in the latter on February 19; all showed little or no gonadal
enlargement; bill grayish horn brown, paler on the mandible; iris light brown
to hazel brown; feet pinkish brown to orange brown; stomach contents, insect
fragments; weight, 9 to 10 grams.

Macrosphenus flavicans hypochondriacus (Reichenow)

The yellow longbill is a common bird in the Malabigambo, where seven
specimens were collected February 1 to 18, four males and three females, all
in non-breeding state except for a male taken on February 2; maxilla dark
brown horn to black, with whitish tomial edges and tip; mandible flesh horn
to pale pinkish gray to whitish gray; iris bright lime yellow; feet grayish brown,
bluish gray, to pale pinkish gray; stomach contents, insect fragments; weight,
12 to 14 grams in the males, 12 to 12.5 grams in the females. The collectors
found these birds chiefly in the undergrowth of the seasonal swamp forest.

Hylia prasina prasina (Cassin)

A series of 12 specimens taken between January 26 and February 16 is
evidence of the abundance of this bird in the Malabigambo Forest. Of these
five were males, six females, one was unsexed; three of the males had large
gonads, the other two and the females were in non-breeding condition; bill
black to dark brown horn, the edge of the mandible yellowish in some; iris
dark gray brown to dark brown; feet pale to dark olive; stomach contents,
beetles and other insect fragments; weight, 14 to 15.5 grams in the males, 11
to 13.5 grams in the females. One of the males has the entire underparts washed
with olive green, and is probably subadult. The birds were netted close to the
ground in the undergrowth of the seasonal swamp forest. Many years ago
Neave collected a specimen of this species in Buddu (Jackson, 1938, p. 1369).

38

Contributions in Science

No. 162

CAMPEPHAGIDAE
Campephaga quiscalina martini Jackson

The purple-throated cuckoo-shrike is common in the Malabigambo For-
est; four males, three females, and one unsexed bird in male plumage were
collected January 28 to February 14, their gonads were small or slightly en-
larged in all but one in which the testes were large; bill black; iris dark brown;
feet dark brown to black; the inside of the mouth in two males bright orange,
tongue also bright orange; stomach contents, chiefly caterpillars but also other
insect remains; weight, 34 to 39 grams in the males, 33 to 40 grams in the
females.

LANIIDAE

Dryoscopus angolensis nandensis Sharpe

This pink-footed puff -back shrike appears to be really abundant nowhere
in its range. In the Malabigambo Forest, two males, one with enlarged testes
and one with only slight gonadal swelling, were taken on January 30 and 31;
bill black; iris dark brown with aquamarine outside the pupil; feet grayish
pink; stomach contents, beetle and other insect fragments; weight, 36 and 37
grams. One of these birds was shot in the top of a very tall wild fig tree; both
specimens are in fairly worn plumage, especially on the upper surface of the
wings.

Lanius collaris smithii (Fraser)

One male, in non-breeding state, was collected at the edge of the forest at
Katera, Sango Bay, February 13; bill black; iris dark brown; feet black;
stomach contents, insect remains; weight, 37 grams. A single male with an
imperfect tail, the outer rectrices largely shot away, is impossible to place with
certainty between smithii and capelli, both of which White (1962b, pp. 44, 45)
records from western Uganda, but it seems that the population of the Sango
Bay area is probably smithii, especially since what is left of the basal portions
of the outer tail feathers of our specimen has no black areas. In capelli these
feathers are said to have considerable black in the basal section of the outer-
most rectrices.

ORIOLIDAE

Oriolus brachyrhynchus laetior Sharpe

One of these orioles was shot in a mixed flock of several kinds of birds in
an open glade in the Malabigambo on February 14; the bird was not sexed by
the collectors; wing, 116 mm; bill reddish brown; iris crimson; feet bluish
gray; stomach contents, insect remains; weight, 53 grams.

1969

Birds of Sango Bay Forests in Uganda

39

STURNIDAE

Lamprotornis splendidus splendidus (Vieillot)

This brightly colored starling was common in the Malabigambo, where
four specimens were collected, one male, two females, and one unsexed by
the collectors, January 28 to 31; all with small gonads; bill black; iris white to
bright yellowish white; feet black; stomach contents, wild figs, small hard
fruits, and fruit seeds; weight, 143 grams in the male, 103 to 115 grams in the
females. The birds were found feeding in the lower trees. The loud swishing
sound made by the wings in flight is remarkable for a bird this size.

Lamprotornis caudatus purpuropterus Ruppell
Ruppell’s long-tailed starling was met with in the Malabigambo but once,
on February 17, when one male with large testes, and one female with no
ovarian enlargement, were collected; bill black; iris pale creamy white; feet
black; stomach contents, fruit skins and small whole fruits; weight, 86 grams
in the female, 105 grams in the male.

NECTARINIIDAE

Anthreptes rectirostris tephrolaema (Jardine and Fraser)

The fact that between January 28 and February 10 the collectors obtained
six specimens of this sunbird indicates that it is common in the Malabigambo
Forest. Of these, three are adult males (one left unsexed, but male by plum-
age), one is a young male, and two are adult females; all in non-breeding state;
bill black to dark gray horn; in the young male the basal half of the mandible
and the gape pale yellow; iris deep red brown to deep brick red; feet blackish;
stomach contents, insect fragments and small fruit seeds; weight, 11 to 12
grams in the males, 10 to 11 grams in the females. The birds were found
feeding among hanging lichens in the seasonal (then dry) swamp forest, where
they moved about actively high in the trees.

Anthreptes collaris garguensis Mearns
Common in the Malabigambo Forest; five specimens were obtained
January 27 to February 7, one adult male, one immature male, two adult
females, one unsexed bird, all in non-breeding state; bill black; iris dark
brown; feet black; stomach contents, largely small fruit seeds (in 4 of the 5
specimens), insect remains and one insect egg in one case; weight, 8 to 9
grams. These examples agree with others from western Uganda and from
Kenya.

Nectarinia seimundi traylori Wolters
This sunbird would seem to be less numerous in the Sango Bay forests
than the preceding two species; a single male, with enlarged testes, was ob-

40

Contributions in Science

No. 162

tained in the Malabigambo on February 1, and a young bird of unrecorded
sex in the Namalala on February 17; maxilla black; mandible yellowish gray
or yellowish horn; iris dark brown; feet black; stomach contents, insect frag-
ments; weight, 6.5 to 7 grams. The birds were found feeding amongst the
flowers on flowering trees in the swamp forests.

Nectarinia olivacea ragazzii (Salvadori)

The olive sunbird is abundant in the Malabigambo; in just about a week,
January 27 to February 4, 15 specimens were collected, seven of each sex
and one of unrecorded sex, with gonads ranging from small to large (in two
examples); bill black, the base of the mandible yellowish; iris dark brown;
feet dark brown to brownish black; stomach contents, small fruit seeds and
bramble seeds, and, in one case, a spider and some insect remains; weight, 10
to 12 grams. These sunbirds were found chiefly in the undergrowth of the
seasonal (dry) swamp forest. This is surprising as Jackson (1938, p. 1359)
wrote that the species is, “. . . not easy to obtain owing to its frequenting the
tallest tree-tops . .

In Williams’s experience, throughout Uganda this is essentially a species
of the forest undergrowth that frequents tall trees only when attracted by their
blossoms. It favors flowers of various species of parasitic Loranthus, a high
forest Erythrina; and the flowers of the flame tree, Spathodea nilotica, seem
to attract it.

Nectarinia verticalis viridisplendens (Reichenow)

Abundant in the Sango Bay forests, this sunbird was met with both in
the Malabigambo and in Namalala, where 13 examples were obtained, Febru-
ary 1 to 18; nine adult males, three adult females, one of unrecorded sex, with
gonads varying from small to large; bill black; iris dark brown; feet black;
stomach contents, small beetles and other insect fragments; weight, 13.5 to 16
grams in males, 12 to 14 grams in females. These green-headed sunbirds were
collected in flowering Erythrina trees near the forest edge. Unlike the olive
sunbird (preceding species), they did not come down to the mist nets in the
undergrowth, and, also unlike that species, they appear to feed wholly on
insects and spiders, not on small seeds.

Nectarinia chloropygia orphogaster (Reichenow)

This species was found commonly in both the Malabigambo and Nama-
lala Forests; ten specimens were taken January 31 to February 19, six males,
three females, one of unrecorded sex, with gonads varying from small to large;
bill black; iris dark brown; feet black; stomach contents, insect fragments;
weight, 5.5 to 7 grams. These sunbirds were largely caught in mist nets in the
dense undergrowth of the seasonal swamp forest. Pittman (in litt., July 1968)

1969

Birds of Sango Bay Forests in Uganda

41

informs us that he collected this species in the Malabigambo years ago (one
female, sent to the British Museum) .

Nectarinia bouvieri (Shelley)

One adult male, with slight testicular enlargement, was taken in the
Namalala Forest on February 19; bill black; iris dark brown; feet black;
stomach contents, insect fragments; weight, 8 grams.

Nectarinia cuprea cuprea (Shaw)

The coppery sunbird was not met with in the Malabigambo, but one adult
male, with enlarged gonads, was taken in the Namalala on February 13; bill
black; iris dark brown; feet black; stomach contents, insect remains; weight, 9
grams. The bird was found feeding among the flowers of a small Erythrina
tree in wet swamp forest.

Nectarinia erythrocerca Hartlaub

The red-breasted sunbird was met with commonly in the Namalala For-
est, February 17 and 18, where ten specimens were collected, but, strangely
enough, only a single example was procured in the Malabigambo, on February
4. Of this total series six were males, four were females and one was unsexed
by the collectors; gonads varying from small to large; bill black; iris dark
brown; feet black; stomach contents, insect fragments; weight, 8 to 10 grams.
Most of the specimens were netted in tall swamp grass at the edge of the
forest.

Nectarinia super ba buvuma (van Someren)

This large sunbird was common in the Malabigambo where five examples
were collected January 28 to February 14, three adult males, one adult female,
one unsexed but apparently male, molting into full adult plumage; bill black;
iris dark brown; feet black, dark blackish brown in the female; stomach con-
tents, insect fragments, insect eggs, spiders; weight, 17 grams in the female,
17.5 to 21 grams in the males. These birds were found in the seasonal swamp
forest where some of them were feeding in the Erythrina trees. Elsewhere in
Uganda the species often visits cultivated areas near the forests, attracted
there by the banana flowers.

Nectarinia kilimensis kilimensis Shelley

This long-tailed sunbird was not met with in the Malabigambo, but a
single adult male was taken in the Namalala on February 13; state of gonads
unrecorded; bill black; iris dark brown; feet black; stomach contents, insect
fragments; weight, 15 grams. Apparently the species was not common in the
Sango Bay forests, but it may be noted that it was recorded years ago from
Buddu by Jackson (1938, p. 1317).

42

Contributions in Science

No. 162

PLOCEIDAE

Ambylospiza albifrons melanota (Heuglin)

The grosbeak weaver is represented by six specimens, all taken in the
undergrowth or in the bushes and trees of the seasonal (dry) swamp forest of
Malabigambo, February 2 to 18; two adult males and four adult females, one
of the males with large gonads, the other five birds in non-breeding condition;
bill dull black, with paler areas on the underside of the mandible in males, the
maxilla horn brown and the mandible yellowish horn brown in females; iris
dark brown; feet dark gray brown to dark brown; stomach contents, crushed
seeds, small whole fruits, and grit; weight, 46 to 56 grams in males, 33 to 42
grams in females.

Ploceus pelzelni pelzelni (Hartlaub)

The slender-billed weaver was not met with in the Malabigambo but was
very common in the Namalala Forest, where eight specimens were collected
February 13 to 19; three males and five females, the males all with enlarged
gonads and the females all with small ones; bill black; iris grayish white to
buff, with darker center; feet gray to brownish gray; stomach contents, small
whole fruits; weight, 14 to 17.5 grams in males, 12 to 15 grams in females. The
birds were found in tall tussock grass in wet swamp forest glades.

Ploceus ocularis crocatus (Hartlaub)

The spectacled weaver was found in the Namalala Forest where two ex-
amples, one of each sex, were taken February 17 and 18, the male with large
testes, the female with slight ovarian enlargement; bill black; iris yellowish
white in male, white in female; feet gray, with a brownish tinge in the male;
stomach contents, insect remains; weight, 27 grams in the male, 37 grams in
the female. The birds were collected in tall grass in a clearing in the forest.

Ploceus nigricollis nigricollis (Vieillot)

Common in the Malabigambo; 12 specimens collected January 26 to
February 10, eight males, four females, mostly with small gonads but two of
the males with enlarged testes; bill black; iris creamy brown to dark brown;
feet gray with a brown or a bluish tinge; stomach contents, fruit seeds and
insect fragments; weight, 28 to 40 grams in males, 26 to 27.5 grams in females.

Ploceus xanthops (Hartlaub)

The larger golden weaver is represented in the collection by a single male
with enlarged testes, taken at Katera, February 13; bill black; iris whitish
cream; feet brownish horn; stomach contents, seeds and some insect remains;
weight, 45.5 grams. The bird was taken from a low bush at the forest edge.

1969

Birds of Sango Bay Forests in Uganda

43

Ploceus castanops Shelley

Two males with enlarged testes, and three females, one with large, the
others with slightly swollen ovaries, were taken in the Namalala Forest,
February 17 and 18; bill black; iris yellowish white in males, white in females;
feet brown; stomach contents, insect fragments only; weight, 19.5 to 21 grams
in females, 21 and 22 grams in males. The birds were netted in tall grass in a
clearing in the swamp forest.

Ploceus nigerrimus nigerrimus Vieillot
Vieillot’s black weaver is a very common bird in the Sango Bay area, and
was met with in numbers in both the Malabigambo and the Namalala Forests,
where 16 examples were collected, January 30 to February 17, five males, 10
females, and one unsexed specimen; four of the males with enlarged testes,
the other male and all the females with no or only slight gonadal swelling; bill
black in males, dark horn gray brown, paler toward the base of the mandible,
in females; iris lemon yellow in males, creamy to whitish, flecked with brown,
in females; feet brown; stomach contents, small fruit seeds, bramble seeds,
insects; weight, males 40.5 to 42.5 grams, females 28 to 33 grams. Some of
these birds were shot from the taller flowering trees while others were caught
in nets in the undergrowth, indicating a wide range of food foraging.

Ploceus weynsi (Dubois)

Weyns’s weaver reaches its eastern limits in southern Uganda; the Sango
Bay area must be close to the southeastern periphery of its range. It is common
in the Malabigambo Forest, where seven specimens were taken January 27 to
29; three adult males and three adult and one young females, all in non-
breeding state; bill black in the males, dark horn brown to horn black, with
the mandible basally somewhat paler, in the females; iris lemon yellow with a
chestnut inner ring; feet brown to grayish brown; stomach contents, fruit seeds;
weight, 36 to 37 grams in males, 30 to 33 grams in females.

This relatively little-known species was found high up in the tall trees in the
seasonal (dry) swamp forest, usually in small groups of 10 to 20 individuals.
Its flight was noted as being very fast.

Williams has noted this weaver from time to time near Entebbe, often in
large flocks. There it frequents both lush lake shore vegetation and forest
tree tops.

Ploceus melanocephalus fischeri Reichenow

This weaver was met with at Katera and in the Namalala Forest, Febru-
ary 13 to 18, when three specimens were obtained, all males with enlarged
testes; bill black; iris dark brown; feet brown; stomach contents, insect frag-

44

Contributions in Science

No. 162

ments and a few seeds; weight, 29 to 31 grams. These specimens agree with a
series from western Uganda.

Ploceus super ciliosus (Shelley)

The compact weaver is represented in the collection by two adult males,
one taken in the Malabigambo Forest, February 9, the other at Katera on
February 17; neither were in breeding condition; maxilla blackish; mandible
pale grayish with a dark gray area in the center of the underside; iris dark
brown; feet pale brown to flesh brown; stomach contents, crushed whitish
seeds; weight, 21 and 24 grams.

Malimbus rubricollis rubricollis (Swainson)

The red-headed malimbe was met with in the Malabigambo Forest on two
occasions; two adult males, with slight testicular enlargement were collected
there on January 30, February 18; bill black; iris dark brown; feet black;
stomach contents, grasshopper, insect remains, and fruit seeds; weight, 41.5
and 47 grams.

Malimbus malimbicus malimbicus (Daudin)

One subadult male not in full adult plumage, without an occipital crest
and with some mixture of black among the red of the crown, occiput, and
breast, was taken in the Malabigambo Forest on February 13. This specimen
record constitutes a considerable eastern extension of the known range of the
species, previously not reported east of western Uganda (Bwamba, Budongo
and Bugoma Forests). Our present example was shot out of a small mixed
flock of several species foraging fairly low down in the seasonal (dry) swamp
forest; bill gray horn brown, paler on the mandible; iris dark brown; feet
blackish brown; stomach contents, insect remains; weight, 38 grams. The
specimen agrees with two others, in the same plumage stage, from the
Bugoma Forest.

Euplectes axillaris phoeniceus (Heuglin)

This whydah is not a true forest bird, but it was found in the tall grass in
open glades in the Malabigambo Forest, where six specimens were collected
February 9 to 13, two males with slight to considerable testicular enlargement,
and four females with small ovaries; bill pale gray in males, brown horn with
the mandible paler in females; iris dark brown; feet black in males, brown
horn in females; stomach contents, small seeds and a little grit; weight, 24 and
25 grams in males, 18 to 22 grams in females.

Vidua macroura (Pallas)

The ubiquitous pin-tailed widow bird is not a forest dweller, but a single

1969

Birds of Sango Bay Forests in Uganda

45

adult male, with enlarged testes and in full nuptial plumage, was taken at
Katera on February 18; bill red orange with black edges; iris dark brown; feet
blackish gray brown; weight, 12 grams.

Hypargos nitidulus schlegeli (Sharpe)

The green-backed twin-spot was found in the undergrowth in fairly open
seasonal swamp forest in both the Malabigambo and the Namalala; five speci-
mens were obtained January 28 to February 18; three males, two unsexed
birds, one of which is in male plumage, the other either female or young; bill
black, with a reddish tinge at the tip and on the distal half or so of the mandible;
iris dark brown; feet brown to pale grayish brown; stomach contents, small
crushed seeds in one, the others empty; weight, 8.5 to 10 grams. The birds were
netted in the low vegetation along a path in the seasonal swamp forest.

Pirenestes ostrinus ostrinus (Vieillot)

This extremely variable weaver is common in the Malabigambo Forest;
eight specimens were taken, February 5 to 16, four of each sex, with gonads
varying from small to large; bill black with a violet wash in the males, black
with less violet in the females; iris dark brown to very deep red brown; feet
brown; stomach contents, crushed small whitish seeds, and grit; weight, 18.5
to 22 grams. The birds were collected in the nets set in the undergrowth in the
seasonal (dry) swamp forests.

Nigrita canicapilla schistacea Sharpe
The gray-headed negro finch was found commonly in the Malabigambo
and was also collected (one specimen) in the Namalala; a series of ten speci-
mens in all, January 30 to February 19; four males and six females, the
gonads large in one of the males and small in the other nine birds; bill black;
iris red; feet dark brown; stomach contents, small fruit seeds; weight, 17 to 23
grams in males, 16 to 19.5 grams in females. These birds were found both in
the undergrowth and well up in tall trees in the seasonal swamp forest.

Spennophaga ruficapilla ruficapilla Shelley
The red-headed blue-bill was found to be very common in the Mala-
bigambo but was not met with (or at any rate, was not collected) in the
Namalala. In the former forest 16 specimens were obtained January 27 to
February 12, eight adult males, seven adult females, one young bird of un-
recorded sex, gonads enlarged in one male, small or only slightly enlarged in
the other examples; bill metallic violet blue to black, with tomial edges and tip
reddish; iris dark brown; feet dark brown; stomach contents, chiefly small
seeds, some insect fragments, and grit; weight, 20 to 25 grams in the males,
20 to 24 grams in females. Most of these birds were netted in the undergrowth
in the seasonal swamp forest.

46

Contributions in Science

No. 162

Nesocharis ansorgei ansorgei (Hartert)

One of the surprises of the collection was to find that this weaver finch
extends east so far as the Sango Bay area, about 170 miles east of its pre-
viously known range. At the edge of the Malabigambo Forest two adult males
and one female were taken, January 30 and 31, all with slight gonadal enlarge-
ment; bill blackish horn, with the portion of both the maxilla and the mandible
bluish white; iris very dark brown; feet grayish brown; stomach contents, small
crushed seeds; weight, 8 to 9.5 grams. These specimens have been discussed
elsewhere (Friedmann, 1968b) in connection with the distributional picture
of the whole genus.

Estrilda paludicola paludicola Heuglin
This little waxbill is not a forest bird but was found in grassy places in
open glades in both the Malabigambo and the Namalala Forests, where seven
specimens were collected, February 9 to 13; the gonads varying from small to
enlarged; bill red; iris bright red; feet brown to dark brown; stomach contents,
seeds and fine grit; weight, 6 to 8 grams.

Estrilda astrild peasei Shelley

Another non-sylvan waxbill, this species was collected in an open glade
in the Malabigambo, where one female was taken on February 15, and at
Katera, where a male was collected February 17; both birds in non-breeding
state; bill bright red to orange red; iris dark brown; feet black in male, dark
gray in female; stomach contents, small seeds and grit; weight, 7.5 grams each.

Lonchura bicolor poensis (Fraser)

The balck and white mannikin was found in the grass along a path in an
open glade in the Malabigambo, February 2 and 5, when five specimens were
obtained, three males, one female, one unsexed bird, with gonads ranging from
small to enlarged; bill pale blue to horn black, with paler gray edges; iris dark
gray brown; feet dark gray; stomach contents, grass seeds and grit; weight,
8 to 10 grams.

Lonchura cucullata cucullata (Swainson)

The bronze mannikin, like the preceding species, was found in the grassy
places in an open glade in the Malabigambo Forest; two females were collected
February 6 and 7, both in non-breeding state; maxilla black, mandible pale
gray; iris dark brown; feet dark gray; stomach contents, grass seeds and grit;
weight, 9.5 grams.

1969

Birds of Sango Bay Forests in Uganda

47

FRINGILLIDAE

Serinus citrinelloides frontalis (Reichenow)

Found in both the Malabigambo and the Namalala Forests, this serin is
represented by a male taken in the former place February 10, and a female
from the latter forest, February 19; both in non-breeding state; bill gray brown
with darker tip; iris dark brown; feet brown; weight, 12 grams each.

Literature Cited

Chapin, J. P. 1939. The birds of the Belgian Congo. Pt. 2. Amer. Mus. Nat. Hist.,
Bull. 75: 1-632.

. 1944. Phyllastrephus icterinus (Bonaparte) and its larger counterpart. Ibis,

86: 543-545.

. 1953. The birds of the Belgian Congo. Pt. 3. Amer. Mus. Nat. Hist., Bull.

75 A: 1-821.

. 1962. Sibling species of small African honey-guides. Ibis, 104: 40-44.

Dale, I. R. 1954. Forest spread and climatic change in Uganda during the Christian
era. Empire Forestry Journ., 33: 23-29,

Dawe, M. T. 1906. Report on a botanical mission through the forest districts of
South Buddu and the Western and Nile Provinces of the Uganda Protectorate.
H. M. Stationery Office, London.

Deignan, H. G. 1964. Timaliinae. In Peters, J. L. (continuation of) Check-List of
Birds of the World, ed. by Ernst Mayr and R. A. Paynter, Jr. Cambridge, Mass.:
Mus. Compar. Zool., Harvard, 10: 240-427.

Friedmann, H. 1966. A contribution to the ornithology of Uganda. Los Angeles
County Mus. Nat. Hist., Bull., Sci., 3: 1-55.

. 1968a. Range and variation of the icterine bulbul in Uganda. Bull. Brit. Or-

nith. Club, 88: 110-112.

. 1968b. The olive weaver-finch, Nesocharis ansorgei ansorgei in Uganda.

Bull. Brit. Ornith. Club, 88: 135-138.

. 1969. The status and distribution in Uganda of the white-bellied kingfisher,

Alcedo leucogaster leopoldi. L. A. County Mus. Nat. Hist., Contrib. Sci., 158:
1-5.

Hall, B. P. 1963. The francolins, a study in speciation. British Museum (Nat. Hist.),
Bull. 10: 107-204.

Holtz, W. 1906-1911. Der Minzirowald in Deutsch-Buddu. In Berichte Land-u-
Forst-Wirt in D. O. Afrika: 3.

Jackson, F. J., ed. by W. L. Sclater. 1938. The birds of Kenya Colony and the
Uganda Protectorate. London: Gurney and Jackson, 1, 2, 3: 1-1592.

Mackworth-Praed, C. W., and C. H. B. Grant. 1955. Birds of eastern and north-
eastern Africa. London: Longmans, Green and Co., 2: 1-1099.

. 1957. Birds of eastern and northeastern Africa. London: Longmans, Green

and Co., 1, ed. 2: 1-806.

Neumann, O. 1899-1900. Beitrage zur Vogelfauna von Ost- und Central-Afrika. Teil
II, III. Journ. f. Ornith., 47: 33-74; 48: 185-228.

Ogilvie-Grant, W. R. 1905. On the birds collected by the late W. G. Doggett on
the Anglo-German frontier of Uganda. Ibis, (8)5: 199-212.

48

Contributions in Science

No. 162

Pitman, C. R. S. 1931a. Occurrence of the Congo forest akalat in Uganda Protec-
torate. Bateleur, 3: 50.

. 1931b. Occurrence of the yellow-crested woodpecker in Uganda Protectorate.

Bateleur, 3: 73.

Rand, A. L. 1958. Notes on African bulbuls. Fieldiana: Zoology, 35: 145-220.

-. 1960. Pycnonotidae. In Peters, J. L., (continuation of) Check-list of the Birds

of the World, ed. by E. Mayr and J. C. Greenway, Jr. Cambridge, Mass.: Mus.
Compar. Zool., Harvard. 9: 221-300.

Reichenow, A. 1892. Zur Vogelfauna des Victoria Njansa. Sammlungen Dr. Emin’s
und Dr. Stuhlmann’s 1890-91. Journ. f. Ornith., 40: 1-60.

. 1893. Die von Herrn Dr. Fr. Stuhlmann in Ostafrika gesammelten Vogel.

Jahrb. Hamburg Wissench. Anstalten X: 1-27.

— — . 1902. Die Vogel Afrikas. Atlas, 3 maps, 50 pp., 30 col. pis.

. 1911. Die Vogelfauna des Mittelafrikanischen Seengebietes. In Wissenschaft-

liche Ergebnisse der Deutschen Zentral-Afrika-Expedition 1907-1908. Band III.,
Zoologie, 1 : 231-374. Leipzig, Klinkhardt & Biermann.

Ripley, S. D. 1964. Turdinae. In Peters, J. L. (continuation of) Check-list of Birds
of the World, ed. by E. Mayr and R. A. Paynter, Jr. Cambridge, Mass.: Mus.
Compar. Zool., Harvard. 10: 13-227.

Uganda, Dept, of Lands and Surveys. 1962. Atlas of Uganda.

van Someren, V. G. L. 1922. Notes on the birds of East Africa. Novitates Zoologi-
cae, 29: 1-246.

. 1932. Birds of Kenya and Uganda, being addenda and corrigenda to my previ-
ous paper in “Novitates Zoologicae,” XXIX, 1922. Novitates Zoologicae, 37:
252-380.

White, C. M. N. 1960. A check list of the Ethiopian Muscicapidae (Sylviinae).
Part I. Natl. Mus. of Southern Rhodesia, Bulawayo, Occas. Papers, Nat, Sci.,
3: 399-430.

~. 1961. A revised check list of African broadbills, pittas, larks, swallows, wag-
tails, and pipits. Lusaka: Govt. Printer, 84 pp.

. 1962a. A check list of the Ethiopian Muscicapidae (Sylviinae). Part II. Natl.

Mus. of Southern Rhodesia, Bulawayo, Occas. Papers, Nat. Sci., 3: 653-738.

. 1962b. Revised check list of African shrikes, orioles, drongos, starlings, crows,

waxwings, cuckoo-shrikes, bulbuls, accentors, thrushes and babblers. Lusaka:
Govt. Printer, 176 pp.

. 1963. A revised check list of African flycatchers, tits, tree creepers, sunbirds,

white-eyes, honey eaters, buntings, finches, weavers and waxbills. Lusaka: Govt.
Printer, 218 pp.

. 1965. A revised check list of African non-passerine birds. Lusaka: Govt. Print-
er, 299 pp.

Accepted for publication February 14, 1969.

LOS

ANGELES

COUNTY

MUSEUM

!

CONTRIBUTIONS
IN SCIENCE

UMBER 163

June 26, 1969

So?, 73

C^LSCB

MAN AND SMI LO DON : A PRELIMINARY REPORT ON THEIR
POSSIBLE COEXISTENCE AT RANCHO LA BREA

By George L Miller

Los Angeles County Museum of Natural History • Exposition Park
Los Angeles, California 90007

CONTRIBUTIONS IN SCIENCE is a series of miscellaneous technical papers
in the fields of Biology, Geology and Anthropology, published at irregular intervals
by the Los Angeles County Museum of Natural History. 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 Robert J. Lavenberg, Los Angeles County Museum of Natural
History, 900 Exposition Boulevard, Los Angeles, California 90007.

Robert J. Lavenberg
Managing Editor

INSTRUCTIONS FOR AUTHORS

Manuscripts for the LOS ANGELES COUNTY MUSEUM, CONTRIBU-
TIONS 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. Al-
though priority will be given to manuscripts by staff members, or to papers dealing
largely with specimens in the Museum’s collections, 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 Edi-
torial Committee. They may also be subject to review by competent specialists out-
side 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.

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. Alterations or changes in the manuscript after galley proof will
be billed to the author. Unless specifically requested, page proof will not 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.

Dorothy M. Halmos
Editor

MAN AND SMILODON : A PRELIMINARY REPORT ON THEIR
POSSIBLE COEXISTENCE AT RANCHO LA BREA

By George J. Miller1

Abstract: Possible artif actual cuts on fossil bone of extinct
animals including the sabre-tooth cat, Smilodon californicus,
from the Rancho La Brea of California, have recently been un-
covered. Preliminary experiments suggest that some of the cuts
were made by primitive stone tools. The bone is at least 10,000
years old.

Introduction

One of the world’s outstanding collections of Pleistocene animals (Stock,
1963), the Los Angeles County Museum of Natural History (LACM) col-
lection from the Rancho La Brea of California, has produced possible evidence
of early man in North America at an earlier time than hitherto accepted.
Bones of three extinct kinds of animals, Smilodon californicus : one tibia and
three femora; Panthera cf. atrox : one femur; and Bison cf. antiquus : one
radius; all with what appear to be artif actual cuts and grooves have been found
in the collection.

Part of specimen number LACM T-6101, (UCLA 1292 L), a right tibia
of Smilodon californicus (Fig. 1), was given to Rainer Berger of the Institute
of Geophysics and Planetary Physics, University of California at Los Angeles,
to obtain a radiocarbon date using the new method of collagen extraction
developed by T. Y. Ho. Until this date becomes available, the age of the bone
can only be grossly estimated from its relative location to dated specimens
from the tar pits, since any stratigraphic interpretations have been questioned
on the basis of “pit churning” and other factors (Miller, 1968). However,
current investigations suggest that there may be more stratigraphy present in
the tar pits than was previously thought. If this is shown to be true, a calcula-
tion of this kind will be even more valid.

When the excavations were made at La Brea, the pit numbers, depth, and
grid were recorded for most specimens. “A grid system of 3-foot squares”
(Howard, 1960) was used. All the specimens discussed in this report have
complete pit data and can thus be placed stratigraphically within the pits.

Discussion

Specimen LACM T-6101 is from Pit 4, Grid C-2, at a depth of IIV2 feet.
Berger and Libby (1966) reported a Carbon-14 date of 33,700 ± 1600 years

iResearch Associate in Vertebrate Paleontology, Los Angeles County Museum of
Natural History, and Assistant Curator of Vertebrate Paleontology, Idaho State
University Museum, Pocatello, Idaho 83201.

1

2

Contributions in Science

No. 163

Figure 1. Right tibia of Smilodon calif ornicus, the extinct sabre-tooth “cat,” with
two possibly artifactual cuts. The two parallel, oblique cuts may have been made
by Early Man using a primitive stone cutting tool. The bone, from the Rancho La
Brea of California, is probably more than 10,000 years old, (LACM T-6101).

1969

Man and Smilodon

3

B.P. (Before Present) for a wood sample from Pit 4 at a depth of 5 feet. Dates
of approximately 10,000 C-14 years B.P. have been obtained by Geochron
Labs (Miller, 1968) for Canis dirus (the dire wolf) humeri from Pits 3 and 16.
No dates have been published for Smilodon or any of the other series of extinct
animals discussed herein. However, the most recent record of Smilodon is at
the end of the Wisconsin (Hibbard, Ray, Savage, Taylor and Guilday, 1965),
approximately 11,000 years B.P. (Ericson, Broecker, Kulp and Wollin, 1956).
The time range for Panthera atrox and Bison antiquus is not as well known,
although Geochron Labs have obtained a date of 13,800 B.P. for wood in
association with Smilodon and B. antiquus at Maricopa, California. It would,
therefore, seem safe to assume the Smilodon tibia to have a C-14 age of be-
tween 11,000 and 34,000 B.P., C-14 dates not being reliable beyond 35,000
to 40,000 years (Broecker, 1965). Using the same kind of extrapolation, the
other specimens could well fall within the same time range.

Pit data and LACM catalog numbers of the other specimens are as
follows:

Smilodon californicus : LACM A-690, Pit 4, Grid D-2, 14 feet, proximal and
medial portion of left femur.

Smilodon californicus : LACM A-726, Pit 4, Grid C-3, 12V^ feet, proximal
and medial portion of left femur.

Smilodon californicus : LACM A-1203, Pit 77, Grid G-ll, 9V2-1 \Vi feet,
proximal and medial portion of left femur.

Panthera cf. atrox : LACM A-647, Pit 3, Grid D-3, 13 feet, proximal portion
of left femur.

Bison cf. antiquus : LACM 46685, Pit 4, Grid C-2, 12 Vi feet, right radius.

The estimate of from 11,000 to 34,000 years B.P. for these specimens is
much older than for any previous evidence of man at Rancho La Brea. The
only human remains that have been found in the tar pits are those of a young
female between 25 and 30 years of age at the time of death. This author had
the good fortune to examine the remains, which consisted of a skull, mandible,
and several post-cranial skeletal parts. The specimen was in good enough
condition forKroeberin 1914 (Kroeber, 1962) and Gilford (1926) to identify
as that of an early California Indian. “La Brea Woman” has not as yet been
C-14 dated; however, it is thought that she is of Recent time (Howard and
Miller, 1939). As Merriam (1914) stated, “The age of this specimen may
perhaps be measured in thousands of years, but probably not in tens of
thousands.” An atlatl dart foreshaft from Pit 61-67 was dated by the La Jolla
Laboratory (Sample No. LJ 121) as 4450 ± 200 years B.P. (Howard, 1960).
Other “artifact materials” have been C-14 dated at about 2500 B.C. (Heizer in
Kroeber, 1962).

The parallel grooves on the Smilodon tibia and on the two femora are
approximately eleven millimeters apart (Fig. 2). Those on the specimens
shown in Figure 2, center and bottom, have sharp edges, while those on the

4

Contributions in Science

No. 163

Figure 2. Smilodon calif ornicus. Top: left femur with two oblique cuts medially,
(LACM A-690). Center: left femur with two oblique cuts medially, (LACM
A- 1203). Bottom: right tibia with two oblique cuts medially. The plaster filling is
visible in place where sample was taken for C-14 dating, (LACM T-6101).

specimen shown in Figure 2, top, are rounded. The depth of the cuts varies
from three to five millimeters.

The cuts on the Panthera femur (Fig. 3, top) are also approximately
eleven millimeters apart but have a maximum depth of fifteen millimeters.
There is no rounding or evidence of wear on the Panthera specimen. The
Smilodon femur (Fig. 3, bottom) follows a different pattern: the cut goes
through the bone to the marrow cavity and shows a high degree of “polishing”
that is not present in any of the other specimens.

The oblique cut on the proximal portion of the Bison radius (Fig. 4) is
seven millimeters deep at the maximum and the middle part goes through to
the marrow cavity. The surface of the cut is much rougher on the Bison
specimen than on any of the others.

Experiments performed by this author on fresh bone using an unre-
touched chalcedony flake, one of the most primitive stone tools known, have
shown that the cuts could have been made with a tool of this type. Fresh bone
has been given to animals in several zoos (Miller, 1969) to determine if the
cuts could have been made by animals. These experiments have explained
many of the marks found on La Brea bone, but nothing has been found to even
approximate the cuts being discussed here.

1969

Man and Smilodon

5

Figure 3. Top: Panthera cf. atrox left femur with two parallel grooves below head,
(LACM A-647). Bottom: Smilodon calif ornicus left femur with hole cut medially,
(LACM A-726).

It may also be noted that the Smilodon tibia and the two femora (Fig. 2)
have what appear to be holes drilled in them. It has been found (Miller, 1969)
that the large cats such as the modern African lion ( Panthera leo), as well as
the Canidae, often bite bone with the canines, leaving a round hole. However,
these holes do not compare at all with those found in the La Brea specimens.
It is therefore tentatively assumed that the holes in the bones of the extinct
animals were made by man.

It is difficult to visualize a purpose or reason for the oblique cuts. The
two parallel grooves seen on the three specimens in Figure 2 suggest the
“groove and splinter technique” of Upper Paleolithic and Mesolithic Europe
(Clark and Thompson, 1953), evidence of which has also been found in
California (Heizer and Kelley, 1962). However, in this process the cuts were
made parallel to the longitudinal axis of the bone and not obliquely as in the
La Brea specimens. Figures 2-4 show a recurrence of the oblique pattern with
the exception of Figure 3 (bottom). The cuts are not cracks, as the Smilodon
femur (Fig. 3, bottom) and the Bison radius (Fig. 4) are the only specimens
having cuts that go all the way through the bone, and these go through only to
the marrow cavity in the middle part of the cut. Experiments by the author with
rapid drying of fresh bone have shown that drying cracks go all the way
through the compact bone and examination of fossil bone from La Brea and
other localities has shown that “weathering” cracks go completely through the
bone. The rapid drying experiments and the examination of fossil bone have

6

Contributions in Science

No. 163

Figure 4. Bison cf. antiquus right radius with oblique cut proximally, (LACM
46685).

also shown the oblique pattern to be unusual, as an overwhelming majority of
the cracks follow either the longitudinal axis of the bone or are transverse.

The possibility of the marks having been made in butchering seems also
to have been eliminated. Theodore E. White, the leading authority on primitive
butchering techniques, after a thorough examination of the specimens, was
convinced that the cuts were not from butchering. Dr. White was of the opinion
that the cuts were artif actual (Theodore E. White, 1968, personal communi-
cation).

The nature of the cuts, especially the two parallel grooves, seems to be
too non-random to have been made during excavation. The Smilodon tibia,
LACM T-6101 (Fig. 1), was microscopically examined by flint-knappers Don
Crabtree and Francois Bordes. Dr. Bordes expressed the opinion that “if the
specimen had been found in France, there would be no question of its being
artifactual” (Francois Bordes, 1968, personal communication). It was during
this examination that minute scratches that may have been made with a stone
cutting tool were observed on the surface of the cuts. Dr. Bordes stated that
the cuts were “clearly done on fresh bone” (Francois Bordes, 1968, personal
communication).

Judged by the large amount of fossil material recovered from Rancho La
Brea — bones of more than 2400 individuals of Smilodon have been reported
(Miller, 1968), it seems quite obvious that many animals were attracted to
the area. The presence of large amounts of game would probably also have
attracted man. As a large percentage of the La Brea material shows signs of
what appears to be weathering before preservation, it may be safely assumed
that there was always bone present on top of the ground as well as in the tar
pools. It is possible that man may have worked bone on the site and since
Early Man commonly left his rejects at the site where they were produced
(Bryan and Tuohy, 1960), the material discussed in this report could be some
of his rejects.

It should be emphasized that the specimens with possible artifactual cuts
represent a very small fraction of the La Brea collection. However, only a very
small sampling has been examined for this kind of evidence.

1969

Man and Smilodon

1

The literature contains numerous records of associations of early man
and Bison antiquus in North America (Harrington and Simpson, 1961). Hind
Sadek-Kooros (1968, personal communication) has evidence of a possible
association of man and Panthera atrox. The bones discussed above, however,
seem to be the first evidence of a possible association of early man and
Smilodon.

The specimens described in this report have been examined by many
archaeologists and paleontologists in the Western United States; and so far no
one has seen anything exactly like them. It is possible that others have seen
bones from other sites having parallel, oblique cuts, and it is hoped that the
photographs in this report will prompt anyone having knowledge of bones with
similar cuts to contact the author.

Acknowledgements

I thank J. R. Macdonald and Theodore Downs for the loan of specimens
and access to documents of the Los Angeles County Museum of Natural
History; and the following persons for examining specimens, casts, or photo-
graphs and/or other help: Robert Ariss, Francois Bordes, Alan Bryan,
Donald Cocke, Don Crabtree, John Cutler, Keith Dixon, F. Gaynor Evans,
La Mar Farnsworth, Franklin Fenenga, Edson Fichter, David Fortsch,
Priscilla Hearst, Robert F. Heizer, Hildegarde Howard, Floyd Humeston,
Andre Isotoff, Leslie F. Marcus, John E. Mawby, Douglas Osborne, Richard
L. Reynolds, Charles Rozaire, John Stanley, Earl H. Swanson, Jr., Theodore
E. White, and G. Davidson Woodard. I also thank Theodore Downs, J. R.
Macdonald, Max Pavesic, John A. White and Theodore E. White for critically
reading the manuscript. Photographs of specimens from the collection of the
Los Angeles County Museum of Natural History are by the author. This
research was partially supported by National Science Foundation Grant
GB 5119.

Literature Cited

Berger, R., and W. F. Libby. 1966. U. C. L. A. Radiocarbon Dates V. Radiocarbon,
8: 1-59.

Broecker, W. S. 1965. The Quaternary of the United States. H. E. Wright and
Frey, D. G., Eds. Princeton Univ. Press, 922 p., illus.

Bryan, A. L., and D. R. Tuohy. 1960. A basalt quarry in northeastern Oregon.
Am. Phil. Soc., Proc., 104(5) :485-510.

Clark, J. G. D., and M. W. Thompson. 1953. The groove and splinter technique of
working antler in the Upper Paleolithic and Mesolithic Europe. Prehistoric Soc.,
Proc., 19(6): 148-160.

Ericson, D. B., W. S. Broecker, J. L. Kulp and G. Wollin. 1956. Late Pleistocene
climates and deep-sea sediments. Science, 124: 385-389.

Gifford, E. W. 1926. California Indian types. Nat. Hist., 26: 50-60.

Harrington, M. R., and R. D. Simpson. 1961. Tule Springs, Nevada, with other
evdience of Pleistocene Man in North America. Southwest Museum Papers,
18: 1-146.

8

Contributions in Science

No. 163

Heizer, R. F., and H. Kelley. 1962. Blades and Bladelets from California. Amer.
Phil. Soc., Proc., 106(2) : 94-105.

Hibbard, C. W., D. E. Ray, D. E. Savage, D. W. Taylor, and J. E. Guilday. 1965.
The Quaternary of the United States. H. E. Wright and Frey, D. G., Eds.
Princeton Univ. Press, 922 p., illus.

Howard, H., and A. H. Miller. 1939. The avifauna associated with human remains
at Rancho La Brea, California. Carnegie Inst, of Wash., 514: 39-48.

Howard, H. 1960. Significance of Carbon- 14 dates for Rancho La Brea. Science,
131: 712-714.

Kroeber, A. L. 1962. The Rancho La Brea skull. American Antiquity, 27(3) : 416-
417.

Merriam, J. C. 1914. Preliminary report on the discovery of human remains in an
asphalt deposit at Rancho La Brea. Science, 40: 198-203.

Miller, G. J. 1968. On the age distribution of Smilodon calif omicus Bovard from
Rancho La Brea. Los Angeles Co. Mus. Contrib. Sci., 131 : 1-17.

. 1969. A study of cuts, grooves, and other marks on Recent and fossil bone.

I. Animal tooth marks. Tebiwa, J. Idaho State Univ. Mus., 12(1): In press.
Stock, C. 1963. Rancho La Brea — A record of Pleistocene life in California. Los
Angeles County Museum, Science Ser., 20 (Paleo. No. 11).

Accepted for publication February 10, 1969.

CONTRIBUTIONS
IN SCIENCE

LOS

ANGELES

COUNTY

MUSEUM

[umber 164

June 26, 1969

fo 7' 7s
, zL

UPPER PENNSYLVANIAN VERTEBRATES FROM THE
SANGRE DE CRISTO FORMATION OF CENTRAL COLORADO

By Peter Paul Vaughn

1

■JUL 9 1969

Los Angeles County Museum of Natural History • Exposition Park
Los Angeles, California 90007

CONTRIBUTIONS IN SCIENCE is a series of miscellaneous technical papers
in the fields of Biology, Geology and Anthropology, published at irregular intervals
by the Los Angeles County Museum of Natural History. 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 Robert J. Lavenberg, Los Angeles County Museum of Natural
History, 900 Exposition Boulevard, Los Angeles, California 90007.

Robert J. Lavenberg
Managing Editor

INSTRUCTIONS FOR AUTHORS

Manuscripts for the LOS ANGELES COUNTY MUSEUM, CONTRIBU-
TIONS 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. Al-
though priority will be given to manuscripts by staff members, or to papers dealing
largely with specimens in the Museum’s collections, 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 Edi-
torial Committee. They may also be subject to review by competent specialists out-
side 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.

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. Alterations or changes in the manuscript after galley proof will
be billed to the author. Unless specifically requested, page proof will not 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.

Dorothy M. Halmos
Editor

UPPER PENNSYLVANIAN VERTEBRATES FROM THE
SANGRE DE CRISTO FORMATION OF CENTRAL COLORADO

By Peter Paul Vaughn1

Abstract: On the basis of fragmentary vertebrate fossils
found at a site low in the Sangre de Cristo Formation in the
Arkansas River valley of Colorado, Brill (1952) tentatively as-
sessed the age of this formation — in Colorado — as Early Permian
and postulated a pre-Wolfcampian truncation of Upper Pennsyl-
vanian strata in this region. Subsequent collection from Brill’s
site has produced many more disarticulated bones, as well as
fragments of plants and shells of small pelecypods. Recognizable
vertebrates include: palaeoniscoid fish, two or more kinds of lab-
yrinthodont amphibians, a new genus and species of aistopod
amphibian, a new species of the diadectid cotylosaur Desma-
todon, an ophiacodont pelycosaur, probably a sphenacodont pely-
cosaur, and an edaphosaurian pelycosaur referable to Edapho -
saurus aff. E. raymondi. The new aistopod has a mode of costal
articulation that seems to be more primitive than that of known
Pennsylvanian forms. Materials of the new species of Desmatodon
greatly expand our knowledge of this hitherto poorly known
genus and indicate greater primitiveness than seen in D. Holland! ,
described by Case in 1908 on the basis of a fragmentary jaw from
the Conemaugh Series of the eastern United States. Edapho-
saurus aff. E. raymondi corroborates approximate time-equiva-
lence with the Conemaugh Series. At least the lower part of the
Sangre de Cristo Formation in central Colorado seems, there-
fore, to be of Missourian (Late Pennsylvanian) age. This is con-
sonant with the position of the Sangre de Cristo Formation
directly above rocks of Desmoinesian age in this region. This is
the first known occurrence of North American Pennsylvanian
tetrapods west of eastern Kansas, and it promises new insights
into the geographic distribution of Early Permian vertebrates.

Introduction

In a stratigraphic study of the uppermost Paleozoic rocks deposited in the
zeugogeosyncline of Colorado and northern New Mexico — the trough between
the Late Paleozoic San Luis-Uncompahgre and Front Range highlands —
Brill (1952:821) redefined the Sangre de Cristo Formation in the region of
central Colorado “to include the Permo-Pennsylvanian strata that overlie the
Whiskey Creek Pass limestone or its equivalent.” That is, he excluded from
this formation the underlying Veta Pass Limestone Member of earlier authors.
In central Colorado, in the valley of the Arkansas River, the Sangre de Cristo
Formation as redefined is about 8800 feet thick. Temporal relationships with

1Research Associate in Vertebrate Paleontology, Los Angeles County Museum of
Natural History; and Department of Zoology, University of California, Los Angeles,
California 90024.

1

2

Contributions in Science

No. 164

the Sangre de Cristo as exposed in northern New Mexico have remained un-
clear. As Brill pointed out, near Pecos, New Mexico, the Sangre de Cristo
overlies strata of Virgilian (latest Pennsylvanian) age and underlies strata of
Leonardian (Early Permian) age and is probably a time-equivalent of the Abo
Formation of central and southern New Mexico — which is Wolfcampian
(earliest Permian) in age. Brill (1952:822) saw two possibilities for the
Sangre de Cristo of the Arkansas River valley: “Either the lower part of the
Sangre de Cristo formation in . . . Colorado is a nonmarine equivalent of the
upper Pennsylvanian marine strata in northern New Mexico, or the Sangre de
Cristo is mainly Wolfcamp and the upper Pennsylvanian has been truncated
by a pre-Wolfcampian unconformity.” Brill tentatively chose the latter as the
correct explanation, on two lines of evidence: (1) the Sangre de Cristo of the
Arkansas River valley rests on rocks of Desmoinesian (Middle Pennsylvanian)
age, and (2) vertebrate fossils found low in the Sangre de Cristo of this region
seemed to indicate Wolfcampian age. Nevertheless, he remained troubled by
the lack of good lithologic evidence for a pre-Wolfcampian truncation, as
shown by his remark (1952:842) that “nonmarine strata [of the zeugogeosyn-
cline] assigned to the Wolfcamp may be upper Pennsylvanian entirely or in
part.” Baltz (1965:2054) expressed this same uncertainty when he referred to
Brill’s paper by saying: “Early Permian vertebrates have been found in the
Sangre de Cristo Formation, but it has not been established whether the for-
mation also contains rocks of Pennsylvanian age in Colorado.”

In the course of a study of Early Permian vertebrates of the Four Corners
area, I have re-examined the fossils noted by Brill, and I have made further
collections at the site whence they came. Study of the fossils convinces me that
Brill’s first alternative is the correct one, that at least the lower part of the
Sangre de Cristo Formation in central Colorado is a nonmarine equivalent of
Upper Pennsylvanian marine strata in northern New Mexico. More specifi-
cally, these fossils seem to indicate Missourian age. Not only is this interesting
from a stratigraphic point of view, but the fossils shed light on some problems
in the understanding of Late Paleozoic vertebrates. Vertebrate-yielding Penn-
sylvanian deposits are rare anywhere in the United States, and this is particu-
larly true of the West. The Sangre de Cristo site provides our first knowledge of
Pennsylvanian tetrapods west of the now well known Garnett quarry of
eastern Kansas (see Peabody, 1952). Further, the fossils from the Sangre de
Cristo greatly supplement previously meager data on the Pennsylvanian
diadectid cotylosaur Desmatodon, and they provide amplification of our
knowledge of the aistopod amphibians.

The Fossils in General

Brill noted bones of cotylosaurian and pelycosaurian reptiles found in
Interval 300 of his measured section of the Sangre de Cristo Formation near
Howard, in Fremont County, Colorado. There is some slight discrepancy

1969

Upper Pennsylvanian Vertebrates

3

within his paper as to the distance of Interval 300 above the base of the Sangre
de Cristo, but his detailed section and a statement on page 834 of his paper
make it clear that it is about 1450 feet. Following his directions to the fossili-
ferous bed, my crew and I found it in NE14 SW!4 sec. 22, T 49 N, R 10 E,
where it is cut by Rock Creek, a branch of Badger Creek, which latter drains
into the Arkansas River about a mile and a half to the south. Brill’s Interval
300 is described in his detailed section (1952:870) as “Shale dark red below,
greenish gray . . . above, silty; basal bed darker, wedges out to north; a few
thin beds of dark-gray limestone; shale contains ganoid fish scales, bones of
Diadectes sp., and pelycosaurs.’’ Interval 300 is about eight feet thick. The
bones come from the dark, almost black, basal bed, which is about two to
three feet thick at its deepest part — where it is cut by Rock Creek — and wedges
out about 80 feet to the northwest. Most of the bones come from about 1 .5 feet
above the two-foot sandstone of Interval 299. On the southeast side of the
creek, Interval 300 is covered by soil and vegetation.

The fossils collected by Brill are catalogued in the collections of the
United States National Museum as USNM 18602, which consists of part of a
vertebral neural arch definitely of a diadectid cotylosaur, the glenoid portion
of a probably diadectid pectoral girdle, a vertebral centrum possibly from the
caudal region of a diadectid, part of a vertebra of a small pelycosaur, a small
piece of dermal bone of a labyrinthodont amphibian, a scale of a palaeoniscoid
fish, and several other, small scraps of bone. The diadectid remains were
identified as pertaining to Diadectes, a characteristic member of faunal as-
semblages known from Lower Permian rocks of north-central Texas, northern
New Mexico and southeastern Utah (see Romer, 1960 and Vaughn, 1964),
and never, as far as is known, found in the Pennsylvanian. It was on the basis
of this identification that Brill assigned the Sangre de Cristo Formation of
central Colorado to the Wolfcamp.

In the summers of 1966 and 1967, my crew and I excavated many more
fossils from Brill’s site in Interval 300. Large blocks of matrix and bone were
removed to be prepared in the laboratory, where we found that — except for
some aistopod vertebrae and osteoderms, and some palatal and anterior limb
elements of a small labyrinthodont amphibian — no elements were in even
partial articulation with one another although they lay close together. The
bones are all quite dark, and often difficult to distinguish from matrix in the
field; but many of them are excellently preserved although fragile, and can be
associated as elements belonging to a number of species on morphological
grounds even though in most instances it is impossible to relate them with
certainty as parts of the same individual animal.

Besides bones, poorly preserved fragments of plants were found, as were
also shells of small pelecypods. Because of the disarticulated nature of the
remains, it has not been possible to identify all the vertebrates represented, but
the following forms may be recognized: a palaeoniscoid fish; of labyrinthodont

4

Contributions in Science

No. 164

amphibians, one small and at least one large rhachitome; of lepospondylous
amphibians, a new genus and species of aistopod; of cotylosaurian reptiles, a
new species of the diadectid genus Desmatodon’, and of pelycosaurian reptiles,
one ophiacodont, probably one sphenacodont, and an edaphosaur referable to
Edaphosaurus aff. E. raymondi. It is Desmatodon that was mistaken for
Diadectes. The aistopod, the edaphosaur and, especially, the new species of
Desmatodon form the principal bases for re-assessment of the age of at least
the lower part of the Sangre de Cristo Formation in Colorado as Late Pennsyl-
vanian rather than Early Permian.

Abbreviations

The following abbreviations are used throughout the text to refer to
repositories of specimens: AMNH, American Museum of Natural History,
New York; CM, Carnegie Museum, Pittsburgh; FM, Field Museum, Chicago;
UCLA, University of California, Los Angeles; USNM, United States National
Museum, Washington.

Palaeoniscoid Fish

Palaeoniscoid fish are represented by a number of smooth, shiny scales
with a diamond-shaped outline and with bevelled edges by which they articu-
lated with adjacent scales. The largest of these scales are about 3.5 mm long
(between acute apices) and about 2.5 mm wide (between obtuse apices).
These scales and the small pelecypods, as well as the lenticular shape of the
fossiliferous bed, offer evidence for aquatic origin of the deposit.

Labyrinthodont Amphibians

At least two kinds of labyrinthodont amphibians are distinguishable
among the materials from Interval 300, one small rhachitome and one large
one, and isolated teeth may represent a third kind.

The small rhachitome is represented by UCLA VP 1698, which consists
mostly of partially articulated palatal elements but also includes part of a
maxilla with teeth, a few fragments of an anterior limb, and perhaps some
vertebral parts. There is also a partial impression of a lower jaw, and a small
fragment of roofing bone that shows a pattern of fine pitting. About a dozen
teeth can be seen on the fragment of maxilla, but only a few are complete.
They are narrow cones with longitudinal grooves in their basal halves; the
largest is 1.5 mm long. Of the few palatal elements present, the pterygoids are
the best preserved, although twisted out of their natural positions with respect
to one another, and there is also a displaced denticulate element that seems to
be from the lateral palatal series. The left pterygoid is the more complete; it
was about 20 mm long from near the posterior end of the quadrate process to
where it met the vomer. To judge from proportions in a primitive rhachitome

1969

Upper Pennsylvanian Vertebrates

5

Figure 1. A, medial, and B, ventral views of the right palatine bone of a labyrintho-
dont amphibian, UCLA VP 1699. C, lateral view of the right premaxillary bone
(and a bit of the right nasal bone) of a labyrinthodont amphibian, UCLA VP 1700.

such as Dendrerpeton (Carroll, 1967), the total length of the skull may have
been about 30 mm. The pterygoids are finely denticulate in their anterior parts,
and each shows a deep, posteromedially facing socket for the basipterygoid
process of the braincase — the palate must have been mobile. Bowing of the
medial edge of the palatine ramus of the pterygoid indicates fairly wide in-
terpterygoid vacuities. Behind the palate lie some bits of shoulder girdle and
free limb, but these permit no more than the observation that the limbs seem
to have been relatively small. Posterior to this are fragments of which some
may be pieces of thin-walled intercentra. The grooves in the teeth make it
clear that this is a labyrinthodont amphibian; the mobile palatal articulation
and the fairly wide interpterygoid vacuities would make it seem that we are
dealing with an edopoid rhachitome, perhaps a dendrerpetontid. In the lack
of more materials, it is not possible to be more definite.

UCLA VP 1699 (Fig. 1, A, B) is a tusk-bearing right palatine bone of a
large rhachitome. The posterior part of the dentigerous portion is missing, but
the maxillary process seems to be almost complete and this allows an estimate
of total length, from where the palatine met the vomer to its junction with the
ectopterygoid : about 65 mm. The maxillary process extends dorsally and
posteriorly as a stout blade with a roughly scored lateral surface — it is difficult
to find illustrations of the maxillary process, but Bystrow (1939) shows it in
the Lower Triassic capitosaurid Benthosuchus. There is a smooth, posteriorly
sloped shelf on the dorsal surface of the dentigerous portion medial to the
maxillary process. The narial surface is well preserved; the vomerine process
of the palatine bounded the choana medially, and the dentigerous portion
bounded it posteriorly. One tusk is preserved but the tip is missing; what

6

Contributions in Science

No. 164

remains is 1 1.5 mm long. The tusk is circular in section near the tip and almost
circular, with a diameter of 6.0 mm, at the expanded base. The proximal two-
thirds or more of the tusk is fluted by numerous longitudinal grooves. There
are three poorly defined empty sockets behind the preserved tusk and perhaps
the remains of a crushed socket in front, and there were probably more in the
lost posterior part; the sockets all lie within a groove with a thin lateral and a
thick medial wall. The palatine is 13.7 mm wide in the plane of the preserved
tusk and thins to 11.5 mm at the posteriormost preserved socket. The greatest
height of the dentigerous portion of the palatine, near the anterior border of
the preserved tusk, is 13.0 mm. The medial surface of the dentigerous portion
of the palatine shows distinct grooves for articulation with the pterygoid.
These grooves end anteriorly at about the plane of the preserved tusk, which
indicates that there was a fairly well developed interpterygoid vacuity, that the
palatine formed part of the lateral boundary of the vacuity, and that the
pterygoid did not meet the vomer. The slimness of the palatine as seen in
ventral view suggests a narrow snout.

A right premaxillary bone with two teeth, UCLA VP 1700 (Fig. 1, C)
possibly belongs to the same amphibian whose palatine bone has just been
described. A lightly “sculptured” bit of the nasal bone is attached to this pre-
maxilla, but the suture is difficult to follow. The premaxilla is about 40 mm in
length along the ventral margin and is about 15 mm thick from medial to
lateral surface as measured at the anterior border of the naris. The bar in front
of the narial opening is massive, about 22 mm from front to back at the ventral
border of the naris. Below the naris, there is a thin, recessed sheet for the
reception of the maxilla. Disturbingly, this sheet is separated on its medial
side, by a smoothly walled concavity, from a smaller sheet that extends caudad
from near the line of articulation of the premaxilla with its partner; the con-
cavity may have formed the anterior portion of an anterior palatal vacuity —
such openings are known in the narrow-snouted archegosaurid rhachitomes
(see Romer, 1947) but in these they are wholly surrounded by the premaxillae.
Should the concavity be part of a choana rather than a palatal vacuity, its
proximity to the midline would suggest an embolomere rather than a rhachi-
tome. Only the very base of the anterior premaxillary tooth is preserved in
UCLA VP 1700, but this is enough to indicate an enormous tooth. The base is
oval, 12.5 mm long from front to back but only 6.5 mm wide. The second
tooth has a base 7.7 mm long and 5 mm wide. This second tooth is 10 mm
long as preserved; the tip is lacking. The greater length of this tooth is fluted by
coarse longitudinal grooves, but near the tip, where there is a slight bulbous
expansion, the surface is smooth. When this premaxilla was first found, it was
thought that it pertained to the new species of Desmatodon described in this
paper, but this now seems unlikely. For one reason, the better preserved tooth
in the premaxilla is similar to the tusk of the palatine bone described above —
but this is not a cogent point because of the unreliability of comparisons be-

1969

Upper Pennsylvanian Vertebrates

7

tween marginal and palatal teeth. A more compelling argument lies in the
comparison of the premaxillary teeth with isolated diadectid “incisors” un-
doubtedly referable to Desmatodon from Interval 300. There are a half dozen
of these incisors, and they look nothing like the teeth of UCLA VP 1700. The
diadectid incisors (described more fully in a later section) have a fine longi-
tudinal striation that contrasts sharply with the coarse fluting in UCLA VP
1700, and there is no sign in the latter of the excavation of one surface of the
tooth to form the chisel-like crown that is so prominent a feature in the
incisors of Desmatodon — and of Diadectes. Rather, the better preserved tooth
of UCLA VP 1700 looks very much like certain other isolated teeth from
Interval 300; in these, too, the fluting is coarse and there is a slight bulbous
expansion proximal to the tip. In the isolated teeth the tip is strongly recurved
— perhaps this was also true of the teeth of UCLA VP 1700. It is mostly on
the basis of the teeth, then, that I think the premaxilla belongs to the laby-
rinthodont amphibian represented by the palatine bone, but it may also be
noted that the shape of the premaxilla is such that the anterior end of the snout
must have been quite narrow; this would fit the shape of the snout as postulated
on the basis of the palatine.

The entrance of the palatine into the border of the interpterygoid vacuity
indicates that this vacuity was fairly well developed, and this denotes a
rhachitome. But what kind of rhachitome? A search through Romer’s review
of the labyrinthodonts (1947) has not allowed even tentative superfamilial
identification. The best that can be said is that the structure of the palatine
suggests some similarity to the narrow-snouted archegosaurids, but these
rhachitomes are not known below the Permian.

A single neural arch (UCLA VP 1702) of rhachitomous aspect is appro-
priate in size to the palatine bone, as is a fragment of a squamosal bone from
near the border of the otic notch. The squamosal has the typically labyrintho-
dont style of pitting on its external surface, and this may also be seen on other
fragments of dermal bone from Interval 300.

Still other isolated labyrinthine teeth from Interval 300 indicate perhaps
a third kind of labyrinthodont amphibian. The largest of these teeth is about
15 mm long. They are coarsely grooved, are oval in section near the base, and
taper with continuous recurvature to a pointed tip.

Obviously, the nature of the labyrinthodont remains, and my misgivings
as to the association of the premaxilla, do not permit the setting up of new taxa.

NEW GENUS AND SPECIES OF AISTOPOD AMPHIBIANS
Class AMPHIBIA
Subclass LEPOSPONDYLI
Order AISTOPODA
Family OPHIDERPETONTIDAE

8

Contributions in Science

No. 164

Coloraderpeton, new genus

Type species : Coloraderpeton brilli, new species

Diagnosis : The vertebrae of Coloraderpeton differ from those of Ophider-
peton — the only other known genus of the Ophiderpetontidae — in that the
neural spine, instead of being low and bladelike, is tall, has a crenulated dorsal
edge, and does not extend the full length of the vertebra. The ends of the
capitulum and “tuberculum” (costal process) of the rib, instead of being
cupped, are condylar; and the posteromedial process of the rib is short and
rounded, rather than long and pointed.

Coloraderpeton brilli, new species

Holotype : UCLA VP 1689, a dorsal vertebra (Fig. 2, A, B, C, D).

Referred specimens : UCLA VP 1690, proximal part of a rib, preserved
portions slightly displaced along a fracture (Fig. 2, E) ; UCLA VP 1691, three
partially articulated dorsal vertebrae, with many scattered, but some partially
articulated, osteoderms on the same small piece of matrix; UCLA VP 1692,
many disarticulated vertebrae and osteoderms.

Horizon and locality : The holotype and referred specimens were collected
by a field party from the University of California, Los Angeles, in the summer
of 1966, 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 forma-
tion, in NE14 SW14 sec. 22, T 49 N, R 10 E, Fremont County, Colorado. The
holotype and referred specimens were all found close together, within an area
of about a square foot, some distance removed from other bones found at this
site. The foramen for the spinal nerve in the vertebrae, the tetraradiate struc-
ture of the rib, and the fusiform shape of the osteoderms are all typically
aistopod features that preclude any possibility of incorrect assignment. While
it may be that all these specimens are parts of a single individual, this is uncer-
tain— hence the designation of a single dorsal vertebra as the holotype. The
age is Late Pennsylvanian, probably Missourian. The species is named in honor
of Kenneth G. Brill, Jr., who discovered the site.

Diagnosis : The same as for the genus.

Description and discussion : Baird (1964) has published a survey of
aistopod amphibians that has made distinction of Coloraderpeton an easy task.
The Mississippian form he mentions has not yet been described, but he was
able to summarize the features of all other aistopods. According to Baird
(1964:6), all, including the members of the family Phlegethontiidae, have
“Vertebrae with low, blade-like neural spines” that extend “the full length of
the vertebra.” Indeed, Baird includes this feature in his list of ordinal charac-
ters. In view of this, it is possible that the new genus may represent aistopods
different enough to form a new family, but it is otherwise so like Ophiderpeton
in general features that I feel it belongs in the Ophiderpetontidae — monotypic

1969

Upper Pennsylvanian Vertebrates

9

Figure 2. Colorciderpeton brilli, new genus and species: A, right lateral, B, dorsal,
C, ventral, and D, anterior views of the holotypic vertebra, UCLA VP 1689; edges
of adjacent matrix indicated by lines; in B and C, anterior is to the top. E, proximal
part of a rib, UCLA VP 1690; note slight displacement along fracture. F, osteoderm.

up until now. In its similarities to Ophiderpeton, Coloraderpeton can be dis-
tinguished without difficulty from the Phlegethontiidae, a monotypic family,
if Dolichosoma is, as Baird suspects, congeneric with Phlegethontia. Differ-
ences are immediately apparent: in Phlegethontia, the ribs have slender,
flexible shafts and the posteromedial process is weakly developed at best,

10

Contributions in Science

No. 164

whereas the shaft is stout in Coloraderpeton and the posteromedial process,
although not long, is prominent; in Phlegethontia, the armor consists only of
thin gastralia, whereas C. brilli has — as seen fairly well in UCLA VP 1691 —
the well developed “sheathing of fusiform osteoderms which are packed to-
gether in a herringbone pattern” that is characteristic of Ophiderpeton (Baird,
1964:8). An individual osteoderm of C. brilli is illustrated (Fig. 2, F); this
osteoderm is 6.5 mm long.

Caudal vertebrae of C. brilli have not yet been recognized among the
materials. The dorsal vertebrae are so like those of Ophiderpeton in general
that only salient features need be discussed. The centrum of the holotypic
vertebra (Fig. 2, A, B, C, D) is 8.7 mm long as measured along the ventral
midline and has a transverse diameter of 6.0 mm at the anterior end; the
lateralmost points of the transverse processes are 12.5 mm apart. The an-
teroposteriorly short, dorsoventrally tall, and dorsally crenulated neural spine
is seen not only in the holotypic vertebra but also clearly in other vertebrae
included among the referred specimens; there can be no doubt of this feature.
As in Ophiderpeton , the neural canal forms a long tunnel the entire length of
the vertebra. Baird (1964:6) noted that “In Phlegethontia a foramen for the
spinal nerve occurs below the anterior end of the posterior zygapophysis or
slightly further forward, behind the transverse process. This foramen appears
to be present also in Ophiderpeton although it is generally obscured by dorso-
ventral compression or concealed by the dermal armor in the specimens at
hand.” The well preserved, uncrushed specimens of C. brilli would seem to
confirm Baird’s observation with regard to Ophiderpeton; all show the fora-
men— a useful identifying mark for aistopods. These foramina in Coloraderpe-
ton are smaller than those in Phlegethontia as illustrated by McGinnis (1967).

Baird refers to the transverse process in aistopods as a parapophysis and
says that it articulated with only the capitulum of the rib. This is undoubtedly
the mode of articulation in Phlegethontia, but the matter is not as clear in the
Ophiderpetontidae. In Coloraderpeton, at least, the articular surface of the
parapophysis is divided into two distinct facets, which lie one in front of the
other and are separated by a slight vertical ridge. Only one proximal part of a
rib (UCLA VP 1690, Fig. 2, E) of C. brilli has been found, and this was not
in natural articulation with a vertebra; but this rib can be easily fitted into the
parapophysial facets in such a way that the “tuberculum” — better termed cos-
tal process as Baird says — articulates with the anterior facet and the capitulum
articulates with the posterior facet, with the ridge between facets projecting
into the angle between costal process and capitulum. It may be that Colora-
derpeton differs in this way from Ophiderpeton, and perhaps the lesser develop-
ment of the posteromedial process of the rib in Coloraderpeton is related to
such difference; Baird notes that both costal process and capitulum have
cupped heads in Ophiderpeton, but in Coloraderpeton the heads of both are
condylar.

1969

Upper Pennsylvanian Vertebrates

11

It is unfortunate that no cranial parts of Coloraderpeton have yet been
recognized; nevertheless, what is known does supplement our knowledge of
these too little understood limbless amphibians. The only aistopod found above
the Pennsylvanian is a species of Phlegethontia known from the Lower Per-
mian (McGinnis, 1967). Otherwise, the aistopods — including Phlegethontia
— are a typically Carboniferous group. This in itself may support the argument
developed more fully below, that at least the lower part of the Sangre de Cristo
Formation in central Colorado is of Late Pennsylvanian age. There is addi-
tional support in apparent degree of primitiveness. In Coloraderpeton both the
capitulum and the costal process of the rib are condylar, and both seem to have
articulated with the vertebra; in Ophiderpeton, according to Baird (1964),
both the capitulum and costal process are cupped, but only the capitulum
articulated with the vertebra; in Phlegethontia, as may be seen in the illustra-
tions published by McGinnis (1967), only the capitulum is cupped, and as she
says, only the capitulum articulated with the vertebra. In this seeming sequence
of decreasing emphasis on the costal process as an articular part, it would
appear that Coloraderpeton has the most primitive aspect; this sequence would
also be consonant with McGinnis’s opinion that Phlegethontia is more ad-
vanced than Ophiderpeton.

NEW SPECIES OF THE DIADECTID COTYLOSAUR DESMATODON

Class REPTILIA
Subclass ANAPSIDA
Order COTYLOSAURIA
Family DIADECTIDAE
Genus Desmatodon Case, 1908
Desmatodon hesperis, new species

Holotype : UCLA VP 1706, a practically entire left maxillary bone, with
complete dentition (Fig. 3, A, B, C, D).

Referred specimens : UCLA VP 1712, “incisor” tooth (Fig. 3, E, F);
UCLA VP 1713, cheek tooth; UCLA VP 1710, left half of atlantal neural arch;
UCLA VP 1707, dorsal vertebra (Fig. 4, A, B) ; USNM 18602, partial neural
arch of dorsal vertebra; UCLA VP 1708 and 1709, two caudal vertebrae (Fig.
4, C, D); UCLA VP 1716, chevron bone; UCLA VP 1714, right humerus
(Fig. 5, A, B, C, D); UCLA VP 1715, ungual phalanx; UCLA VP 1717,
various elements including teeth, dorsal and caudal vertebrae, fragments of
chevrons and limb bones, and phalangeal elements.

Horizon and locality : The holotype and some of the referred specimens
were collected by a field party from the University of California, Los Angeles,
in the summer of 1966; the rest of the referred specimens were collected in
1967. The site is in Interval 300 of the section of the Sangre de Cristo Forma-
tion measured by Brill (1952:870), about 1450 feet above the base of the

12

Contributions in Science

No. 164

Figure 3. Desmatodon hesperis, new species: A, lateral, B, medial, and D, ventral
views of the holotypic left maxillary bone and dentition, UCLA VP 1706; unshaded
areas in A and B represent matrix; in D, anterior is to the left. C, outline sketch in
medial view of the holotypic maxilla and dentition to indicate, by hachure, areas of
attrition on teeth. E, lateral, and F, posterior views of an “incisor” tooth, UCLA VP
1712.

1969

Upper Pennsylvanian Vertebrates

13

formation, in NE!4 SW14 sec. 22, T 49 N, R 10 E, Fremont County, Colo-
rado. None of the elements was articulated with another. Probably many of
the specimens are from a single individual, but the fact that one cheek tooth
(UCLA VP 1713) is significantly wider than any of the teeth in the holotype
indicates that more than one individual is represented. The age is Late Penn-
sylvanian, probably Missourian.

Diagnosis : The maxillary bone and its dentition in Desmatodon hesperis
may be contrasted with those in D. hollandi, the only other known species of
the genus. In D. hesperis the ventral border of the lingual wall of the maxillary
bone lies farther dorsally than does the ventral border of the labial wall, so that
the bases of the teeth are more exposed on the lingual side; in D. hollandi the
lingual and labial walls extend to about the same level. The central cusps of the
teeth in D. hesperis are more prominent than in D. hollandi, but the lingual
cusps are less well developed. The teeth in D. hesperis have crowns that are
taller — measured on the labial side — than they are wide, whereas in D. hollandi
the crowns are wider than they are tall. In D. hesperis the maxillary teeth —
with the exception of the first and second — are set very close together; in
D. hollandi each tooth is relatively shorter from anterior to posterior surface,
and the space behind each tooth is equal to almost half the anteroposterior
length of that tooth.

Description and discussion : Although the point is not crucial to this dis-
cussion, it may be appropriate at the outset to comment on my placement of
diadectids as cotylosaurian reptiles — the classic classification. Romer (1964)
has recently argued that the diadectids are better regarded as labyrinthodont
amphibians closely allied to the seymouriamorphs. Although I have no doubt
that the relationship to seymouriamorphs is close, many features of the dia-
dectid skeleton indicate achievement of the reptilian structural grade. The
matter is not settled, and I feel obliged to maintain a conservative position until
more definite evidence of an amphibian life history is presented.

In 1908, E. C. Case described bones collected by P. E. Raymond at
Pitcairn, about 15 miles east of Pittsburgh, Pennsylvania, at a level about mid-
way in the Conemaugh Series. Case referred to the formation as the Pittsburgh
Red Shale, but subsequently it has been called the Round Knob Formation (as
in Romer and Price, 1940). From this site, Case noted remains of an eryopid
rhachitome, an edaphosaurian pelycosaur, and a diadectid cotylosaur. The
diadectid is represented by a fragment of a maxilla with four complete teeth
and the root of a fifth, and by a number of chevron bones. Case designated the
fragment of maxilla — CM 1938 — as the holotype of Desmatodon hollandi and
based the genus on this species. Case (1908:236-237) described the dentition
as follows: “The teeth are transversely elongate; the crown is slightly wider
than the root and is also somewhat swollen in the antero-posterior direction.
The outer half of the crown rises gently into a sharp apex from which there is
a sharp descent to the inner half, which is lower than the outer half. The inner

14

Contributions in Science

No. 164

side of the apex is nearly vertical and presents a flat face inwards, this more
prominent on the posterior and largest of the teeth than on the anterior ones.
The surface of the crown is marked with fine lines and the sides of the root
show the same character, but here the lines are coarser. The inner half of the
crown of the anterior and the posterior teeth shows no wear, but on the two in
the middle there are surfaces worn by attrition on both the inner half of the
crown and on the apex.”

Case noted the basic similarity of these teeth to those of the Lower
Permian Diadectes and regarded the dentition of Desmatodon as more primi-
tive. Romer (1952:86) elaborated on this comparison: “The Desmatodon
teeth . . . appear to be built on the fundamental Diadectes pattern, but are far
more archaic than in Diadectes; the most highly developed Desmatodon teeth
are comparable to the least developed molars of Diadectes — those at the front
or back ends of the ‘molar’ batteries. Desmatodon may well be, as its strati-
graphic position suggests, a direct ancestor of Diadectes .” Case illustrated the
holotype (1908, PI. 59, fig. 1) with the teeth pointing upwards and repeated
this illustration soon afterwards (1911), although comparison with Diadectes
would make it seem that his description was of upper teeth. Romer (1952:86),
in a restudy of the holotype, said that “Diadectid ‘molar’ teeth show a pattern
reversed in upper and lower jaws; in the upper teeth the major cusp is lateral in
position, in the lower teeth, medial. Since Case refers to the major cusp in
Desmatodon as lateral, it is obvious that he considered the specimen as per-
taining to the upper jaw. Although this is uncertain (not improbably the speci-
men is the anterior portion of the tooth battery of the left dentary), we retain
his orientation.” Romer is quite correct in his description of the lateral place-
ment of the major (central) cusp in the upper teeth of diadectids, but it is not
at all clear that Case thought that CM 1938 was from the upper jaw — and it is
not from a dentary. In a description of the dentition in Diadectes, Case (1911:
70) said that “The wear of the teeth produced flat faces on the inner cusps of
the lower teeth and in advanced stages a deep triangular groove on the inner
side of the medial cusp. The same effect was produced on the outer half of the
upper teeth.”

Not only is this the exact reverse of the pattern of attrition that I have
observed in Diadectes dentitions at various museums (e.g., AMNH 4349 and
FM UC 675), but it is even contradictory to the evidence presented by Case
himself in a figure (1911, PI. 1, fig. 6) of the right maxilla of AMNH 4349. As
Romer (1952) has made clear, in the upper “molars” of diadectids the major
cusp is lateral, and the attrition is concentrated on the inner (lingual) side.
In the lower teeth, the attrition is concentrated on the labial side (see Welles,
1941). I have removed more of the recalcitrant matrix from CM 1938 and
have exposed, near the base of one of the teeth, a pit of the kind called a “basal
notch” by Welles (1941). These pits are found in Diadectes too, and they
always lie on the lingual side of the teeth, in both upper and lower jaws. This

1969

Upper Pennsylvanian Vertebrates

15

permits the positive statement that the major signs of attrition on the teeth in
CM 1938 occur on the lingual side, and this identifies the holotype of D.
hollandi as part of an upper jaw. Comparison with the holotype of D. hesperis
makes it clear that CM 1938 is from the left side, but this can be adduced as
well from a consideration of the relative positions of the labial and lingual
edges of the maxillary teeth in Diadectes. In this form, as in D. hesperis, the
greatest width of the tooth is not perfectly transverse; the labial edge lies
somewhat anterior to the lingual edge. The teeth in D. hollandi are not per-
fectly transverse either; the angle of implantation indicates that the holotype is
from the left maxilla, and this shows that Case was correct in his judgment that
the larger teeth lie posterior to the smaller. That the holotype of D. hollandi is
a fragment of the left maxilla is fortunate because it allows direct comparison
with the holotype of D. hesperis. Except for differences in the amount of wear
on individual teeth, the teeth of the holotype of D. hesperis fit the description
given by Case for the teeth of D. hollandi; detailed distinctions are given above
in the diagnosis of the former. Because D. hollandi is so poorly known, I prefer
not to revise formally the generic diagnosis; obviously, it would include those
features of the maxillary teeth that override the specific distinctions.

UCLA VP 1706, the holotype of Desmatodon hesperis (Fig. 3, A, B, C,
D), is a left maxillary bone, 77 mm long, with twelve teeth. In Diadectes, as
far as I have been able to learn from the literature and through personal exami-
nation of specimens, the number of maxillary teeth varies from eleven to
thirteen with eleven the usual number. All the teeth except the most posterior
one are excellently preserved; in the most posterior, the crown is fractured and
the distal part displaced forwards. As in Diadectes, the maxilla has its greatest
depth— 19 mm in UCLA VP 1706 — somewhat less than halfway back from
the anterior end. Posterior to this, the gently convex dorsal border decreases
gradually in height. Anteriorly, the thin lateral wall is broken along its dorsal
rim, but it would appear that it did not decrease much in height. The external
surface of the maxilla is bowed into a slight concavity; this surface is lightly
rugose, but it is not “sculptured” as in labyrinthodonts. On the medial side,
about 6 to 7 mm ventrad of the dorsal edge of the lateral wall, the maxilla
thickens abruptly into a longitudinal bar that is convex dorsomedially. The
greatest width of this bar, 15.5 mm, is in the plane between the eighth and ninth
teeth; anterior to this the bar narrows slowly, but posteriorwards the narrow-
ing is rapid. The part of the maxilla anterior to the first tooth is abruptly re-
cessed to present an anteromedially facing concave surface that overlapped
the premaxilla. This surface is continued posteriorly into a notch on the medial
side of the bar above the first tooth. The lower border of the lateral surface of
the maxilla extends farther ventrad than the lower border of the bar on the
medial side, with the effect that the exposed part of the sixth tooth, for example,
is only 8.5 mm long as measured on its labial side but 10.5 mm on the lingual
side.

16

Contributions in Science

No. 164

At the plane between the third and fourth teeth, a narrow channel begins
just above the tooth row and passes posteriorly and dorsally across the medial
and dorsal surfaces of the described bar to fade out at about the plane of the
sixth tooth. Such a channel may be seen in Diadectes maxillae too (AMNH
4349 and FM UC 675 for examples), in which it also has its ventral end at
about the plane between the third and fourth maxillary teeth. In Diadectes the
ventral end of the channel lies immediately in front of the anteriormost part
of the palatal process of the palatine bone. Olson (1947) thinks that the palatal
process formed part of a secondary palate that was completed by a membrane.
The similar placement of the channel in Desmatodon and Diadectes may indi-
cate that a palatal process was present in the former too, although it was
obviously not as tightly bound to the maxillary bone as in Diadectes — -in which
the two bones are practically inseparable. I cannot convince myself of the
presence of the channel in the fragmentary maxilla of D. hollandi, although it
may be represented by a poorly defined groove.

The general features of the maxillary teeth have already been noted in the
diagnosis of D. hesperis and in the discussion of the holotype of D. hollandi,
but some further description is required. As in D. hollandi — and also in Dia-
dectes— the crowns of the teeth of UCLA VP 1706 are marked with light
longitudinal striae; near the sockets the striation is coarser, and there seems
to be a modified bone serving as cementum. Well developed basal notches are
present on the lingual sides of the bases of at least the fourth and sixth teeth.
As may be seen in the illustrations, there are marked gaps separating the
bases of the first, second and third teeth. Posterior to these, adjacent teeth are
set close together. The first two teeth are narrow from side to side and have a
somewhat caniniform aspect: they are more acuminate than the succeeding
teeth, they jut out slightly beyond the general tooth row, and their posterior
edges are broadly concave and rather bladelike. The postero-lingual surfaces
of the first two teeth are excavated in such a way that they approach the shape
of the “incisors” described below. The third and fourth maxillary teeth are
intermediate in structure between the first two and the more posterior teeth.
The third overlaps the fourth on the lingual side, and to a lesser extent the
fourth overlaps the fifth. The fifth through twelfth are readily recognizable as
diadectid “molars” and resemble the teeth in the holotype of D. hollandi
except for differences in proportions already noted. As in D. hollandi, the
labial cusp is not the distinctly set off entity found in Diadectes ; “shoulder”
might be a better word than “cusp.” The lingual cusp is more distinct, but far
from as well developed as in Diadectes. As in D. hollandi, the lingual cusp does
not extend quite as far distally as does the labial cusp. The central cusp in
D. hesperis is longer than in D. hollandi. The difference in general aspect be-
tween the dentitions in the two species might be summed up by saying that the
teeth in D. hesperis tend more toward an overall conical shape, and presumably
this is more primitive than the condition in D. hollandi.

1969

Upper Pennsylvanian Vertebrates

17

The crowns of the third through twelfth teeth in UCLA VP 1706 have
their greater width in the transverse plane. They gradually increase in width
from the third (6.6 mm wide) to the seventh, which has a width of 7.8 mm but
a length from front to back surface of only 5.2 mm. The sixth through tenth
are all almost as wide as the seventh, but the eleventh is only 5.8 mm wide and
the twelfth is narrower than this, although measurement is difficult due to
fracture.

The first, and the third through tenth teeth of UCLA VP 1706 show dis-
tinct wear facets on their lingual sides, on the lingual cusps in the teeth to
which this term is appropriate (Fig. 3, C). Attrition on the lingual side is seen
also in D. hollandi and is the pattern on the maxillary teeth in Diadectes too.
Welles (1941 ) analyzed the motion of the lower jaw in Diadectes as propalinal
and suggested that this animal, instead of being herbivorous as it is usually
regarded, may have been molluscivorous. He based this suggestion partly on
the abundant presence of thin-shelled pelecypods in association with certain
remains of Diadectes. In this light, it is worth noting again that many small
pelecypods were found along with the specimens of D. hes peris at the Interval
300 site.

It is difficult to determine the exact serial positions of the teeth in the
holotype of D. hollandi. In comparison with the holotype of D. hesperis, they
(including the tooth represented only by a root) could conceivably be the third
through seventh, the fourth through eighth, or the fifth through ninth. The
first two possibilities seem unlikely because of the lack of a clearly discernable
channel on the medial side of the maxilla such as is seen in UCLA VP 1706,
and also because of the degree of “molarization” of the first two teeth in CM
1938. It is likely that CM 1938 includes the fifth through eighth maxillary
teeth and the root of the ninth. If this is so, then, unlike the condition in
D. hesperis where the sixth through tenth teeth are almost equal in width, there
would be a slight but steady increase in width from the sixth through ninth
teeth in D. hollandi. The width of the eighth tooth in UCLA VP 1706 is 7.7
mm, and that of the probably corresponding tooth in CM 1938 is 9.5 mm —
note also that the greatest width of the maxilla in this specimen is 18.7 mm as
compared to 15.5 mm in the Sangre de Cristo maxilla. The teeth in D. hollandi
are shorter from anterior to posterior surface — for the probably eighth tooth
in CM 1938 this distance is 4.2 mm, and the corresponding measurement for
the eighth tooth in UCLA VP 1706 is 5.3 mm — but the total length taken up
in the maxilla by the five teeth is the same in both forms, due to the greater
distance between successive teeth in D. hollandi. Details of these comparisons
must, of course, be regarded with caution because of the uncertainty in deter-
mination of the serial positions of the teeth of CM 1938.

An isolated distal part of the crown of a cheek tooth from Interval 300,
UCLA VP 1713, is conspicuously wider- — 10.5 mm — than the widest tooth in
the holotype of D. hesperis. It is difficult to be sure which jaw the crown is

18

Contributions in Science

No. 164

from, but the wear facet is on the shoulder that reaches slightly farther distally
— the reverse of the condition in the maxilla— and this may indicate a lower
tooth. It is not impossible to imagine that the lower cheek teeth might have
been significantly wider than the uppers within the same individual, but this
seems unlikely. This isolated tooth serves as a caution that all the disarticulated
diadectid remains from Interval 300 are probably not from the same individual.

Six isolated “incisor” teeth from Interval 300 are confidently referred to
D. hes peris on the basis of their great similarity to premaxillary and anterior
dentary teeth in Diadectes. The longest and best preserved of these, UCLA
VP 1712, is illustrated (Fig. 3, E, F). The root is lacking, but almost all of the
crown seems to be present, and this is about 18 mm long. The fine longitudinal
striation found in the “molars” is seen here too. The distal two-thirds of the
posterior surface is broadly and concavely excavated so that the tooth is
chisel-shaped in lateral view, and the distal half is gently expanded from side
to side so that the tooth looks spadelike in posterior view. These incisors pro-
vide further confirmation of the correct assignment of Desmatodon to the
Diadectidae.

The left half of an atlantal neural arch, UCLA VP 1710, from Interval
300, is similar to that of FM UR 27, a specimen of Diadectes sideropelicus
from the Wichita Group of north-central Texas, although it is only about two-
thirds as large.

Two almost complete diadectid dorsal vertebrae were found at Interval
300, and there are fragments of others; the most complete, UCLA VP 1707, is
illustrated (Fig. 4, A, B), and it will be seen that this specimen has suffered
some crushing. Perhaps the best way to characterize these vertebrae is to say
that, had they been found as isolated elements in a Lower Permian deposit,
they would almost certainly have been referred to Diadectes— within which
there is considerable variation in details of vertebral structure — and it was
exactly this mistake in identification of the fragmentary neural arch included
in USNM 18602 that led Brill (1952) to his assessment of the age of the
Sangre de Cristo Formation in Colorado as Wolfcampian. UCLA VP 1707 is
about 100 mm high from the bottom of the centrum to the top of the slim
neural spine; the almost horizontal zygapophysial surfaces lie about 35 mm
above the bottom of the centrum; the left posterior zygapophysial buttress is
incomplete, but the distance between the lateral tips of the zygapophyses must
have been about 65 mm. Measured on its posterior face, the centrum is 23 mm
wide and 24 mm deep. The nature of the episphenes is not clear, but
hyposphenes were present on the posterior surface above the neural canal.
Their articular surfaces faced ventrolaterally; they were not twisted to face
upwards as they do in some Diadectes vertebrae, but there are also many
specimens of the Permian form that show the presumably more primitive con-
dition seen here. As in Diadectes, the transverse processes are elongate ridges
that pass anteroventrally from behind the anterior zygapophysis, and the neural

1969

Upper Pennsylvanian Vertebrates

19

Figure 4. Desmatodon hesperis, new species: A, anterior, and B, posterior views
of a dorsal vertebra, UCLA VP 1707. C, lateral view of a caudral vertebra, UCLA
VP 1708. D, lateral view of a somewhat more posterior caudal vertebra, UCLA VP
1709.

spines are quadrangular in section with the angles directed fore, aft, and side-
ways. These vertebrae help demonstrate the truly diadectid nature of
Desmatodon.

A number of caudal vertebrae from Interval 300 were at first difficult to
identify, but I am now convinced that they are of D. hesperis. Two of these,
UCLA VP 1708 and 1709, are illustrated (Fig. 4, C, D). The lack of trans-
verse processes indicates that these are from well behind the region of the
caudal ribs. The striking difference from Diadectes is in the slim neural spine,
which slants strongly backwards, and which in at least some of the more
anterior caudals — such as UCLA VP 1708 — extended over part of the next
posterior vertebra. In Diadectes (see Case, 1911 and Romer, 1944) the spine
is stout and is directed almost straight upwards. Nevertheless, there is a com-
pelling reason for recognition of these caudals as diadectid: traces of (much
reduced) hyposphenes can be seen on the slight ridge beneath the almost
conjoined posterior zygapophyses. The length of the centrum in UCLA VP
1708 is 17 mm, and in 1709 (from a more posterior position), 16 mm. These
vertebrae are narrow in cross-section — the centrum in 1708 is only two-thirds
as wide as it is tall — and the zygapophyses are set very close together. The

20

Contributions in Science

No. 164

anterior zygapophysial surfaces face as much medially as they do dorsally. The
ventral portion of the centrum is bevelled at both the anterior and posterior
faces for the reception of the chevron bones. Actually, except for the slanted
spines, these caudals are not unlike those in Diadectes, in which the zyga-
pophysial surfaces are also inclined and in which the zygapophyses are set
fairly close together — there is an abrupt decrease in the distance between
contralateral zygapophyses within the length of the first sacral vertebra. As-
suming that I have correctly identified the caudal vertebrae from Interval 300,
the tail in Desmatodon must have resembled to a certain degree that in Sey-
mouria, in which the neural spine also slanted backwards to cover part of the
succeeding vertebra. White (1939) regarded the tail of Seymouria as a sign of
a good swimmer. A similar condition of the neural spines may be seen in the
obviously aquatic reptile Mesosaurus. Possibly then, there is some reason to
believe that Desmatodon had somewhat more aquatic habits than did Dia-
dectes, and it may be, but is not clear, that the more “fish-like” condition of the
caudal neural spines in itself denotes greater primitiveness.

Two kinds of chevron bones were taken from Interval 300. One, repre-
sented by a single, incomplete specimen, is quite large and is probably of a
labyrinthodont amphibian. The other kind is represented by an almost com-
plete bone (UCLA VP 1716) and several fragments. These fit the description
given by Case (1908) for diadectid chevrons found in association with the
holotype of D. Holland i. The largest chevron described by Case is 51 mm long.
UCLA VP 1716 is incomplete distally, but it appears to have been at least
that long.

There are a number of partial ribs from Interval 300. Some of these seem
to be appropriate in size to the dorsal vertebrae described above, but none has
a complete head, and identification is uncertain.

Brill’s (1952) collection from Interval 300, USNM 18602, includes a
fragment from the glenoid region of a right scapulocoracoid that I identify,
with some hesitation, as pertaining to D. hesperis. Part of the supraglenoid
buttress is preserved, and this is pierced by the supraglenoid foramen — as in
Diadectes. But not only is this condition common among cotylosaurs, it occurs
in ophiacodont pelycosaurs too and, as I note in the next section, there are
signs of a fairly large ophiacodont from Interval 300. Unfortunately, that part
of the bone where there is a prominent process for the coracoid head of the
triceps muscle in ophiacodonts is not preserved. What is present agrees with
the illustrations of the Diadectes scapulocoracoid published by Case (1911).
The glenoid facet is twisted into a somewhat screwlike shape and extends
ventrally at its front end to overhang the area of the coracoid foramen. The
anteroposterior length of the facet must have been about 45 mm, and this
seems appropriate to the humerus described below.

UCLA VP 1714 (Fig. 5, A, B, C, D), a right humerus from Interval 300,
has lost the deltopectoral crest and the posterior part of the head, as well as

1969

Upper Pennsylvanian Vertebrates

21

Figure 5. Desmatodon hesperis, new species: A, proximal-dorsal, B, distal-dorsal, C, proximal-ventral, and D,
distal-ventraPviews of the right humerus, UCLA VP 1714.

22

Contributions in Science

No. 164

most of the supinator process and the entepicondyle. Nevertheless, it may be
assigned to D. hesperis with complete confidence. As in Diadectes (see Romer,
1956), the basic structure is tetrahedral, with the proximal surfaces set at a
right angle to the distal surfaces; this is especially obvious in distal-dorsal and
distal-ventral views (in the terminology of Romer). The total length of the
humerus as preserved is 96 mm. A detailed description of UCLA VP 1714
would merely repeat the available descriptions of the humerus in Diadectes;
only differences need be stressed. In general aspect, the humerus is somewhat
slimmer than in Diadectes. This appearance is partly due to the fact that the
anterior and posterior parts of the head and most of the supinator process were
broken away, but it is nevertheless clear that there is in D. hesperis at least
something of a shaft between the proximal and the distal parts of the element;
in Diadectes the proximal and distal parts are, so to speak, simply turned on
one another and a shaft is barely recognizable. The head has suffered the most
crushing and is incomplete, but it seems not to have differed in any substantial
way from Diadectes. The ectepicondyle is less prominent than in Diadectes.
Much of the entepicondyle has been lost, and the outer bar of the entepicondy-
lar foramen has been fractured at its proximal end and displaced towards the
shaft, but what is left indicates that this region was similar to what is seen in
Diadectes. In D. hesperis the base of the supinator process lies a little farther
distally — nearer to the capitellum — than in Diadectes , probably in correlation
with the more distinct shaft. What remains of the stump of the supinator
process shows that it was more constricted at its base than in Diadectes, in
which the base is broadly flared along a longitudinal plane.

Case (1911, PI. 11, figs. 4, 5) illustrated a humerus, FM UC 6541, from
the Conemaugh Series near Danville, Illinois. This humerus has had a compli-
cated nomenclatorial history that I need not discuss, but lately it has been
informally reputed to be diadectid, and it probably forms part of the basis for
Romer’s statement (1952:87) that “It is possible that a few fragments of bones
of diadectid appearance from the vertebrate bonebed at Danville, Illinois, and
of the same general late Pennsylvanian age, may pertain to Desmatodon .”
Other fragments from Danville may indeed be diadectid, but I have examined
the humerus and found that, except in characters common to very primitive
cotylosaurs, it is not like that of D. hesperis; it much more likely has limnosce-
loid affinities, and Dr. R. L. Carroll, who has also examined this specimen, has
written to tell me that this is also his opinion.

Four ungual phalanges from Interval 300 are obviously of D. hesperis
because they are so completely like the unguals of Diadectes figured by Case,
and they fit the description (Case, 1911:83) in being “broad and spatulate,
with thin and very rugose edges. . . .” The largest of these, UCLA VP 1715, is
12 mm wide and 14 mm long.

There cannot be any doubt that Desmatodon is related to the Early
Permian Diadectes, and it seems quite obvious on the basis of the dentition that

1969

Upper Pennsylvanian Vertebrates

23

Desmatodon is much the more primitive genus — as Romer (1952) has already
said. The other Early Permian diadectids, including Diasparactus and Phanero-
saurus, are not as well known, but Romer (1956) feels that they form a com-
pact group with Diadectes as apart from Desmatodon. Even without our
knowledge of D. hollandi, the presence of specimens of D. hesperis in Interval
300 would argue strongly for Late Pennsylvanian, rather than Wolfcampian,
age; but the remains of D. hollandi from the Conemaugh Series make this clear
and indicate a position within the Missouri Series for at least the lower part of
the Sangre de Cristo Formation in central Colorado. It has been noted above
that the dentition in D. hesperis may be regarded as somewhat more primitive
than that in D. hollandi, but this does not seem to be sufficient reason to
postulate much greater age.

Romer (1964) has made clear his opinion that diadectids are closely
related to the seymouriamorph labyrinthodont amphibians, and the recent
description of the Lower Permian form Tseajaia (Vaughn, 1964), which seems
in some ways to demonstrate a morphologically intermediate stage, supports
this opinion, although it does not necessarily follow that diadectids must be
regarded as amphibians rather than reptiles. The specimens described in this
paper show that Desmatodon was similar to Diadectes not only basically but
also in many details. The line leading to diadectids must have split away from
the more general seymouriamorph stock at a time well before Conemaugh-
Missouri deposition.

Pelycosaurian Reptiles

In any collection from a quarry where many disarticulated and jumbled
bones are found, it may be expected that some elements will be unidentifiable
below the level of major taxonomic grouping. This is true of pelycosaurian
elements from Interval 300. Assorted items, including teeth, fragments of
jaws, and pointed claws, indicate the presence of several kinds of pelycosaurs,
but except for the recognition of a few elements as ophiacodont, the probably
correct identification of a toothed fragment of maxilla as sphenacodont, and
the fairly detailed identification of an edaphosaur, these items must remain
labelled simply “pelycosaurian” for the present.

A part of a neural arch — UCLA VP 1695 — including the left posterior
zygap°physis and the base of the spine seems quite clearly to be of an ophia-
codont. What remains of the spine shows that it had the bladelike shape char-
acteristic of ophiacodonts, and the neural arch is not laterally excavated as it
is in sphenacodonts. The base of the spine, from anterior to posterior edge, is
about 12.5 mm long; the greatest thickness of the spine is 3.0 mm.

Two partially preserved ilia, a right and left (UCLA VP 1696 and 1697),
are also of ophiacodont aspect and are of a size that may be considered appro-
priate to the neural arch described. Measurements are hard to take on these
specimens; the right one seems to have suffered some post-mortem flattening,

24

Contributions in Science

No. 164

but the left is less altered and in this the neck of the ilium is about 34 mm from
anteriormost to posteriormost point at the level of the supra-acetabular
buttress. These ilia look similar to those of the Lower Permian Ophiacodon
figured by Romer and Price (1940). The posterior parts of the blades are
incomplete, but the anterior parts are fairly well preserved, and these show no
indication of the flaring expansion seen in sphenacodonts and edaphosaurs.
When these ilia were first extracted from the matrix, it was thought that they
might belong to Desmatodon hesperis, but this seems highly unlikely because
the neck is not constricted above the acetabulum as in Diadectes, and prepara-
tion revealed no sign of the external iliac shelf that is so prominent a feature in
Diadectes. Fox and Bowman (1966) have argued convincingly that an external
iliac shelf — indicating the presence of a considerable mass of axial musculature
external to the upper part of the iliac blade — was not necessarily present in all
of the most primitive reptiles. Nevertheless, the conspicuous presence of the
external shelf in Diadectes, and also in the labyrinthodont Seymouria and the
somewhat diadectid-like Tseajaia (Vaughn, 1964), as well as in the primitive
captorhinomorph Limnoscelis, would seem to make it hazardous to postulate,
on the basis of these two disarticulated ilia, such a major shift in pelvic mor-
phology between Late Pennsylvanian and Early Permian diadectids — espe-
cially in view of Romer’s (1956) reasons for regarding the external shelf as a
primitive feature.

In the past, all Pennsylvanian ophiacodonts were almost automatically
assigned to the genus Clepsydrops, but Romer (1961) has shown that Penn-
sylvanian ophiacodonts were probably quite diverse. The neural arch and,
probably, the ilia described above may be identified as ophiacodont, but not
more precisely at this time.

UCLA VP 1704, a small fragment from the anterior region of a right
maxilla, suggests a sphenacodont. There are two teeth; the anterior one is
6 mm long, and the posterior is 1 1 mm long. Each tooth is somewhat flattened
from side to side, is broad at the base, and tapers to a narrow tip. As in most
sphenacodonts, and in contradistinction to other pelycosaurs (see Romer and
Price, 1940), there is a “step” in the maxilla such that the ventral edge of this
bone anterior to the first tooth lies at a conspicuously higher level than does
the rest of the ventral edge. The first tooth arises partly from the region of the
step — this resembles the general condition in Dimetrodon.

A number of parts of neural spines from Interval 300 represent a small
edaphosaurian pelycosaur similar to one known in association with Desmato-
don hollandi. Case (1908) set up the species Naosaurus raymondi on the basis
of a fragment of neural spine— CM 1941 — found at the Pitcairn site near
Pittsburgh, in the middle of the Conemaugh Series, Pennsylvanian System.
Naosaurus was later synonymized with Edaphosaurus (see Romer and Price,
1940). CM 1941 does seem definitely to represent a species of Edaphosaurus,
in that it has the lateral tubercles so characteristic of that genus. This fragment

1969

Upper Pennsylvanian Vertebrates

25

is about 12 mm long and is, as Case (1908:238) says, “oval in section with the
greatest diameter antero-posterior.” The greatest diameter is about 4 mm, and
the least diameter, measured transversely at a “non-tubercular” level, is 3.4
mm. The collection from Interval 300 in the Sangre de Cristo Formation in-
cludes nine fragmentary neural spines closely comparable to CM 1941. In one
of these, UCLA VP 1693, the greatest diameter is from side to side, but others
are thicker from front to back. The transverse diameter taken between tuber-
cles in UCLA VP 1693 is 3.3 mm. The shortest distance between two tubercles
of a side in CM 1941 is about 4 mm, but this fragment is too short to give any
idea of what variation may have existed; among the Sangre de Cristo spine
fragments, which range in length up to 45 mm, the distance between two
tubercles of a side varies from 2.5 to 15 mm. None of the tubercles is complete
in CM 1941 ; in the Sangre de Cristo specimens they vary from about 1 to 5 mm
in length. In some of the Sangre de Cristo spines the tubercles lie nearly oppo-
site one another as in CM 1941, but in others the tubercles are staggered. Some
show a longitudinal groove along one surface between the rows of tubercles.

The Sangre de Cristo spines are quite unlike those of the other known
American Pennsylvanian species of Edaphosaurus, E. ecordi from the Garnett
quarry in the Missouri Series in eastern Kansas (Peabody, 1957) ; the spine in
E. ecordi is very much expanded from front to back in its upper portion, and
the tubercles are only weakly developed. E. raymondi, as Romer and Price
(1940) point out, is not easily separable on morphological grounds from the
tiny E. mirabilis (Stephanian) and somewhat larger E. credneri (Lower Per-
mian) of Europe. The Sangre de Cristo edaphosaur is best referred to as
Edaphosaurus alf. E. raymondi.

Centra of the Sangre de Cristo edaphosaur have been difficult to recog-
nize, but UCLA VP 1694 probably belongs to this form. This centrum has a
broadly rounded ridge along the ventral midline in the style typical of Eda-
phosaurus. The posterior face of the centrum has a transverse diameter of
8.4 mm and a depth of 6.7 mm, but the fragmentary nature of the specimen
makes other measurements unreliable, although it seems to have been about
1 1 mm long.

The presence of this edaphosaur at Interval 300 in the Sangre de Cristo
Formation is of considerable interest in its association with Desmatodon
hesperis, because the holotype of E. raymondi was found, at Pitcairn, in
association with the holotype of Desmatodon hollandi. This supports the
thesis of approximate equivalence in age of the Interval 300 site (probably
Missouri Series) and the Pitcairn site (Conemaugh Series).

The presence of probably the three major kinds of pelycosaurs at Interval
300 is not surprising. All three, ophiacodont, sphenacodont and edaphosaur,
are known from the Garnett quarry of Missourian age in Kansas (Peabody,
1957).

26

Contributions in Science

No. 164

Concluding Remarks

Detailed comments on morphological peculiarities and the probable
phylogenetic positions of the vertebrates known from Interval 300 have been
made under the descriptions of the individual forms and will not be repeated
here. The indications of age presented by the fossils may, however, be
summarized.

The scales of palaeoniscoid fish are interesting as providing, along with
the pelecypods, evidence for aquatic origin of the deposit, but they do not help
in determination of age. Neither do the fragmentary remains of labyrinthodont
amphibians help, although their presence does promise a future source, on an
expanded geographic scale within North America, of greater knowledge of
Late Pennsylvanian members of this group.

The new aistopod, Coloraderpeton brilli, does provide a definite indica-
tion of age. Aistopods, except for an Early Permian species of the apparently
advanced genus Phlegethontia, are typically Carboniferous, and Ophiderpeton,
the nearest known relative of Coloraderpeton as far as I can determine, is a
Middle and Late Pennsylvanian genus (Baird, 1964). The evidence from the
nature of the costal articulation in Coloraderpeton indicates that this genus is
more primitive than Ophiderpeton, and this points to an age at least as old as
Late Pennsylvanian.

The new species of the diadectid cotylosaur Desmatodon, D. hesperis, is
certainly very close to Desmatodon hollandi, which is known only from the
Conemaugh Series of the eastern United States. Detailed differences in denti-
tion between these two species denote somewhat greater primitiveness for
D. hesperis, but this seems not to be enough reason to postulate a great differ-
ence in age between the two; the conservative course is to regard them as
approximately equal in age.

Edaphosaurus aff. E. raymondi provides corroboration of Late Pennsyl-
vanian age, inasmuch as the holotype of E. raymondi was found in association
with the holotype of Desmatodon hollandi about midway in the Conemaugh
Series.

The fossils from Interval 300 thus provide strong evidence for assignment
of at least the lower part of the Sangre de Cristo Formation in central Colorado
to the Missouri Series, which is roughly the western equivalent of the Cone-
maugh Series. This fits perfectly the fact (Brill, 1952) that the Sangre de Cristo
Formation in the Arkansas River valley of Colorado rests on rocks of Des-
moinesian (Middle Pennsylvanian) age; and it makes it vastly preferable to
accept Brill’s (1952:822) alternative that “the lower part of the Sangre de
Cristo formation in . . . Colorado is a nonmarine equivalent of . . . upper
Pennsylvanian marine strata in northern New Mexico,” rather than his tenta-
tive suggestion, based mostly on the quite natural misidentification of frag-
mentary Desmatodon remains as Diadectes, that the Upper Pennsylvanian of
that region had been truncated by a pre-Wolfcampian erosional interval.

1969

Upper Pennsylvanian Vertebrates

27

As already noted, the Sangre de Cristo site provides us with our first
knowledge of Late Pennsylvanian North American tetrapods west of eastern
Kansas. It may be hoped that this knowledge will be extended; there is promise
of additional fossils from Interval 300, and further search of the Sangre de
Cristo Formation in Colorado is needed. Further, the Sangre de Cristo fossils
so far found, although hardly varied enough to be spoken of as a “fauna,” do
indicate that Late Pennsylvanian tetrapod faunas — in similar environments —
were much alike throughout North America; and they will be important in
considerations of the geographic distributions of Early Permian vertebrates
( e.g ., Vaughn, 1966); that is, they provide evidence of “local” derivations of
Early Permian faunas such as those of the Four Corners region.

Acknowledgements

I am indebted to my able field assistants, David Berman, limmy Hara,
Edward Mercurio and lohn Moss; and to Miss Sally Dieckman, who drew the
illustrations for this paper. Thanks are due Dr. E. H. Colbert of the American
Museum of Natural History, Dr. C. Black of the Carnegie Museum, Dr. R.
Zangerl of the Field Museum, and Dr. N. Hotton of the United States National
Museum, for the loan of specimens under their care. I am grateful to the
National Science Foundation for support of this work through grants GB-5104
and GB-7784,

Literature Cited

Baird, Donald. 1964. The ai'stopod amphibians surveyed. Breviora, (Harvard,
Univ., Mus. Comp. Zool.) 206: 1-17.

Baltz, E. H. 1965. Stratigraphy and history of Raton basin and notes on San Luis
basin, Colorado-New Mexico. Amer. Assoc. Petrol. Geol., Bull., 49: 2041-2075.

Brill, K. G., Jr. 1952. Stratigraphy in the Permo-Pennsylvanian zeugogeosyncline
of Colorado and northern New Mexico. Geol. Soc. Amer., Bull., 63: 809-880.

Bystrow, A. P. 1939. Blutgefassystem der Labyrinthodonten (Gefasse des Kopfes).
Acta Zool., 20: 125-155.

Carroll, R. L. 1967. Labyrinthodonts from the Joggins Formation. J. Paleont.,
41: 111-142.

Case, E. C. 1908. Description of vertebrate fossils from the vicinity of Pittsburgh,
Pennsylvania. Pittsburgh, Carnegie Mus., Ann., 4: 234-241.

. 1911. A revision of the Cotylosauria of North America. Carnegie Inst.

Washington, Publ. 145: 1-122.

Fox, R. C., and M. C. Bowman. 1966. Osteology and relationships of Capto-
rhinus aguti (Cope) (Reptilia: Captorhinomorphia). Univ. Kans. Paleont.
Contrib., Vertebrata, 11: 1-79.

McGinnis, H. J. 1967. The osteology of Phlegethontia, a Carboniferous and Perm-
ian ai’stopod amphibian. Univ. Calif. Publ. Geol. Sci., 71: 1-49.

Olson, E. C. 1947. The family Diadectidae and its bearing on the classification of
reptiles. Fieldiana: Geol., 11(1): 1-53.

Peabody, F. E. 1952. Petrolacosaurus kansensis Lane, a Pennsylvanian reptile from
Kansas. Univ. Kans. Paleont. Contrib., Vertebrata, 1: 1-41.

28

Contributions in Science

No. 164

. 1957. Pennsylvanian reptiles of Garnett, Kansas: edaphosaurs. J. Paleont.,

31: 947=949.

Romer, A. S. 1944. The Permian cotylosaur Diadectes tenuitectus. Amer. J. Sci.,
242: 139-144.

— =. 1947. Review of the Labyrinthodontia. Harvard, Univ., Mus. Comp. Zool.,

Bull., 99: 1-368.

— . 1952. Late Pennsylvanian and Early Permian vertebrates of the Pittsburgh-

West Virginia region. Pittsburgh, Carnegie Mus., Ann., 33: 47-113.

— -. 1956. Osteology of the Reptiles. Univ. Chicago Press, 772 pp.

-. 1960. The vertebrate fauna of the New Mexico Permian. New Mexico Geol.

Soc. Guidebook of Rio Chama Country, 11th Field Conf., p. 48-54.

— ■ — — . 1961. A large ophiacodont pelycosaur from the Pennsylvanian of the
Pittsburgh region. Breviora (Harvard, Univ., Mus. Comp. Zool.), 144: 1-7.

— . 1964. Diadectes an amphibian? Copeia, 1964: 718-719.

, and L. I. Price. 1940. Review of the Pelycosauria. Geol. Soc. Amer., Spec.

Paper 28: 1-538.

Vaughn, P. P. 1964. Vertebrates from the Organ Rock Shale of the Cutler Group,
Permian of Monument Valley and vicinity, Utah and Arizona. J. Paleont.,
38: 567-583.

— . 1966. Comparison of the Early Permian vertebrate faunas of the Four

Corners region and north-central Texas. Los Angeles County Mus. Nat. Hist.,
Contrib. in Sci., 105: 1-13.

Welles, S. P. 1941. The mandible of a diadectid cotylosaur. Univ. Calif. Publ.
Geol. Sci., 25: 423-432.

White, T. E. 1939. Osteology of Seymouria baylorensis Broili. Harvard, Univ., Mus.
Comp. Zool., Bull., 85: 325-403.

Accepted for publication February 17, 1969

LOS

ANGELES

CONTRIBUTIONS

COUNTY

MUSEUM

IN SCIENCE

SMBER 165

June 26, 1969

7 73

Li BC®

HABITS AND HABITATS OF CERTAIN
SOUTH AMERICAN JAYS

By John William Hardy

Los Angeles County Museum of Natural History • Exposition Park
Los Angeles, California 90007

CONTRIBUTIONS IN SCIENCE is a series of miscellaneous technical papers
in the fields of Biology, Geology and Anthropology, published at irregular intervals
by the Los Angeles County Museum of Natural History. 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 Robert J. Lavenberg, Los Angeles County Museum of Natural
History, 900 Exposition Boulevard, Los Angeles, California 90007.

Robert J. Lavenberg
Managing Editor

INSTRUCTIONS FOR AUTHORS

Manuscripts for the LOS ANGELES COUNTY MUSEUM, CONTRIBU-
TIONS 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. Al-
though priority will be given to manuscripts by staff members, or to papers dealing
largely with specimens in the Museum’s collections, 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 Edi-
torial Committee. They may also be subject to review by competent specialists out-
side 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.

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. Alterations or changes in the manuscript after galley proof will
be billed to the author. Unless specifically requested, page proof will not 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.

Dorothy M. Halmos
Editor

HABITS AND HABITATS OF CERTAIN
SOUTH AMERICAN JAYS

By John William Hardy1

Abstract: Notes are given on eight species of jays, consist-
ing of Aphelocoma viridicyanci (two races), Cyanocorax chrys-
ops, mystacalis, cyanopogon, affinis, cristatellus, violaceus, and
caeruleus. The information is taken largely from the author’s
personal observations made in July and August, 1965, in Vene-
zuela, Brazil, and Peru. Habits discussed include feeding, flock-
ing, vocal, care of young, and behavioral relationship to habitat.

Habitats discussed include Andean cloud forest (Aphelocoma) ,
subtropical evergreen forest (C. chrysops ), mesquite woodland,
cactus steppe woodland, and deciduous forest (C. mystacalis) ,
tropical evergreen forest (C. affinis and C. caeruleus) , “islands”
or riparian growths of gallery forest (C. violaceus and C. cyano-
pogon), and “cerradao” (C. cristatellus) .

Introduction

For most species of birds of tropical latitudes, the habitat and ecological
relationships of each species to its environment are poorly known. In July and
early August of 1965, I made a reconnaissance of selected areas of South
America that allowed me to visit the habitats of nine species of jays. I photo-
graphed and studied the habitats, attempted to assess the ecological vicissitudes
to which avian types such as jays might be expected to be responsive, or to be
controlled by in the habitat, and observed the behavior and recorded the voices
of six of the nine species.

Vocal repertoire will be discussed by me within the context of a taxonomic
revision of the New World jays. This revised taxonomy is used herein, the
genus Cyanolyca is submerged in Aphelocoma, and Psilorhinus is submerged
in Cyanocorax.

Species Accounts

Aphelocoma viridicyana

Like most of its congeners, this species is an inhabitant of montane forests.
Richards (1964: 364) divided the northern Andean forests into three zones.
Of these, the lower montana does not differ much in aspect and flora from
the rain forest of the Amazon lowlands, and extends from 1200 to 1500 meters
in elevation. Above this lie the two zones inhabited by A. viridicyana. These
are the transitional “medio yungas” (1800-2000 m) and the “ceja de la
montana,” the vegetation of which differs markedly from that of the lower
zones. The “ceja” begins around 2000 m and extends to 4000 m, giving way

Moore Laboratory of Zoology, Occidental College, Los Angeles, California 90041

1

2

Contributions in Science

No. 165

to an Alpine region consisting of open paramo or steppe. The “medio yungas”
is luxuriant tropico-temperate cloud forest which closely resembles the Mexi-
can cloud forests. It has a mixed flora, including cinchona ( Cinchona suc-
cirubra ), bamboos, tree ferns, and rank epiphytic growth. Composition of the
“ceja” is similar to that of the temperate elements of the “medio yungas” but
the aspect of the forest is different. Richards (1964: 364) terms the “ceja” an
Elfin Woodland or mist forest “. . . dominated by low growing trees with
gnarled and twisted trunks abundantly overgrown with epiphytic bryophytes;
the foliage is predominantly small, coriaceous and entire . . .” This forest re-
minds me superficially of the heavier growth of coastal chaparral in California.
There are openings in it that include grassy clearings and rock outcroppings
only partially obscured by shrubby vegetation.

I encountered the dark purplish race, A. v. meridana, in the “medio
yungas” below Paramo Zumbador along the road toward Queniquea. This
locality is approximately 40 km NE of San Cristobal, Venezuela, at about
2700 m elevation. In this area I saw no evidence of a well-developed “ceja,” and
the “medio yungas” on the leeward slopes gave way rather sharply to Alpine
conditions on windward slopes, high exposed ridges, and peaks.

In a ravine where there was dense herbaceous undergrowth, bamboo, and
tree ferns beneath a forest canopy approximately 20 m in height, I observed
three or four of these jays foraging in a very loose group from 5 to 10 m up in
the sub-canopy and second story trees. Their behavior in this situation was
much like that of Aphelocoma coerulescens except that they were more vocal.
The birds foraged in typical corvid fashion, peering into crevices, pecking into
epiphytes, and holding food under a foot while feeding. They remained in
sight for about 15 minutes and did not seem shy or take much notice of me
(about 30 m away) . Other avian species that I observed here were wood-wrens
( Henicorhina leucophrys ) and flower piercers ( Diglossa sp.), but the severe
weather conditions kept bird activity to a minimum during my four hour stay.

In Peru, I visited the habitat of Aphelocoma viridicyana turcosa with the
late Dr. Walter Markl, who had encountered small flocks of up to 5 birds on
several occasions above the village of Canchaque on the road leading up from
the Pacific Coast to Huamachuco, at around 330 m elevation. The forest here
is clearly the “ceja de la montana” described by Richards (1963: 363-4). On
these Pacific slopes, the “ceja” occurs upward from approximately 2500 m. A
dominant tree here is Polylepis spp. (Rosaceae), in which Dr. Markl has often
seen the jay foraging.

Other bird species inhabiting the area, according to Markl (letter, 10
November 1965), include Turdus fuscater and Colaptes rupicola.

Cyanocorax chrysops

This species, with three geographic races, has the widest range of any of
the strictly South American jays. As a consequence, although it apparently is

1969

South American Jays

3

restricted throughout its range to the heaviest forest available, it exists in
several types of forest. These include the islands of “mato” and gallery forest
of tropical nature that dot the scrub savannah (both “pampas” and “cerrado”),
the more nearly unbroken tropical rain forest south of the central Amazon
River (the race diesingi ), and the temperate rain forest of the interior of
southern Brazil (the race chrysops).

Wetmore (1926: 365) has briefly noted the habitat of the two races
chrysops and tucumanus in northern Argentina, where the former is found
together with Cyanocorax cyanomelas , while Gyldenstolpe (1945: 236) ob-
served that it was an inhabitant of “isolated forest islands on the Pampas.”
Eckelberry (1963: 341 ) briefly notes the forest habitat of chrysops in Misiones
Province, Argentina.

I examined the habitat of C. chrysops 30 km NE of Tres Lagoas, on
Fazenda Beya-flor (Hummingbird Farm), state of Matto Grosso, Brazil, where
it is sympatric with C . cristatellus, and 55 km NW of Piricicaba at Hacienda
Barriera Rico, owned by Dr. Jose Carlos Reisde Malgahaes. The habitat there
consists of dense islands of the tropical “mato” in the rolling “cerrado” and
similar forest and woodlands along streams. The vegetation has been cut over
and most of the forest is second growth. There are many dwellings and small
farm clearings, often planted with fruit trees. According to natives, the jays
often emerge from the forest in small groups early in the morning and late in
the evening to pillage fruit. We spent several hours visiting groves of such
trees seeing only two jays for a few moments, though Camargo insists that they
are common here.

Other bird species which I observed here were Syrigma sibilatrix, Buteo
magnirostris, Gampsonyx swainsonii, Ara arauana, Brotogeris chiriri, Rham-
phastos toco, Casiornis rufa, Platypsaris rufus, Thamnophilus dioliatus,
Pitangus suphuratus, Myiarchus ferox, Elaenia parvirostra , Stelgidopteryx
ruficollis, Pygochelidon cyanoleuca, Mitnus saturninus, Parula pitiayumi,
Thraupis palmarum, T. sayaca, Hemithraupis guira, and Nemosia pileata.

Near Hacienda Barriera Rico, a remarkable area of virgin forest has
been set aside as a wildlife preserve. One access road leads from the Hacienda
a mile into the forest and thence into the interior as a one lane trail suitable for
jeep travel. At the terminus of the road, a permanent observation tower about
4 m high had been erected. Here it was possible to observe C. chrysops at close
range, as they periodically came to a bird feeder.

The forest here is subtropical broadleaved, and evergreen, and reminds
me greatly of the northern cloud forest at Rancho del Cielo, Tamaulipas,
Mexico. The resemblance is suggested by the lack of tree ferns, the prevalence
of relatively straight tall trees, rather than those with tortuous growth form,
and the existence of a rather open canopy that promotes a second story of
smaller trees at between 3 and 10 meters. There are few really large trees and
many of perhaps one-half to one meter diameter at breast height. The second

4

Contributions in Science

No. 165

story forms such an effective light screen that ground level herbaceous growth
is slight.

We visited this locality in late evening of 29 July and early morning of
30 July. Both days were heavily overcast, with intermittent rain and tempera-
tures in the 70’s (F.). Both days bird activity was low, probably due to the
rain, but some birds were highly vocal. Trogon rufus called incessantly, as did
a dove ( Leptotilat ) and a robin (probably Turdus albicollis) . Occasionally, a
band of monkeys ( Cebus sp.) were vocal, and small birds sang in the under-
growth ( Dysithamnus spp. was tentatively identified) . No jays were seen in the
evening hours, but a loose flock of perhaps 20 were observed intermittently for
an hour and a half (7:00-8:30) in the morning. The flock was highly vocal as
it moved loosely through the upper strata of the forest. This species, like
C. dickey i and C. yncas, has a large repertoire of vocalizations; in a highly
active flock, such as I observed, a large variety of these are uttered in seemingly
haphazard order, or without reference to immediately obvious environmental
circumstances. Especially prominent calls, however, were several types of
rapid-fire chatter, and a mellow call very much like the plaintive voice of the
Laughing Falcon (Her pet other es cachinnans) . Near the feeding station, the
birds descended very warily, usually one at a time toward the food cache. As a
bird neared the feeder, it would stop abruptly at intervals, call loudly, jerk its
tail upward, peer intently at the blind and the food, and then proceed, until
finally it reached its goal. Ordinarily it then seized a kernal of corn and hastily
retreated to concealment on a low limb nearby but not out in the open. Occa-
sionally a bird would feed at the food source momentarily. In the observation
period, the flock withdrew and returned to the feeder area several times, and
when they left the vicinity, they travelled far enough away in the forest that
their calls could be heard only faintly or not at all. There was no evidence of
breeding activity, adults caring for young, or recently independent young
birds in the flock. At this latitude, it would seem that breeding occurs some-
time between October and February and that in July the birds are the least
active in reproductive behavior.

Cyanocorax cyanopogon

C. cyanopogon is broadly sympatric with its congeneric but distant rela-
tive C. cristatellus in the tree savannah “cerradao” of the tableland of Brazil.
The two are segregated ecologically since cyanopogon remains consistently in
gallery forests along streams and the islands of “mato” (Fig. 1A) in poorly
drained or swampy areas that everywhere dot the “cerrado.”

I examined the habitat of C. cyanopogon in the vicinity of Brasilia, al-
though I did not see the bird. The forest is evergreen, the trees range from
15 to 60 meters in height, and the undergrowth of smaller broadleafed trees,
palms, shrubs, vines, and herbaceous growth is dense. Thus although the domi-

1969

South American Jays

5

nant trees resemble lowland rain forest types, the dense understories clearly
are in contrast to rain forest.

Local farmers informed me that small bands of the jays often emerged
from the forest in early morning and late evening hours to feed on the fruit of
orchard trees. They firmly distinguished between cristate 1 1 us as “Grallia,” and
cyanopogon as “Urraca,” and seemed to consider only the latter as invaders
of the orchard.

Cyanocorax mystacalis

C. mystacalis is the only species of the tribe to have reached the Pacific
lowlands across the Andes. Its presence there, I believe, is the result of a single
pioneering colonization from the Amazonian side of the mountains that prob-
ably occurred when the ancestral form, now replaced in the region of origin
of the assemblage by the chrysops-cyanopogon populations, occurred all over
the Amazon valley south of the main river itself. The range of C. mystacalis
extends from lowlands to 2500 meters (Taczanowski, 1884: 399) in extreme
southern Ecuador and northern Peru, but, like a successful avian pioneer, the
bird displays a wide and efficient adaptability to the several habitats within its
relatively small and inhospitable geographic range. It is abundant to common
in all but one of the natural habitats extensively available; like our North
American jays but unlike many of its congeners, it thrives in the vicinity of
man and is found in residential sections of villages nesting in trees in yards, as
well as in orchard groves.

The natural habitats frequented by this species, all within Chapman’s
(1926: 37) equatorial arid zone, include arid mesquite woodland ( Prosopis
juliflora, Fig. IB), cactus steppe woodland (Fig. 2A), riparian broadleaved
tropical evergreen woodland, and deciduous forest (Fig. 2B). Only the treeless
creosote bush and bare desert areas are unsuitable. The mesquite habitat is
the most widespread of the chosen habitats and supports most of the popula-
tion of this jay. It persists as islands within cultivated areas and along dry river
courses (Fig. 3A). Cactus steppe is found mainly in the littoral zone and often
is abruptly contiguous to or intermingled with mangrove thickets along tortu-
ous shallow bays and inlets near the mouths of streams. Jays can thus often
be heard calling seemingly from mangrove growth, when they are in fact in
islands of cactus woodland surrounded by mangrove. Cactus steppe is charac-
terized by an organ pipe cactus (probably both N eoraimondia sp. and Espostea
lanata), and by the peculiarly “swollen” tree Bombax discolor, as well as by
other spiny vegetation. The deciduous forest (Fig. 2B) of the foothills and
mountain slopes is the most humid part of Chapman’s Equatorial Arid Zone.
He describes it as a “transition area connecting the luxuriant forest of the
Andes . . . with the desert. . . .” Ferreyra’s (1960) presentation of the major
aspects of this forest may be summarized as follows: it is a deciduous forest
with Bombax discolor a prominent tree. The deciduous tree species include

6

Contributions in Science

No. 165

B

Figure 1. A, View across open savannah of the “cerrado” toward an island of dense
evergreen forest habitat of Cyanocorax cycinopogon, Parque Nacional, near Brasilia,
Brazil; B, Mesquite ( Prosopis juliflora ) woodland in open parklike stand, forming
principal habitat of Cyanocorax mytitacalis, northwestern Peru, near Piura.

1969

South American Jays

7

B

Figure 2. A, Cactus steppe woodland habitat of Cyanocorax mystacalis, northwestern
coastal area of Peru, near Tumbes; B, Leaf-shedding forest habitat of Cyanocorax
mystacalis, Peruvian Andes, northwestern Peru.

8

Contributions in Science

No. 165

Bursera graveolens and Loxopterygium huasango, abundantly clothed with
epiphytes of which the principal one is Tillandsia usneoides. The two cacti
mentioned above are also present. Although Chapman states that the inland
limit of this forest separating deserts from typical Andean vegetation is the
base of the Andes, the lower foothills and their valleys also bear leaf-shedding
forest. Trees in this zone reach heights of 20 to 40 meters and bear fairly dense
foliage only in the short rainy season which lasts from January or February
through March or April (Chapman). Chapman says rainfall is “comparatively
low” and practically non-existent in the dry season when most of the trees lose
their leaves.

North American observers would be reminded somewhat of the taller
stands of tropical deciduous forest that blanket lowlands and foothills in
southwestern Mexico, but epiphytic growth in the leaf-shedding forest is much
greater, and there are more tall trees.

I examined the habitat of and observed C. mystacalis in the following
places: Hacienda San Jacinto, 12 km SW Piura, Depto. Piura, Peru; Porto
Pizaro, near Tumbes, Depto. Tumbes; and below 1000 m in foothills along
the road leading from the Pacific lowlands toward Canchaque in the Andes.

At the first locality, the jays were very common in parklike islands of
mesquite remaining within the vast irrigated cotton fields. As the photograph
shows (Fig. IB), there is practically no vegetation in these woodlands except
for the trees. This is partially due to intensive grazing by cattle. Where damp
ground exists, a dry shrubby vegetation abruptly changes to luxuriant broad-
leaved evergreen forest in a very narrow riparian strip. In the vicinity of human
dwellings, there are planted stands of various broadleaved trees, including
Ficus sp., that also exist in parklike stands. Although the jays are most abun-
dant in the mesquite woods, they also are found in these other nearby vege-
tational complexes.

I did not observe the jays in close association with other bird species, but
other species that were prominent in the same areas included Mimus longi-
caudatus (an unusually abundant bird in the mesquite), Falco sparverius,
Crotophaga sulcirostris, Ensifera ensifera, V eniliornis collonotus, Lepidocolap-
tes souleyetii, Furnarius leucopus pallidus, Elaenia leucospodia, T odirostrum
cinereum, Thryothorous super ciliaris, and Polioptila plumbea (mostly in the
mesquite woodland). I observed only one interaction, on 8 August 1965, be-
tween the jays and other birds. On that occasion two of a loose group of about
five jays were seen poking their heads into an opening which they apparently
had made in the mud nest chamber of Furnarius leucopus. The jays were ex-
tracting and eating the eggs of the ovenbirds, while the latter perched nearby
and chattered nervously but did not approach the marauders.

The jays seen were in loose, non breeding foraging groups usually varying
in size from four to about ten birds. Often lone birds or pairs were seen. My
impression was that this species is somewhat less social than its close relative

1969

South American Jays

9

C. dickeyi or C. morio. Individuals often strayed loosely from the bounds and
activities of the flock, reacted toward predators or humans independently, and
dispersed in twos and threes in different directions while being observed.

According to Markl, the jay is common in the cactus steppe and un-
common but regular in the deciduous forest. Taczanowski (1884: 399) states
that at Tumbes the jays remain in thickets at the edge of the stream, in flocks
composed of several pairs, but on two occasions we saw them ranging through
open cactus steppe several hundred meters from the water.

Cyanocorax cristatellus

Much of central Brazil is a rolling tableland from 150 to 900 m above sea
level. This is the “cerrado,” the vegetation of which is largely short trees and
a savannah grass — the “cerradao” — which has been described and illustrated
by Rizzini and Heringer (1962) and is the habitat (Fig. 3B) of the Curly-
crested Jay. Although to the casual observer the gnarled short trees between
1 and 8 m in height might seem at first to comprise no more than half a dozen
species, they in fact include approximately twenty-five very common forms
and an equal number of somewhat less common species. Most are semi-
deciduous or evergreen. Many have coriaceous leaves, rough bark, and tortu-
ous limb form. In the rather frequent swampy areas and along streams, the
“cerradao” gives way to heavy tropical gallery forest (Fig. 1A). Flere some
of the same tree species, Lafoensia pacari, for example, have entirely dif-
ferent growth form and reach great heights. Many species of trees on the
“cerrado” bear fruit that probably provides food for the jays as well as for
many other birds, as pointed out by Dr. Heringer. The following is a brief
annotated list of the trees, in systematic order (Willis, 1966). Anona crassi-
fiora (Annonaceae) — large fruit, sold in market. Possibly avian food; Bow-
dichia virgilioides (Leguminosae) — furniture wood; Dalbergia violaceus
(Leguminosae) — violet wood; Hymenaea stilbocarpa (Leguminosae) — fruit
provides food for humans; Machaerium opacum (Leguminosae) — important
only for its wood; Schlerolobium paniculatum (Leguminosae) — wood for
fence posts; Stryphnodendron barbatimao (Leguminosae) — small but typical
tree of the “cerrado.” Known for its tannin content; Byrsonima muricie (Mal-
pighiaceae) — bears fruit eaten by humans; Qualea parvifiora ( Vochysiaceae)
— one of the most typical flowering trees of the “cerrado;” Ouratea ochnacea
(Ochnaceae) — a prominently flowering and fruiting species that provides food
for birds; Caryocar brasiliensis (Caryocaraceae) — fruit for man, possibly
birds. Souari or butternut; Didymopanax macrocarpum (Araliaceae) —
“Fruto de Tucan,” provides food for birds; Lafoensia pacari (Lythryaceae) —
scrub form in the “cerrado,” large tree in gallery forest; Pouteria vamiflora
(Sapotaceae) — bears a succulent fruit; Aspidosperma tomentosa (Apocyna-
ceae) — this plus another in the same genus are fairly common, and produce

10

Contributions in Science

No. 165

a milky fluid characteristic of the family; Tocoyena brasiliensis (Rubiaceae) —
one of the most prominent trees of the “cerrado”; small palm or palmetto,
genus and species unknown.

In the area of Brasilia, the most common grass of the savannah is Aristida
pallescens (Fig. 3B). It grows to an average height of about 1 m and forms a
dense rather diffuse ground cover broken only by occasional areas of nearly
bare reddish soil. The rainy season in the “cerrado” lasts from November
through March and total rainfall varies from 100 to 1600 mm per year. During
the dry season (April through September) many of the trees lose most of their
leaves and regain them when rains begin in October. This is also the major
time of flowering.

Jays which I saw on 21 July about 15 miles from Brasilia in the Parque
Nacional and on 28 July about 15 miles from Tres Lagoas, Matto Grosso, were
obviously non-breeding (Dr. Camargo informs me that the breeding season is
generally from November through February). Both times the birds were in
flocks of 8 to 12, the members of which kept close together and engaged in
coordinated flying, foraging, and resting. They were easily attracted to squeak-
ing sounds which I made and would approach closely in the trees and give their
alarm cries. Like C. violaceus, the Curly-crested Jay has a simple vocal reper-
toire that consists largely of a loud jeering call. Conversational notes are low
throaty croaks, barely audible more than a few feet away. Its long wings and
concomitant strong flight reflect its adaptation to the wide open habitat which
it prefers. North American observers would be reminded much of the Nut-
cracker, Nucifraga Columbiana, similarly proportioned and adapted, as well
as to the Pinon Jay, Gymnorhinus cyanocephalus.

At least in the non-breeding season, this species seems to move about in
nomadic bands. In contrast to groups of its forest dwelling congeners, a flock
frightened may take wing and fly straight away from an observer, landing after
a flight of perhaps one-quarter mile and subsequently continuing in such rela-
tively long flights completely out of the area where originally observed.

Nothing seems to have been published on the nesting habits of this jay.
From the unity of the flocks that I observed in the non-breeding season, plus
experience with other species, it seems likely that this species will prove highly
social in its breeding habits — like Aphelocoma ultramarina, Cyanocorax morio
and C. dickeyi.

On only one occasion did we observe Curly-crested Jays frequenting
other than “cerradao.” At the Matto Grosso locality, we surprised a group of
perhaps six birds just within the edge of an island of dense gallery forest along
a stream. This occurred at about noon. The weather was very warm and the
birds had seemingly retired to the shaded forest to rest, for we detected no
movement or calling until we disturbed them. They were reluctant to move,
called only a few times, and flew not into the forest but along the edge to rest
again a quarter mile away where we later disturbed them again.

1969

South American Jays

11

Cyanocorax affinis

Sturgis (1928: 438) and Phelps and Phelps (1963: 258) agree that this
species is an inhabitant principally of lowland humid forest. Chapman (1917:
636) however, stated that in Colombia it is also prevalent in the more arid,
seasonally green, woodlands of the tropical zone. I examined the jay and its
habitat near La Fria, state of Jaure, Venezuela, in the lowlands at the base of
the Andes. There the forest has been much replaced by agriculture, and the
rural human population is large. There are frequent patches of only partially
disturbed, mature, climax humid tropical forest, however, with trees ranging
to taller than 30 m and little understory (there is a subcanopy of trees) except
at the edges where the dense forest border gives way abruptly to cultivated
fields.

I followed a pair of adults that were feeding nearly grown young. No
other adults or young were seen or heard in this tract of forest. The birds stayed
mostly in the underpart of the canopy and in the subcanopy. Slud (1964: 281 )
however, found them in Costa Rica in “noisy bands the members of which
keep low in the trees on or near the ground.” He makes no mention of what
time of year his observations were made.

Adults observed by me gave the characteristic alarm and social call of the
species, which Slud (ibid) has accurately translated as “pyou” or “kyoop.”
This first notified me of their presence as they flew furtively and weakly into
my vicinity (Slud, ibid, also mentions their poor flight). Once more they
changed positions in the forest and again uttered this call while moving through
the trees.

Cyanocorax affinis is of low sociality in the nesting season, members of
pairs caring exclusively for their young. It occurs in small flocks (Slud’s
“bands”) in the non-breeding phases of its cycle. It is a forest dweller, weak
in flight, and it is not restricted to one layer of the forest, frequenting levels
from ground to canopy.

Cyanocorax violaceus

This species is an inhabitant of the tropical evergreen forest of lowlands
and foothills, but does not require large continuous tracts of such vegetation.
Therefore, it inhabits not only the relatively unbroken forests of the Amazon
lowlands but also the islands of tropical selva and riparian gallery forests
within the savannahs or “llanos” of Venezuela (Fig. 4A) . Many of these islands
of forest are but a few acres in area, while others are several square miles in
size. The Violaceus Jay of the north coastal area of Venezuela (C. v. pallidus )
is restricted to the narrow coastal belt of tropical forest in the vicinity of
Barcelona and Piritu, in the state of Anzoategui (Phelps and Phelps, 1963:
257).

William Phelps, Sr. (personal communication) considers the nominant

12

Contributions in Science

No. 165

B

Figure 3. A, Mesquite habitat of Cyanocorax mystacalis, along dry water course,
northwestern Peru, near Piura; B, Short tree-savannah growth of the “cerrado,” near
Brasilia, habitat of Cyanocorax cristatellus.

W

m

1969

South American Jays

13

Figure 4. A, View across the grassland to gallery forest in the “llanos,” habitat of
Cyanocorax violaceus, near Barinas, Venezuela; B, From a clearing, viewing the edge
of the humid evergreen forest habitat of Cyanocorax caeruleus, southeastern Brazil.

14

Contributions in Science

No. 165

race of this species to be most at home in the llanos. He directed me to such
habitat in the state of Barinas, Venezuela. Six miles southwest of Barinas, I
encountered a mixed flock of jays and two species of oropendolas (probably
Psarocolinus decumanus and Cacicus cela ) traveling and foraging together in
forest within 100 feet of the highway. The flock consisted of about 6 to 8 jays
and about 12 to 15 oropendolas, all but about two or three of which were
thought to be P. decumanus. The jays were a closely knit, highly social group
that included two fully grown young which were still dependent on the adults
for food. The young gave begging calls, flapped their wings, and on several
occasions I saw adults feed them. I determined the nature of the food in only
one instance, when I saw one bird devour a small fig. The bird held the morsel
securely between his toes and tore off and swallowed a large piece of it. When
I made squeaking alarm noises, the adult jays became excited and gave alarm
cries as they bobbed and lifted their tails. They also approached the edge of the
forest and descended from the canopy to positions about 20 feet from me. I
withdrew slightly and watched them continue to forage and feed young. After
about 25 minutes, the mixed flock withdrew farther into the forest and then
flew across the open “llano” to another island of forest.

I was impressed with the general similarity of this jay to the Brown Jay
{Cyanocorax morio ) in its appearance, demeanor, and voice. Individuals bear
a short but prominent crest which they often erect, but otherwise their form is
like that of the Brown Jay. The coordinated actions of the birds as a flock and
the simple vocal repertoire are much like those of C. morio, and also suggestive
of Aphelocoma ultramarina. C. morio and A. ultramarina are highly social
nesters (see Skutch, 1935; also, Hardy, 1961: 62-63). The flock of Violaceus
Jays including young being fed by flock members suggests a high degree of
sociality in this species.

Cyanocorax caeruleus

The typical habitat of this species is in the rugged hills and on the narrow
coastal plain of southern Brazil, and adjacent parts of Argentina and Paraguay.
These hills and intersecting narrow valleys are clothed with a luxuriant,
humid, tropical and subtropical forest (Fig. 4B) . There are many epiphytes and
tree ferns. To North Americans familiar with vegetational complexes of
southern Mexico, the aspect of this forest is much like the more mesic tropical
deciduous forest in the rainy season, as for example in Oaxaca. The forest also
reminds me of cloud forest because of the tree ferns. A prominent tree in the
forest is a palm Acromia sclerocarpa, called Indaia. The fruit of this palm is
said by Camargo (personal communication) to be a prominent food of the jay.

I visited this habitat in foothills of the Serra do Paranapiacaba between
Jacupiranga and Eldorado, approximately 150 km south of Sao Paulo, state of
Sao Paulo, with Camargo. We were not able to observe well any jays. In a day-
long search we encountered two birds at dawn. They called twice, a down-

1969

South American Jays

15

wardly inflected peer, peer (something like the Brown Jay, Cyanocorax
morio), and we got a glimpse of them as they flew off down a ravine into dense
forest. They had been perched in a grove of palm trees along a path on a ridge
above the forest bordering a stream.

Birds were abundant and active in early morning hours, but our atten-
tions were so thoroughly directed in the brief period we had to finding the
jays that we identified only a few other species. Among these were: Pipro-
morpha rufiventris, Tangara cyanocephala, and Tachyphonus cristatus, the
last two of which were very common.

SUMARIO

Habitos y habitats de ocho especies de grajos (Corvidae) de America del
Sur, basados, en parte, en observaciones hechas en el campo por el autor, en
Venezuela, Brasil y Peru, durante Julio y Agosto de 1965. A phelocoma viri-
dicyana habita los bosques en montanas altas y selvas de los Andes. La sub-
especie meridana se encuentra en bosques exuberantes nublados tropical-
templados en Venezuela, en donde fue observada buscando alimento en
pequenas bandadas, en la region immediatamente debajo de la copa de los
arboles y en el “segundo nivel” de los bosques. La subespecie turcosa vive en
la selva dominada por Polylepis o “ceja de montana,” sobre los 2500 m. Esta
constituye una vegetacion baja, mas clara que la del habitat de A. v. meridana.
En la parte sur de Brasil Cyanocorax chrysops se encuentra en el bosque sub-
tropical “siempre-verde,” en donde se mueve en bandadas de por lo menos 20
pajaros, alimentandose entre el piso del bosque y la copa de los arboles. La
especie C. mystacalis, que esta limitada a las tierras bajas y pie de las montanas
al Oeste de los Andes en Peru, es menos exclusiva en la seleccion del habitat,
siendo caracteristica de los bosques de mesquite ( Prosopis ), estepas de cactus,
y bosques dehiscentes del pie de las montanas. Aparentemente esta especie no
es tan social como la anterior, pues durante la epoca en que no esta en repro-
duction aparece en grupos de 4 a 10 pajaros. C. cyanopogon del centro de
Brasil y C. caeruleus del sur de Brasil viven en el bosques humedo tropical. El
habitat de la primera especie, por lo general, se limita a parches ailados en las
sierras altas del centro de Brasil, mientras que el de la segunda es, general-
mente una serie de parches continuos o de naturaleza riberena. En el mismo
tipo basico de bosque en los llanos de Venezuela se encuentra C. violaceus.
Esta es una especie altamente social aun durante la epoca de reproduction, tal
y como lo prueba el hecho de que pajaros jovenes mezclados en bandadas de
adultos son protejidos por varies pajaros. Al norte de los Andes, en las tierras
bajas de Venezuela, el bosque tropical humedo es el habitat caracteristico de C.
affinis, la cual se ha observado ser poco social; una pareja joven que buscaba
alimento no estaba acompanada por otros adultos, ni otros adultos fueron
vistos en la misma area del bosque. C. cristatellus es quiza el grajo Cyanocorax
mas aberrante en America del Sur; una especie con alas bastante largas, esta

16

Contributions in Science

No. 165

bien adaptada a su existencia nomadica en las tupida sabanas de arbustos en
las planicies de Brasil. Esta especie parece ser moderadomente social en grupos
de 8 a 12 pajaros.

Literature Cited

Chapman, F. M. 1917. The distribution of bird-life in Colombia; a contribution to a
biological survey of South America. Amer. Mus. Nat. Hist., Bull., 36: 1-729.

— -. 1926. The distribution of bird-life in Ecuador. Amer. Mus. Nat. Hist., Bull.,

55: 1-784.

Eckelberry, Don R. 1963. A note on the jays of northern Argentina. Wilson Bull.,
75: 341.

Ferreyra, R. 1960. Algunos aspectos fitogeograficos del Peru. Inst, de Geogr., Univ.
Nac. Mayor de San Marcos, Publ., Ser. 1, Mon. y Ensayos Geogr., 3: 41-88.

Hardy, J. W. 1961. Studies in behavior and phylogeny of certain New World jays
(Garrulinae). Kansas Univ. Sci. Bull., 42: 13-149.

. 1964. Behavior, habitat, and relationship of jays of the genus Cyanolyca, in

press. A taxonomic revision of the New World jays. Condor, in press. Adams
Ctr., Ecol. Studies, Occas. Pap., 11: 1-14.

Phelps, W. H., and W. H. Phelps, Jr. 1963. Lista de las aves de Venezuela con su
distribucion. Vol. 1, part 2. Passeriformes. Socied. Venez. de Cien. Nat., Bol.,
24 (104-5): 1-479.

Richards, P. W. 1964 (revised printing). The tropical rain forest. Cambridge, Eng-
land, University Press. 450 pp.

Rizzini, C. T., and E. P. Heringer. 1962. Preliminares acerca das fora§oes vegetais
e do reflorestamento no Brasil central. Rio de Janeiro, Brazil, Min. da Agr. Ser.
de Informa?ao Agricola. 79 pp.

Skutch, A. F. 1935. Helpers at the nest. Auk, 52: 257-273.

Slud, P. 1964. The birds of Costa Rica. Amer. Mus. Nat. Hist., Bull., 128: 1-430.

Sturgis, B. B. 1928. Field book of birds of the Panama Canal Zone. New York.
466 pp.

Taczanowski, W. 1884. Ornithologie du Perou. Rennes. Vol. 2, 556 pp.

Wetmore, A. 1926. Observations on the birds of Argentina, Paraguay, Uruguay,
and Chile. U. S. Natl. Mus., Bull., 133: 1-448.

Willis, J. C., and H. K. A. Shaw. 1967. A dictionary of the flowering plants and
ferns. 7th ed. Cambridge, England, University Press. 1214 pp. and appendix.

Accepted for publication February 14, 1969.

LOS

ANGELES

COUNTY

MUSEUM

CONTRIBUTIONS
IN SCIENCE

Ijmber 166

June 26, 1969

\ 07.73

i 2 L

EARLY PERMIAN VERTEBRATES FROM
SOUTHERN NEW MEXICO AND THEIR
PALEOZOOGEOGRAPHIC SIGNIFICANCE

By Peter Paul Vaughn

|

Los Angeles County Museum of Natural History

Los Angeles, California 90007

Exposition Park

i!

CONTRIBUTIONS IN SCIENCE is a series of miscellaneous technical papers
in the fields of Biology, Geology and Anthropology, published at irregular intervals
by the Los Angeles County Museum of Natural History. 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 Robert J. Lavenberg, Los Angeles County Museum of Natural
History, 900 Exposition Boulevard, Los Angeles, California 90007.

Robert J. Lavenberg
Managing Editor

INSTRUCTIONS FOR AUTHORS

Manuscripts for the LOS ANGELES COUNTY MUSEUM, CONTRIBU-
TIONS 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. Al-
though priority will be given to manuscripts by staff members, or to papers dealing
largely with specimens in the Museum’s collections, 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 Edi-
torial Committee. They may also be subject to review by competent specialists out-
side 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.

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 wilLbe returned
after the manuscript has been published.

PROOF. — Authors will be sent galley proof which should be corrected and
returned promptly. Alterations or changes in the manuscript after galley proof will
be billed to the author. Unless specifically requested, page proof will not 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.

Dorothy M. Halmos
Editor

EARLY PERMIAN VERTEBRATES FROM
SOUTHERN NEW MEXICO AND THEIR
PALEOZOOGEOGRAPHIC SIGNIFICANCE

By Peter Paul Vaughn1

Abstract: A rhipidistian crossopterygian fish and the long-
spined pelycosaur Dimetrodon are reported for the first time
from the Lower Permian of New Mexico. The rhipidistian comes
from a band that seems to pass from a nodular layer in the upper-
most part of the Laborcita Formation southward into stream-
channel deposits in the lowermost part of the Abo Formation,
along the Sacramento Mountains escarpment in Otero County.
In the same band are found remains of a pleuracanth shark, the
primitive fish Acanthodes sp., palaeoniscoid fishes, an edopid
labyrinthodont amphibian similar to Edops, Platyhystrix cf.
P. rugosus and another dissorophid labyrinthodont, and pelyco-
saurs referable to Ophiacodon, Sphenacodon and Edaphosaurus.
Dimetrodon is represented by vertebral parts from the Abo For-
mation in the Caballo Mountains in Sierra County. This dis-
covery of a rhipidistian and Dimetrodon in southern New
Mexico, in the region of the “mean shoreline of the Wolfcampian
seas” plotted by Kottlowski (1963), seems to partially verify the
prediction that these vertebrates are to be found only as mem-
bers of faunas that lived near the borders of persistent seaways,
in “eu-deltaic” situations. Although these two forms are known
also from southeastern Utah, northern Oklahoma and north-
central Texas, they remain unreported from northern New
Mexico despite the collection of a large and varied fauna from
that area. The Wolfcampian vertebrates of northern New Mexico
seem to have lived under “somewhat more upland” conditions.
These analyses are consonant with paleogeographic reconstruc-
tions of positive elements and seaways. Evidence is reviewed to
show that the “Midcontinental seaway” could have formed only
a partially effective obstacle to distribution of vertebrates be-
tween the Four Corners and the Midcontinent. An analysis of
Permian floras by Read and Mamay (1964) corroborates evi-
dence drawn from vertebrates that drier conditions in the Early
Permian set in earlier in the Four Corners than in the Midconti-
nent. It is suggested that the extensive system of Early Permian
positive elements in Colorado and New Mexico formed the back-
bone of an upland partial obstacle to distribution at least as effec-
tive as the “Midcontinental seaway.”

1Research Associate in Vertebrate Paleontology, Los Angeles County Museum of
Natural History; and Department of Zoology, University of California, Los Angeles,
California 90024.

1

Contributions in Science

No. 166

2

Introduction

In 1966, 1 published a comparison of the Early Permian vertebrate faunas
of the Four Corners region and north-central Texas. That paper was prompted
by new finds and by the prevalent opinion that many of the differences between
the well known faunas of north-central Texas and northern New Mexico were
to be explained by the presence of a water barrier between these areas. I pointed
out that new discoveries of vertebrates in the Cutler Group in southeastern
Utah, including a rhipidistian crossopterygian fish and the long-spined pely-
cosaurian reptile Dimetrodon, cast doubt on the long-term effectiveness of
such a barrier, and noted that the presence of these animals in southeastern
Utah indicates special resemblance to the Early Permian fauna of north-
central Texas. I suggested that the resemblance is due to environmental simi-
larity in that both north-central Texas and southeastern Utah were deltaic
regions in Early Permian time, near the borders of persistent seaways, whereas
the absence of certain forms — including rhipidistians and Dimetrodon — in
northern New Mexico could perhaps be accounted for by the “somewhat more
upland” condition of the latter area in Early Permian time.

For Wolfcampian (earliest Permian) time, then, I strove to draw a dis-
tinction between what may be called, for lack of better designations, by the
guarded terms “eu-deltaic” and “somewhat more upland” vertebrate faunas.
Naturally, the northern New Mexico deposits — in the Cutler Formation in Rio
Arriba County and the Abo Formation in Sandoval County — do not represent
truly upland conditions inasmuch as they were, after all, formed in basins of
sedimentation; the term “somewhat more upland” is used here simply to indi-
cate regions relatively farther removed from borders of persistent seaways.
What was attempted was analysis, albeit crude, of environmental “preferences”
of certain Wolfcampian forms based on their known distributions as correlated
with data on positions of highlands — or at least positive areas — and seaways.
The remarkable fact is that the method, however crude, did seem to help
explain distributions. Correlation with paleogeographic reconstructions based
on non-faunal grounds seems to otfer a useful aid to our understanding of
distributional patterns of ancient vertebrates, and it may offer direct indices to
the habitats of particular forms. It is unhappily true that such studies must, by
their very nature, rely to some degree on negative evidence for non-presence of
faunal elements, and many conclusions must be regarded as tentative. Never-
theless, continuation of such efforts seems worthwhile in an attempt better to
understand distributions of Late Paleozoic vertebrates, on a continental scale,
in terms of environment, as far as possible without recourse to analogy with
modern forms.

The four faunal elements that I listed in 1966 as common to southeastern
Utah and north-central Texas — but unreported from northern New Mexico —
are: the rhipidistian Ectosteorhcichis, the nectridean amphibian Diplocaulus,
the labyrinthodont amphibian Seymouria, and the pelycosaur Dimetrodon.
Since that time, it has been learned that the rhipidistian found in Utah is really

1969

Permian Vertebrates from New Mexico

3

not Ectosteorhachis, but a new form called Lohsania (Thomson and Vaughn,
1968), that is only the second known Permian rhipidistian; nevertheless, no
rhipidistian at all had been known from the New Mexico Permian, and the
presence of such fishes in both Texas and Utah still seems to indicate special
resemblance. The discovery of Dimetrodon in Utah removed what was per-
haps the most dramatic difference between the Early Permian vertebrate
faunas of the Four Corners and the Midcontinent, but until now Dimetrodon
has remained unreported from New Mexico.

To test the hypothesis that these four elements are to be found in “eu-
deltaic” as opposed to “somewhat more upland” deposits, my field party went
southward in New Mexico in the summers of 1966, 1967 and 1968, to prospect
along the “ ‘mean’ shoreline of the Wolfcampian seas” plotted by Kottlowski
(1963:57, and Fig. 12). Good exposures along this “mean shoreline” were
found in the Caballo Mountains in Sierra County and along the western flank
of the Sacramento Mountains in Otero County. Figure 2 indicates both these
areas. They lie, roughly, along the transitional belt between the red-beds facies
of the Abo Formation to the North and the limestone facies of the Hueco to
the South. The figure, based on maps published by McKee, Oriel, et al. ( 1967) ,
gives a less detailed boundary for the “Orogrande seaway” separating the
Caballo Mountains and Sacramento Mountains localities than may be found
in Kottlowski’s paper, because it is a larger-scale map intended to serve a
broader-ranging discussion below.

Among other items, readily identifiable remains of a rhipidistian and of
Dimetrodon were found, and this is the first report of either of these forms
from New Mexico. Although Diplocaulus and Seymouria have not (yet) been
found in southern New Mexico, the presence of a rhipidistian and of Dimetro-
don would seem to provide at least partial verification of the prediction.

Lower Permian Vertebrates from the Sacramento Escarpment

Otte (1959) and Pray (1961) have presented thorough studies of the
uppermost Paleozoic strata exposed along the Sacramento Mountains escarp-
ment in Otero County, New Mexico. Pray’s maps cover the area east and south-
east of Alamogordo, and Otte’s cover the area east of Tularosa and south-
eastward to merge with Pray’s area near the town of La Luz. Particularly im-
portant are the Laborcita and Abo Formations. Because of the interfingering
relationships between these two formations, it is best to quote from Otte ( 1959:
96) at length:

“In the central part of the Sacramento Mountains and the southeastern part of
the map area, the Abo formation overlies with sharp angular unconformity
strata of Pennsylvanian and Mississippian age. The area to the west and north-
west was one of essentially continuous deposition from late Pennsylvanian
through early Permian time, with no major unconformity separating the de-
posits. Gradual emergence of the area and retreat of the marine waters toward

4

Contributions in Science

No. 166

the west and northwest caused interfingering of the uppermost Laborcita and
lowermost Abo strata in the area north of Domingo Canyon . . . The base of
the Abo formation occurs 200 feet stratigraphically above the upper contact
of the Bursum formation as mapped by Pray (1952 [a reference to Pray’s then
unpublished dissertation]).”

Otte named the Laborcita Formation and defined it to include a lower portion
previously known as the Bursum Formation plus approximately 200 feet of
strata previously assigned to the lower part of the Abo Formation. Otte’s
general description (1959:95) of this formation is as follows:

“The Laborcita formation is about 500 feet thick in the southeastern part of
the area . . . The thickness increases to about 1 ,000 feet . . . toward the north-
west. The lithologic and faunal characteristics of the sediments show that
abrupt lateral transitions from open-marine conditions, in the northwest and
west, to terrestrial flood-plain environments, in the southeast and east, occurred
repeatedly within a distance of a few miles . . . On the basis of fusulinid
identifications, the Laborcita formation is very late Virgilian and early Wolf-
campian in age. The uppermost 250 feet of the Laborcita formation near
Tularosa is in part the time equivalent of the lowermost Abo beds toward the
south and east. The Pennsylvanian-Permian boundary, which by earlier strati-
graphers was taken at the base of the Abo formation, occurs 90 feet above the
base of the Laborcita formation, as determined on the basis of fusulinids.”

Recently, Steiner and Williams (1968) have restudied the Laborcita fusulinid
fauna and have come to the conclusion that all of the Laborcita Formation is
Wolfcampian in age. There is some possibility of confusion in reading the
papers published by Otte and Pray. Otte referred to Pray’s work while it was
still in the form of an unpublished dissertation. Although Pray’s paper bears a
lower bulletin number, in the same series, than Otte’s, Pray’s paper was pub-
lished later (1961) than Otte’s (1959). This is important in a choice between
the designations Bursum and Laborcita. Although Pray mapped the Bursum
Formation along the Sacramento Mountains escarpment, he preferred Otte’s
broader designation, Laborcita, as shown by his remark (Pray, 1961:91):
“Otte . . . named the section between the top of the Fresnal group and his Abo
contact the Laborcita formation. This term is clearly preferable to the use of
Bursum in the Sacramento Mountains area; however, as the lower horizon was
used as the base of the Abo in preparing the writer’s geologic map and sections
of the area, the term Bursum is used in this report.” The present paper follows
Otte’s nomenclature.

Figure 1, based on Otte’s maps, shows the relationship between the
Laborcita and Abo Formations in much simplified style — traces of faults,
Pennsylvanian formations, Tertiary intrusives, and covering Quaternary de-

1969

Permian Vertebrates from New Mexico

5

6

Contributions in Science

No. 166

posits are omitted. In the area north of Tularosa, Otte drew the Laborcita-Abo
boundary at the top of his marker bed 55 (a limestone); south of this, in
approximately the area between Tularosa and Domingo Canyons, he drew the
base of the Abo lower down, at the bottom of his marker bed 53 (a conglomer-
ate); still farther south, beginning between Domingo and Laborcita Canyons,
the base of the Abo is dropped to the bottom of marker bed 49 (a conglomer-
ate). These considerations are pertinent to the present report because the
vertebrate fossils found by my field party come from both the Laborcita and
Abo Formations — but all from about the same horizon.

Figure 1 also shows the general areas in which vertebrates were found;
these are indicated by filled-in circles. For ease in reference, they will be called:
the “locality north of Tularosa” (in the upper left-hand corner of the figure),
the “Tularosa Canyon locality,” the “locality south of Tularosa Canyon”
(almost due east of Tularosa), the “locality north of Domingo Canyon,” the
“Domingo Canyon locality,” the “Laborcita Canyon localities,” the “locality
between Laborcita and Cottonwood Canyons,” and the “Cottonwood Canyon
localities.” The northern two — the locality north of Tularosa and the Tularosa
Canyon locality — are in the uppermost part of the Laborcita Formation; all
the others are in the lowermost part of the Abo Formation.

At the Tularosa Canyon locality, in NE14 sec. 21, T. 14 S, R. 10 E, the
vertebrates are found in nodules weathered from a green-gray shale several
feet thick that lies about 40 feet below marker bed 55, about 50 feet below
the top of the Laborcita Formation as Otte draws the boundary in this region.
The nodules are of various shapes from spheres to dumbbells, but mostly they
are biscuit-shaped. There is a great range of sizes, but most of the biscuits are
about 10 to 15 cm in greatest length. Some of the nodules can be traced into
irregular calcareous masses within the shale. It is evident that the nodules are
concretions formed in place, that is, they were not secondarily deposited. The
bedding planes, with very low-angle crossbedding, pass horizontally through
the nodules into the shale. Polished sections reveal no concentric structure
except for a dark, unoxidized core whose border parallels the weathered outer
surface. Most of the sediment is very fine-grained, but there are lenses of
slightly coarser clastic materials in which most of the larger bones occur. The
nodules are readily attacked by weak acids; most of the fossils were prepared
in this way.

The most common remains, by far, are of the acanthodian fish Acanthodes
sp., and scales of this form have been recovered from both the nodules and the
shale, although most of the specimens at hand are in nodules because of the
ease with which these can be collected. The fossils may lie anywhere within
the nodules but, naturally, most of the collection consists of nodules with bones
exposed at the surface. Scales of the acanthodian occur singly and also in
various sized patches in which they are articulated with one another — -the
largest patch found is about 5 by 7 cm. Many pectoral spines, a scapula, and

1969

Permian Vertebrates from New Mexico

7

other acanthodian elements have also been found and these, as well as the
scales in both their gross and microscopic structure, correspond in every way
with descriptions of these elements in Acanthodes. One of the pectoral spines,
UCLA VP 17 18, 2 is 54 mm long. To my knowledge, remains referable to
Acanthodes have been reported from only two other sites in the Lower Permian
of the western hemisphere: from the Dunkard Group of West Virginia
(Romer, 1952) and from the Wichita Group of north-central Texas (Dunkle
and Mamay, 1956). The presence of this fish provides no definite indication of
the kind of waters in which the deposit was formed; according to Zangerl and
Richardson (1963), at least the Pennsylvanian species of Acanthodes with
which they dealt lived in waters of a wide range of salinity.

Besides Acanthodes, the nodules from the Tularosa Canyon locality have
yielded numerous partial stems of Calamites, scales of palaeoniscoid fishes, a
vertebra of the pelycosaurian reptile Sphenacodon cf. S. ferox — a form known
also from northern New Mexico — and, what is most pertinent to this report,
scales and a partial dermal skull bone with the cosmoid microstructure typical
of rhipidistian crossopterygian fishes.

Prior to this, a fragmentary Lower Permian rhipidistian would have been
almost automatically referred to Ectosteorhachis, but, as mentioned earlier, a
second genus of Lower Permian rhipidistians is now known on the basis of
remains from the Cutler Group of southeastern Utah (Thomson and Vaughn,
1968). Although the vertebrae in this new rhipidistian, Lohsania, are quite
different from those in Ectosteorhachis, no difference in the structure of the
scales has yet been noted; hence, in the absence of other than dermal parts, it
seems unwise to attempt at this time even generic identification of the
rhipidistian from southern New Mexico. Nevertheless, this discovery is inter-
esting inasmuch as it is the first record of a rhipidistian from New Mexico, and
it is noteworthy that it was found in the region of Kottlowski’s (1963) “mean
shoreline of the Wolfcampian seas.”

A small collection of similar nodules and intermodular shale was made
at the locality north of Tularosa, in NE14 sec. 31, T. 13 S, R. 10 E, not far
above a porphyry sill. The distance below marker bed 55 is about 50 feet; this
is greater than the interval at the Tularosa Canyon locality, but Otte points out
that the Laborcita Formation thickens considerably toward the Northwest,
and this is well shown in his diagram of stratigraphic sections (1959: PI. 4).
Thus it is likely that the nodular shale here lies at the same horizon as at the
Tularosa Canyon site. Fossils from this locality include acanthodian scales, a
rhipidistian scale, and a tooth of a pleuracanth shark — a common fresh-water
component of Early Permian vertebrate faunas.

The fossiliferous layer of the Tularosa Canyon locality is easily recog-

2The abbreviation “UCLA VP” stands for the vertebrate paleontological collection
of the University of California, Los Angeles.

8

Contributions in Science

No. 166

nized again at the locality south of Tularosa Canyon, in SW14 sec. 22, T. 14 S,
R. 10 E. Acanthodian scales are as abundant as at the Tularosa Canyon
locality. Calamites is present, and there is an impression of a partial leaf,
probably of a seed fern. Palaeoniscoid scales are not uncommon, and there is
also a fragmentary palaeoniscoid jaw in which the small enamel cap can be
seen on some of the teeth. A plane within one nodule shows a “splatter” of
palaeoniscoid scales similar to remains that have been interpreted as gastric
residues by Zangerl and Richardson ( 1963) . There are well preserved rhipidis-
tian scales; for future reference, one of these has been catalogued as UCLA
VP 1720. There are also some fragments of large tetrapod bones, but these
cannot be identified. Although this locality is only about a mile south-south-
east of the Tularosa Canyon locality (uppermost Laborcita Formation), it lies
within the area in which Otte (1959) drops the base of the Abo Formation
from the top of marker bed 55 down about 135 feet to the bottom of marker
bed 53, and it is thus technically in the lowermost part of the Abo Formation —
as are also all the localities to the South.

The fossiliferous layer of the above localities cannot be continuously
traced southward, due to extensive Quaternary cover. The next place south in
which bone was found is the locality north of Domingo Canyon, in about the
center of sec. 36, T. 14 S, R. 10 E. All that was found here is a small fragment
of tuberculated bone, probably of a labyrinthodont amphibian, in a green-gray
sandstone.

We searched at various horizons farther south, but all the productive
localities lie within one band about 50 feet thick. This band is generally in the
form of a sandwich with a sandstone-conglomerate above, a coarse conglomer-
ate with quartzite pebbles below, and mudstone with some nodular limestone
between. As measured in Laborcita and Cottonwood Canyons, the lower
conglomerate lies about 225 feet above the base of marker bed 49 — which,
according to Otte, forms the base of the Abo Formation in the area of these
canyons. This conglomerate-mudstone complex is shown in Otte’s measured
section taken northeast of La Luz (1959: PI. 11). Otte shows the lower con-
glomerate as 1 5 feet thick and the upper conglomerate as four feet thick, but
there is much lateral variation in thickness. Besides the vertebrate fossils
described below, this band contains petrified wood, sometimes fairly large logs,
in many places including some where bone was not found; the wood is
commonly associated with ores of copper. Bones come from both the con-
glomerate and the mudstone, but most were taken from finer-grained lenses
within the conglomerates. Most of the fossils are fragmentary and there are
many items that remain unidentified.

It cannot be said with certainty that the conglomerate-mudstone complex
of the Domingo, Laborcita and Cottonwood Canyon localities lies at exactly
the same horizon as the nodular layer of the northern localities, and it would
be even more difficult to say that there is lithologic continuity between the

1969

Permian Vertebrates from New Mexico

9

two. Nevertheless, there are reasons to suspect a relationship. The northern
nodular layer and the southern conglomerate-mudstone complex are the only
bands in which we found vertebrate fossils, and these bands lie within the zone
of time-equivalence of the uppermost part of the Laborcita Formation and the
lowermost part of the Abo Formation. If one inspects Otte’s diagram of strati-
graphic relationships of the Abo and Laborcita Formations (1959: PI. 13), it
will be seen that his projection northward of the base of marker bed 49 brings
it to lie about 260 feet below the top of the Laborcita Formation (marker bed
55) in the region of the nodular layer in Tularosa Canyon. The nodular layer
is about 50 feet below the top of the Laborcita Formation and thus about 210
feet above the projected base of bed 49. This is very close to the stratigraphic
distance of the conglomerate-mudstone complex above the base of bed 49 —
about 225 feet. An estimate of at least approximate equivalence of the nodular
layer and the conglomerate-mudstone complex would be in accord with Otte's
observation (1959:95) that “The lithologic and faunal characteristics of the
sediments [of the Laborcita Formation] show that abrupt lateral transitions
from open-marine conditions, in the northwest and west, to terrestrial flood-
plain environments, in the southeast and east, occurred repeatedly within a
distance of a few miles.” Although none of the vertebrate elements from the
nodular layer in Tularosa Canyon indicates a marine origin for the deposit,
there are beds both above and below that contain marine invertebrates. Some of
these invertebrates occur in nodules similar to those of the vertebrate-bearing
layer, and such nodules with marine invertebrates are found also in the area of
the locality north of Domingo Canyon. Thus, it seems not unreasonable to
suggest that the Tularosa Canyon vertebrate-bearing layer may be a somewhat
marineward extension of the conglomerate-mudstone complex of the localities
in Domingo Canyon and farther south.

The principal Domingo Canyon locality is in NWV4 sec. 6, T. 15 S, R. 11
E, but fossils were also collected from nearby sites. Fine-grained sandstones in
this area have yielded a number of fragments of bone, and there is abundant
wood in association with copper ores; impressions of “Walchia” leaves were
also found. From a coarse sandstone have come a rhipidistian scale, and also
partial neural spines similar to the complete spines found in Cottonwood Can-
yon and tentatively identified below as Edaphosaurus cf. E. novomexicanus.

The most interesting fossil from the region of the Domingo Canyon
locality consists of parts of the posterior moiety of the skull of a large labyrin-
thodont amphibian, UCLA VP 1727, recovered from a rounded block of red-
brown sandstone that was found loose in the bed of the wash, about a half to
three-quarters of a mile downstream from the fossiliferous layer. The dermal
roofing bones show the pattern of “sculpture” typical of labyrinthodonts. Only
fragments of the circumorbital and cheek regions remain, but the skull table
between the orbits and back to the hind border of the postparietals and tabulars
is fairly well preserved even though a few of the elements are incomplete and

10

Contributions in Science

No. 166

somewhat displaced. Not all the sutures can be easily followed, but it can be
seen that, as is characteristic of rhachitomes, there is a broad contact between
the supratemporal and postparietal bones; and the boundaries of an inter-
temporal bone can be made out— this primitive element is extremely rare
among Permian rhachitomes. Fragments of the braincase are present, and of
the palate there are preserved large portions of the pterygoids in the region of
their articulation with the braincase. The palatal surface of the pterygoid bears
a shagreen of tiny denticles that extends not only forward from the basal articu-
lation, but also back onto the quadrate ramus. The bowed medial borders of the
pterygoids enclose moderately developed interpterygoid vacuities, that is, the
vacuities were much wider than in any described anthracosaur, but they were
somewhat narrower than in the rhachitome Eryops. The processes of the
pterygoids that articulated with the braincase are smoothly finished medially,
and it is quite clear that they were not suturally united with the basipterygoid
processes; the basal articulation was mobile. These features are characteristic
of edopoid rhachitomes, and the specimen fits closely the description of Edops
craigi given by Romer and Witter (1942). Displacement of elements prevents
the taking of a reliable suite of measurements, but the width of the dermal skull
roof across the tabulars was about 115 mm. The corresponding measurement
calculated from the illustrations given by Romer and Witter is about 190 mm,
but these authors point out that the illustrated skull is close to maximum size
and that other, fragmentary remains indicate smaller individuals. The only
other large edopoid from the American Permian is Chenoprosopus — known
only from northern New Mexico— but there are obvious differences between
it and the form from Domingo Canyon. In Chenoprosopus the skull is narrow
and the hind border of the postparietals is markedly convex, whereas the
Domingo Canyon form resembles Edops in that the width across the pterygoids
indicates a wide, flattened skull and the hind border of the postparietals is
concave. Besides, the basal articulation in Chenoprosopus is sutural, at least
in the adult (Langston, 1953). In the lack of better materials, the conservative
course is to identify UCLA VP 1727 as an edopid rhachitome close to, if not
congeneric with, Edops craigi. E. craigi is known only from the lower part of
the Wichita Group in north-central Texas and is thus Wolfcampian in age. The
presence of a closely similar edopid in the lowermost Abo Formation in Otero
County is interesting in that it increases the resemblance of the Early Permian
vertebrate fauna of southern New Mexico to the deltaic fauna of north-central
Texas.

Vertebrate fossils were taken from a number of sites in Laborcita Canyon,
but the two most productive localities are in NE14 sec. 18 and NW14 sec. 17,
T. 15 S, R. 11 E. The more western of these two lies at the end of a spur and
hence may appear, in Figure 1, to lie stratigraphically lower, but the two
are actually within the same conglomerate-mudstone complex. Petrified logs
were found in abundance along the north wall of the canyon. Many of the

1969

Permian Vertebrates from New Mexico

11

fragments of bone from the Laborcita Canyon localities seem to pertain to
pelycosaurian reptiles, and it has been possible to identify one fragmentary
vertebra as of Ophiacodon sp. and one neural spine as of Sphenacodon sp. The
Sphenacodon spine is of a size intermediate between those of S. ferox and
S. ferocior, both of which are known from northern New Mexico — a vertebra
of Sphenacodon cf. S. ferox was found at the Tularosa Canyon locality. One
poorly preserved vertebra may represent Edaphosaurus sp.

The Laborcita Canyon localities have also yielded remains of at least
two kinds of dissorophid labyrinthodont amphibians. From the eastern lo-
cality, a “sculptured” neural spine about 240 mm long above the zygapophyses,
thin from side to side and anteroposteriorly expanded, unquestionably marks
the presence of a species of Platyhystrix, a genus known also from south-
eastern Utah (Vaughn, 1966), southwestern Colorado (Lewis and Vaughn,
1965) and northern New Mexico (Langston, 1953). This specimen, UCLA
VP 1732, may be safely identified as Platyhystrix cf. P. rugosus. Seven articu-
lated dorsal vertebrae with attached ribs (UCLA VP 1722), also from the
eastern locality, are of quite a different dissorophid, although one of about the
same general size as P. rugosus. These vertebrae, of normal rhachitomous
structure with short neural spines, are each about 50 mm tall from the bottom
of the intercentrum to the top of the pitted plate of dermal “armor” that is
solidly fused to the spine. The plates are small, not much wider than twice the
diameter of the narrowest part of the spine. Detailed features of this form
cannot be quite reconciled with published descriptions of any of the numerous
dissorophids presently known — within which there is much variation in the
pattern of the “armor” — but the materials are not sufficient to justify the
naming of a new taxon. A partial skull (UCLA VP 1721) from the western
locality is also of a dissorophid. Large parts of the braincase, of weathered
dermal bones in the region of the orbits, and of the pterygoids are present.
Unfortunately, only very small areas of the surface of the dermal bones are
preserved, and the pattern of the “sculpture” cannot be characterized other
than by saying that it exists. The pterygoids carry a shagreen of tiny teeth, and
the interpterygoid vacuities are enormous. In transverse section, the skull has
the deep, almost subrectangular outline typical of advanced dissorophids. This
was a large form; the width between maxillary rims, measured in a transverse
plane through the centers of the orbits, was about 115 mm. This skull seems
to be appropriate in size to either the Platyhystrix spine or the short-spined
vertebrae. It would be most interesting to be able to demonstrate that it belongs
to the mysterious Platyhystrix, in which the cranium is so far known only from
tuberculated scraps collected in northern New Mexico (Langston, 1953), but
more and better materials are needed. A pelvis of dissorophid aspect was found
in close proximity to the Platyhystrix spine. It may be hoped that further
collection in this area will soon make it possible to supplement our meager
knowledge of this strange, long-spined labyrinthodont.

12

Contributions in Science

No. 166

Only scraps of bone were recovered from the locality between Laborcita
and Cottonwood Canyons, in SW14 sec. 17, T. 15 S, R. 1 1 E, but these serve
to demonstrate the continuity of the fossiliferous band.

Both Cottonwood Canyon localities are in NEV4 sec. 20, T. 15 S, R. 11E.
The locality on the south side of the canyon yielded some dermal bone and a
few phalangeal elements of a labyrinthodont amphibian. On the north side, in
a calcareous sediment that lies just above a conglomerate and that passes
laterally into a mauve sandstone with abundant petrified wood, we found well
preserved remains definitely referable to the pelycosaurian reptile Edapho-
saurus. These materials, UCLA VP 1719, include four very closely associated
vertebrae, two of which are nearly complete, and parts of several ribs. The
direction of curvature of one of the neural spines indicates that the vertebrae
are from the posterior dorsal region. The centra are about 26 mm long, 24 mm
high, and 23 mm wide. The posterior zygapophysial surfaces lie about 40 mm
above the bottom of the centrum. The neural spine of one of the vertebrae was
somewhat more than 410 mm long, and it shows the lateral tubercles charac-
teristic of the genus. Comparison with measurements of centra in various
species of Edaphosaurus published by Romer and Price (1940) shows that
UCLA VP 1719 fits into the size range of both E. novomexicanus — known
from northern New Mexico — and E. boanerges — known from the Wichita
Group of north-central Texas. Romer and Price point out that because E. novo-
mexicanus is inadequately known, it is difficult to distinguish the two species
without good materials of cervical vertebrae. Geographical grounds alone
would not seem to offer sufficient basis, but a difference in horizon might
provide some justification for tentative identification. Steiner and Williams
(1968), on the basis of the fusulinid fauna, come to the conclusion that the
Laborcita Formation is correlative with the Pueblo and lower Moran Forma-
tions of north-central Texas. E. boanerges is known from the Putnam and
Admiral Formations, which lie above the Moran within the Wichita Group.
Correlation with the Texas column of the beds in northern New Mexico from
which E. novomexicanus is known is inexact, but since Romer and Price feel
that E. boanerges is slightly more advanced morphologically, it may not be
unreasonable to assume that E. novomexicanus is from a somewhat lower
horizon. In view, then, of the fact that UCLA VP 1719 comes from that part
of the Abo Formation that is a time-equivalent of the uppermost part of the
Laborcita Formation, stratigraphic considerations perhaps warrant tentative
identification as Edaphosaurus cf. E. novomexicanus. There is no evidence
that the partial neural spines found at the Domingo Canyon locality belong to
a different species.

Vertebrates found in the Laborcita and Abo Formations of Otero County:
Elasmobranch fishes
A pleuracanth
Acanthodian fishes

1969

Permian Vertebrates from New Mexico

13

Acanthodes sp.

Palaeoniscoid fishes

One or more kinds
Crossopterygian fishes
A rhipidistian

Labyrinthodont amphibians
Edopid rhachitomes

A form similar to Edops
Dissorophid rhachitomes

Platyhystrix cf. P. rugosus
A probable new form
Pelycosaurian reptiles
Ophiacodon sp.

Sphenacodon cf. S. ferox
Sphenacodon sp.

Edaphosaurus cf. E. novomexicanus

DIMETRODON from the Caballo Mountains
Using the geologic map published by Kelley and Silver (1952), my field
party searched for vertebrates in the Abo Formation of the Caballo Mountains
and found productive localities in about the center of sec. 36, T. 14 S, R. 4 W
and in NW!4 sec. 6, T. 15 S, R. 3 W, Sierra County. Plants, bones of labyrinth-
odont amphibians, and footprints were found at the first locality, and laby-
rinthodont bones were also found at the second locality. Because this new
source of Lower Permian vertebrates seems so promising, a more extensive
search is planned in this area, but the remains described below, from the
second locality, are of immediate interest to this report and will be discussed
now.

UCLA VP 1724, 1725 and 1726 are fragments of neural spines that
show the figure-8 cross-section characteristic of the pelycosaurian reptile
Dimetrodon. The largest (1725) is about 70 mm long. That the fore and aft
grooves are nearer to one another than the spine is wide shows that this frag-
ment is from about midway along a spine. The width is about 15 mm. Another
fragmentary neural spine, UCLA VP 1733, found along with the others, shows
the characteristic transition from the compressed, laterally flattened proximal
portion to the transversely expanded distal portion. A centrum with a partial
neural arch was found in immediate proximity to the fragmentary spines and
is probably part of the same animal. The centrum has a sharp ventral keel, the
floor of the neural canal is thoroughly ossified, and enough is left of the neural
arch to show that it was laterally excavated; it is thus clear that this specimen,
UCLA VP 1723, is of a sphenacodontid pelycosaur, and this rules out any
possibility of confusion with the poorly known edaphosaur Lupeosaurus. The
centrum is slightly distorted, but it was about 33 mm long, 28 mm wide, and

14

Contributions in Science

No. 166

36 mm high. The posterior zygapophyses lie about 50 mm above the bottom of
the centrum. These dimensions fit into the range of Dimetrodon limbatus
known from north-central Texas and described in detail by Romer and Price
(1940). The Caballo Mountains specimens may be provisionally referred to
Dimetrodon aff. D. limbatus.

In Texas, D. limbatus occurs in the Admiral and Belle Plains Formations
of the Wichita Group and thus straddles the boundary between the Wolf-
campian and Leonardian Series (see Dunbar, et al, 1960). It is noteworthy
that similar vertebrae have been reported from the Organ Rock Shale of the
Cutler Group, southeastern Utah, and that the Organ Rock is probably earliest
Leonardian in age (Vaughn, 1966). The few vertebral parts from the Abo
Formation of the Caballo Mountains hardly offer sufficient grounds for de-
tailed assessment of age, but it would seem that they are from a higher horizon
than the fossils described from the uppermost Laborcita and lowermost Abo
Formations of the Sacramento escarpment, and it cannot be said with assur-
ance that they are Wolfcampian, although the Abo Formation is generally
considered to be almost wholly of that age (Dunbar, et al., 1960). Neverthe-
less, they are certainly at least very close to Wolfcampian in age, and it is
hence of considerable interest that these remains of a species of Dimetrodon
were found in the region of Kottlowski’s (1963) “mean shoreline of the
Wolfcampian seas.” This is the first report of Dimetrodon from New Mexico.

Paleozoogeographic Considerations

The rhipidistian from the Laborcita and Abo Formations of Otero County
and the species of Dimetrodon from the Abo of Sierra County seem to verify
the prediction that these elements were to be found in New Mexico in strata
in the region of interfingering of marine and terrestrial sediments. The presence
of these elements along Kottlowski’s “mean shoreline of the Wolfcampian
seas” lends support to the distinction between “eu-deltaic” and “somewhat
more upland” Early Permian faunas, and perhaps the edopid from Otero
County supplements this support.

Figure 2 is a simplified map of Wolfcampian highlands— or at least
positive areas— and major seaways superimposed on outlines of states in the
Four Corners and Midcontinental regions. Areas of collection of Wolfcampian
— and to some extent lowest Leonardian — vertebrate fossils are indicated by
letters within circles: “C” and “S” represent the Cabello Mountains and
Sacramento escarpment areas in southern New Mexico discussed above; “L”
and “M” stand for localities in the Cutler Group in Lisbon Valley and the
general vicinity of Monument Valley in southeastern Utah discussed in my
earlier paper (1966), which gives references to other papers on these faunas;
“P” stands for the fauna from the undifferentiated Cutler Formation near
Placerville in southwestern Colorado described by Lewis and Vaughn (1965) ;
“A” stands for the vertebrates known from the Cutler Formation around

1969

Permian Vertebrates from New Mexico

15

Figure 2. Simplified map of Wolfcampian positive elements and seaways, based on
maps by McKee, Oriel, et al. (1967). The names of the seaways are informally
derived from names of geosynclines, basins and negative belts. Positive elements
are drawn in black: AP, Apishapa; AR, Arbuckle Mountains; DF, Defiance; FI,
Florida Islands; FR, Front Range; NC, Nacimiento; OM, Ouachita Mountains; PD,
Pedernal; SG, Sierra Grande; SL, San Luis; UN, Uncompahgre; WI, Wichita Moun-
tains; ZN, Zuni. Letters within circles indicate general areas of collection of Wolf-
campian— and to some extent lowest Leonardian — vertebrates: in Utah, L, Lisbon
Valley, and M, Monument Valley; in Colorado, P, Placerville; in New Mexico, A,
Arroyo de Agua and Jemez Springs, C, Caballo Mountains, R, Ribera, and S, Sac-
ramento escarpment; in Oklahoma, O, northern; and in Texas, T, north-central.

Arroyo de Agua in Rio Arriba County and the Abo Formation near Jemez
Springs in Sandoval County, northern New Mexico, that have been sum-
marized by Langston (1953) and Romer (1960); “R” stands for as yet un-
described vertebrates known from the Sangre de Cristo Formation near Ribera
in San Miguel County, northern New Mexico; “T” stands for the well known
Lower Permian vertebrates from the Wichita Group in north-central Texas
summarized by Romer (1958); and “O” represents vertebrates from the
Wellington Formation of northern Oklahoma summarized by Olson (1967).

The outlines of the positive areas and seaways are based on maps pub-
lished by McKee, Oriel, et al. (1967). The positive areas are taken from their
Plate 9, which shows these areas in the time of deposition of rocks of “interval
A” — essentially equivalent to the Wolfcampian Series. The outlines of the
seaways are taken from their Figure 4, which shows the “Distribution of
significant chemical sediments of interval A at time of maximum extent of
sea.” They draw boundaries between limestone, dolomite, anhydrite and gyp-
sum, and halite deposits, but this differentiation is not made in the simplified

16

Contributions in Science

No. 166

diagram presented here. The map in the present report differs in substantial
ways from the one I used in 1966. Better information, from McKee, Oriel,
et al. , now permits the drawing of a much more extensive “Midcontinental
seaway” rather than the restricted “Leonardian seaway” I drew then. The
“Sonoran” and “Orogrande seaways” have been added in southern Arizona
and New Mexico, with the boundary of the latter corresponding essentially to
the belt of gradation mapped by Kottlowski (1963: Fig. 12) from the red-beds
facies of the Abo Formation into the Hueco Limestone. I have included the
southeastern Utah “Elephant Canyon seaway” of my 1966 map in the
“Cordilleran seaway” of the present map. The positive areas of the “Florida
Islands” in New Mexico and the “Ouachita Mountains” in Texas have been
added, and other positive areas have been broken into smaller units. The cluster
of small units north and northwest of the Pedernal positive element in New
Mexico seems to be especially important.

Of the collecting areas indicated in Figure 2, the northern Oklahoma and
north-central Texas localities in the Midcontinent, the Lisbon Valley and
Monument Valley localities in southeastern Utah, and the Caballo Mountains
and Sacramento escarpment localities in southern New Mexico seem to repre-
sent “eu-deltaic” conditions, near borders of persistent seaways. Romer (1958)
has described the deltaic nature of the Wichita Group of north-central Texas,
and I have pointed out the evidence for deltaic conditions during the times of
deposition of the red-beds units of the Cutler Group in southeastern Utah. The
faunas from the Midcontinent and southeastern Utah all include a rhipidistian,
the nectridean amphibian Diplocaulus, the labyrinthodont amphibian Sey-
mouria (with some question in Oklahoma; see Olson, 1967), and the pely-
cosaur Dimetrodon. A difference between the faunas from the two sides of the
“Midcontinental seaway” is that the Midcontinental rhipidistian is Ectosteor-
hachis, whereas only Lohsania has been definitely recognized in southeastern
Utah (Thomson and Vaughn, 1968), although it is not yet clear that Ecto-
teorhachis does not occur there too. It is not possible with the limited materials
at hand to determine the generic identity of the rhipidistian from southern
New Mexico. Ectosteorhachis is not included in Olson’s (1967) faunal lists of
the Wellington Formation of Oklahoma, but he has told me that it does occur
there. It must be noted that the part of the Wellington Formation from which
the “eu-deltaic markers” come is probably an upper Wichita equivalent (Olson,
1967) and thus probably earliest Leonardian rather than strictly Wolfcampian
in age, but this seems close enough for a broad-ranging comparison. Although
the rhipidistian and Diplocaulus from southeastern Utah have been reported
from Wolfcampian horizons within the Cutler Group — the rhipidistian from
both the Halgaito Shale in Monument Valley and a conglomerate low in the
undifferentiated Cutler sediments of Lisbon Valley, and Diplocaulus from the
latter place — Seymouria and Dimetrodon have heretofore been reported in
that region only from a lowermost Leonardian horizon, the Organ Rock Shale

1969

Permian Vertebrates from New Mexico

17

of Monument Valley (Vaughn, 1966). Recently, however, a fragmentary
neural spine of Dimetrodon sp., UCLA VP 1728, has been found in the Hal-
gaito Shale; and it is also of interest that Diplocaulus (UCLA VP 1729) is
now known from the Halgaito too. To date, of the four “markers,” only a
rhipidistian and Dimetrodon are known from the southern New Mexico areas
discussed in previous sections of this report, but it is fair to say that the program
of collection in these areas is still in its very early stages.

The fossiliferous parts of the Cutler Formation around Arroyo de Agua
in Rio Arriba County and of the Abo Formation near Jemez Springs in Sando-
val County, northern New Mexico, seem clearly to be equivalent to the lower —
Wolfcampian — parts of the Wichita Group of north-central Texas (Langston,
1953; Romer, 1960). Despite intensive search in this region since the last
century, a search that has resulted in collection of a large and varied fauna, no
signs of rhipidistians or of Diplocaulus, Seymouria or Dimetrodon have been
reported. The lack of a rhipidistian record is especially striking; when present,
scales of these fishes are usually abundant and obvious. The seaway between
northern New Mexico and the Midcontinent can no longer be cited as an
effective barrier to faunal distribution, now that the above elements are known
from southeastern Utah. The explanation, as I indicated in 1966, probably lies
in a “somewhat more upland” position of the northern New Mexico areas in
Wolfcampian time. Figure 2 shows these areas as relatively far removed from
the borders of seaways, and it may be important to note that an arc of small
positive elements, extending northwestward from the Pedernal highland, sep-
arates these areas from the basin of the “Orogrande seaway.” Much of the
difference between the Early Permian faunas of northern New Mexico and
the Midcontinent may have been due, not to a water barrier, but to environ-
mental dissimilarity in terms of relative degree of approach to truly deltaic
conditions; and the finds of a rhipidistian and Dimetrodon along the “mean
shoreline of the Wolfcampian seas” in southern New Mexico would seem to
corroborate this hypothesis. There is, nevertheless, the possibility that the
“Midcontinental seaway” did form a partially effective barrier — or better,
filter. The pelycosaur Sphenacodon is known from southeastern Utah, southern
New Mexico, and also northern New Mexico, but it is unreported from the
Midcontinent. The long-spined labyrinthodont amphibian Platyhystrix is
known from southeastern Utah, southwestern Colorado, northern New Mexico
and now also southern New Mexico, but not from the Midcontinent, although
somewhat similar forms have been found in north-central Texas.

Beds low in the Sangre de Cristo Formation near Ribera in San Miguel
County, northern New Mexico, have recently yielded a fair number of verte-
brate remains, but none of the “eu-deltaic markers” has been seen there,
although the nature of the deposits should have made rhipidistian scales easily
visible if present. A few vertebrates have also been found farther south in New
Mexico, in the Abo Formation in Socorro County — but still north of the

18

Contributions in Science

No. 166

Caballo Mountains and Sacramento escarpment localities; these are elements
known also from northern New Mexico (Langston, 1953).

The undifferentiated Cutler Formation near Placerville in southwestern
Colorado has yielded a fair-sized vertebrate fauna that may show some special
resemblance to the north-central Texas fauna in the inclusion of the pely-
cosaur Mycterosaurus (Lewis and Vaughn, 1965). An approach to the “eu-
deltaic” faunas may also be indicated by the presence of a primitive sey-
mouriid, known unfortunately from only a single vertebra, but the lack of any
sign of rhipidistians is noteworthy. As in southeastern Utah and northern and
southern New Mexico, Platyhystrix is present. Although it lived close to the
southern end of the Uncompahgre highland, the southwestern Colorado fauna
may have been somewhat intermediate between the faunas of southeastern
Utah and northern New Mexico.

The complete picture is still vague, but it would seem that on a large scale,
Early Permian vertebrate faunas were fairly much alike throughout at least
the western part of the United States although seaways may have formed
filters affecting distribution of certain elements; and on a smaller scale, many
of the differences between local faunas can be accounted for by environmental
dissimilarities such as differing degrees of proximity to truly deltaic condi-
tions. Invocation of water barriers should be a last resort in attempts to explain
differences.

The predictability of discovery of rhipidistians and Dimetrodon in “eu-
deltaic” as opposed to “somewhat more upland” deposits, which has been at
least partially verified in New Mexico, provides a more refined analysis of the
habitat of these forms, and it may be hoped that such refinement can be
extended. Certain vertebrates, whose fossils are rare in either the Four Corners
or the Midcontinent, may lengthen the list of strictly “eu-deltaic” elements. For
example, the pelycosaur Ctenospondylus is known only from southeastern
Utah and north-central Texas, but its remains are extremely rare in Texas.
The edopid labyrinthodont from southern New Mexico reported above, which
may belong to the genus Edops known otherwise only from north-central
Texas, may be another of these elements. It must also be noted that the various
local faunas each seem to contain elements uniquely their own; for examples,
the seymouriamorph-diadectomorph intermediate Tseajaia is known only from
southeastern Utah, the pelycosaur Cutleria is known only from southwestern
Colorado, and the cochleosaurid edopoid Chenoprosopus is known only from
northern New Mexico (see Vaughn, 1966, for references).

A recent analysis of upper Paleozoic floras by Read and Mamay (1964)
corroborates my earlier opinion that drier conditions set in earlier in the
Four Corners region, at least in southeastern Utah, than in the Midcontinent.
Because such considerations are potentially important to a continuing analysis
of faunal differences, the evidence may be briefly reviewed.

The lithology and faunas of the famous Lower Permian red beds of

1969

Permian Vertebrates from New Mexico

19

north-central Texas indicate a history of increasing dryness. Romer (1958:
165-166) has summarized his view of conditions during the time of deposition
of the Wichita Group (the upper part of which straddles the Wolfcampian-
Leonardian boundary): “the flora is in part xerophytic . . . but in some
deposits . . . was of a lush nature, and the considerable quantity of remains
of fishes and of tetrapods obviously aquatic to amphibious in habits indicates
an abundance of water. The picture as a whole suggests conditions comparable
to those of lower Mesopotamia where swamp and desert regions may lie in close
proximity.” In the Clear Fork Group (Leonardian) , above limestones of the
Lueders Formation deposited during an interval of extensive marine trans-
gression, the sediments are of a more oxidized nature except in pond deposits,
and Romer pictures conditions similar to those of the Colorado River delta. In
the Arroyo Formation, burrows of the aestivating lungfish Gnathorhiza are
found, and this is a clear indication of seasonal drought; such burrows are also
found in the overlying Vale Formation. Olson (1958) has interpreted the
evidence from the lithology and faunas of the Vale and the succeeding Choza
Formation as an indication of a continuing trend to drier conditions during
later Clear Fork time. There were apparently extended periods of drying in
Vale time; in Choza time, vertebrates and plants became restricted to the
vicinity of water courses. The higher parts of the Choza Formation are char-
acterized by evaporite deposits.

In southeastern Utah, the Cutler Group consists of, in ascending order:
the Halgaito Shale, the Cedar Mesa Sandstone (apparently of shallow-water
marine origin), the Organ Rock Shale, and the De Chelly Sandstone (aeolian
dunes with vertebrate fossils in the form of trackways only) . These formations
are discussed in my earlier paper (1966) which gives references to reports on
stratigraphy and faunas. The Halgaito and Organ Rock Shales are deltaic red-
beds units, the former of Wolfcampian age and the latter probably earliest
Leonardian. That the transition from Halgaito to Organ Rock conditions
included increasing dryness is indicated even by the scant flora known to date:
the Halgaito yields such plants as Calamites, arborescent lycopods, Callipteris
and Neuropteris, whereas Supaia, Yakia and Walchia come from the Organ
Rock. But there is also evidence that the onset of drier conditions occurred
earlier than in the Midcontinent. Even as low as the Halgaito, vertebrate re-
mains are restricted to the vicinity of stream-channel deposits, and this re-
striction persists in the Organ Rock. In neither of these formations are there
pond deposits, although they are common in the Wichita Group. Pleuracanth
sharks are known from the Halgaito but not from the Organ Rock. Palaeonis-
coid fishes are not common in the Halgaito, and they are exceedingly rare in
the Organ Rock. Whereas in Utah rhipidistians are not found above the horizon
of the Halgaito, in the Midcontinent they persist practically throughout the
Wichita Group in north-central Texas (Romer, 1958) and in Oklahoma reach
as high as the Garber Formation, which is probably a lower Clear Fork equiva-

20

Contributions in Science

No. 166

lent (Olson, 1967). In Utah, the aestivating lungfish Gnathorhiza appears at a
much lower horizon, in the Halgaito Shale, than in north-central Texas. In
Texas, the lungfish of the Wichita Group is the apparently non-aestivating
Sagenodus, and Gnathorhiza does not appear until the horizon of the Arroyo
Formation — approximately middle Leonardian — is reached (Romer, 1958)
although it is present in an upper Wichita equivalent — lowest Leonardian —
in Oklahoma (Olson, 1967). The faunas from the Cutler Group in south-
eastern Utah are poor in the aquatic labyrinthodonts that are such common
elements of the Wichita faunas. Although Eryops and Platyhystrix are known
from the Halgaito, and possible remains of Eryops and a zatracheid from the
Organ Rock, there are as yet no signs of the obviously highly aquatic forms
T rimer orhachis and Archeria. Conditions may have been somewhat moister in
parts of northern New Mexico, as indicated by the presence of an amphibian
similar to Archeria in the Abo Formation near Jemez Springs (Langston,
1953), but to the southeast near Ribera, lungfish burrows high in the Sangre
de Cristo Formation indicate at least seasonal drought (see Vaughn, 1966) .

Read and Mamay (1964) show that Wolfcampian floras were essentially
alike throughout the United States, but that the Leonardian floras show marked
geographic diversity. Of particular significance to the present discussion are
the differences between the Supaia flora — west of the ancestral Rocky Moun-
tains— and the Gigantopteris flora — east of the mountains. Roughly speaking,
the Supaia flora is the Leonardian flora of the Four Corners, and the Gigantop-
teris flora is the Leonardian flora of the Oklahoma and Texas parts of the Mid-
continent, extending into southeastern New Mexico. Read and Mamay make a
distinction between “older” (upper Wichita) and “younger” (Clear Fork)
Gigantopteris floras. They (1964:17) characterize the older Gigantopteris
flora as “a vigorous association that grew under much milder ecologic condi-
tions than did the [contemporaneous] Supaia flora.” Of the latter, they say
( loc . cit.) that “The sediments contain mud cracks and molds of salt crystals,
which suggest a relatively rigorous climate . . . The flora itself lacks . . . ele-
ments that normally grew under swampy or nearly swampy conditions. The
evidence thus suggests that the Supaia flora represents the remnants of a lush
flora that became impoverished by an unfavorable environment and was barred
from eastward dissemination by the Ancestral Rocky Mountains.” Not only is
this broad analysis applicable in detail to the changes seen in progression from
the Halgaito to the Organ Rock Shale, but in the general trends it portrays it
complements the evidence from the vertebrate faunas that drier conditions
appeared earlier in the Four Corners than in the Midcontinent. The causes for
this differential are obscure, but Opdyke (1961), on the basis of directions of
slope in aeolian deposits, came to the conclusion that the prevailing Early
Permian winds of the southwestern United States were from the North and
Northeast; he related this to a lower paleolatitude calculated from paleomag-
netic data for continental drift. Is it possible that the ancestral Rocky Moun-

1969

Permian Vertebrates from New Mexico

21

tains threw something of a rain shadow over parts of the Four Corners area?

The developing climatic picture may permit certain refinements in ex-
planation of detailed differences between the Lower Permian vertebrate faunas
from the two sides of the ancestral Rocky Mountains. It may help explain, for
example, the earlier appearance of aestivating lungfish in southeastern Utah
and perhaps also the earlier disappearance of rhipidistians. In broader terms,
it helps in an understanding of the generally less aquatic aspect of the Four
Corners faunas; Romer (1960:52) has remarked that Wichita conditions seem
to have been “less arid than those of the Abo-Cutler, and this is reflected in the
relative composition of the fauna, the New Mexican beds being conspicuously
poor in their representation of aquatic forms.”

These considerations also suggest that the extensive system of Early
Permian positive elements in Colorado and New Mexico may have formed the
backbone of an upland partial obstacle to vertebrate distribution at least as
effective as the “Midcontinental seaway.”

Acknowledgements

I am grateful to my able field assistants, David Berman, Jimmy Hara,
Edward Mercurio, James Miller and David Vaughn; and to the National
Science Foundation for support of this work through grants GB-5104 and
GB-7784.

Literature Cited

Dunbar, C. O., et al. 1960. Correlation of the Permian formations of North Ameri-
ca. Geol. Soc. Amer., Bull., 71: 1763-1806.

Dunkle, D. H., and S. H. Mamay. 1956. An acanthodian fish from the Lower
Permian of Texas. Wash. Acad. Sci., Jour., 46 : 308-3 10.

Kelley, V. C., and C. Silver. 1952. Geology of the Caballo Mountains. Univ. N.
Mex., Publ. in Geol., 4: 1-268.

Kottlowski, F. E. 1963. Paleozoic and Mesozoic strata of southwestern and south-
central New Mexico. N. Mex. Inst. Min. and Tech., State Bur. Mines and
Mineral Res., Bull. 79: 1-100.

Langston, W., Jr. 1953. Permian amphibians from New Mexico. Univ. Calif. Publ.
Geol. Sci., 29: 349-416.

Lewis, G. E., and P. P. Vaughn. 1965. Early Permian vertebrates from the Cutler
Formation of the Placerville area, Colorado. U. S. Geol. Survey, Prof. Paper
503-C: 1-50.

McKee, E. D., S. S. Oriel, et al. 1967. Paleontectonic maps of the Permian System.
U. S. Geol. Survey, Misc. Geol. Invest. Map 1-450, 164 pp., 20 pis.

Olson, E. C. 1958. Fauna of the Vale and Choza: 14, Summary, review and integra-
tion of the geology and the faunas. Fieldiana, Geol., 10: 397-448.

. 1967. Early Permian vertebrates. Okla. Geol. Survey, Circ. 74: 1-111.

Opdyke, N. D. 1961. The palaeoclimatological significance of desert sandstone. In
Descriptive Palaeoclimatology. New York, Interscience, Chapt. 3: 45-60.

Otte, C. 1959. Late Pennsylvanian and Early Permian stratigraphy of the north-
ern Sacramento Mountains, Otero County, New Mexico. N. Mex. Inst. Min.
and Tech., State Bur. Mines and Mineral Res., Bull. 50: 1-111.

22

Contributions in Science

No. 166

Pray, L. C. 1961. Geology of the Sacramento Mountains escarpment, Otero County,
New Mexico. N. Mex. Inst. Min. and Tech., State Bur. Mines and Mineral Res.,
Bull. 35: 1-144.

Read, C. B., and S. H. Mamay. 1964. Upper Paleozoic floral zones and floral prov-
inces of the United States. U. S. Geol. Survey, Prof. Paper 454-K: 1-35.

Romer, A S. 1952. Late Pennsylvanian and Early Permian vertebrates of the Pitts-
burgh-West Virginia region. Pittsburgh. Carnegie Mus., Ann., 33: 47-113.

. 1958. The Texas Permian redbeds and their vertebrate fauna. In Studies on

Fossil Vertebrates presented to David Meredith Seares Watson, T. S. Westoll,
ed. London, Athlone Press, p. 157-179.

. 1960. The vertebrate fauna of the New Mexico Permian. N. Mex. Geol.

Soc., Guidebook of Rio Chama Country, 11th Field Conf., p. 48-54.

, and L. I. Price. 1940. Review of the Pelycosauria. Geol. Soc. Amer., Spec.

Paper 28: 1-538.

, and R. V. Witter. 1942. Edops, a primitive rhachitomous amphibian from

the Texas red beds. Jour. Geol., 50: 925-960.

Steiner, M. B., and T. E. Williams. 1968. Fusulinidae of the Laborcita Formation
(Lower Permian), Sacramento Mountains, New Mexico. Jour. Paleont., 42:
51-60.

Thomson, K. S., and P. P. Vaughn. 1968. Vertebral structure in Rhipidistia (Os-
teichthyes, Crossopterygii) with description of a new Permian genus. Postilla,
Peabody Mus., Yale Univ., no. 127: 1-19.

Vaughn, P. P. 1966. Comparison of the Early Permian vertebrate faunas of the
Four Corners region and north-central Texas. Los Angeles County Mus., Con-
trib. in Sci., no. 105: 1-13.

Zangerl, R., and E. S. Richardson, Jr. 1963. The paleoecological history of two
Pennsylvanian black shales. Fieldiana, Geol. Mem., 4: 1-352.

Accepted for publication March 10, 1969.

LOS

ANGELES

COUNTY

MUSEUM

MBER 167

CONTRIBUTIONS
IN SCIENCE

June 30, 1969

MAMMALIAN ORGANIZATIONAL SYSTEMS

By George F. Fisler

JUL 24 tow
J/9RARi^

Los Angeles County Museum of Natural History • Exposition Park
Los Angeles, California 90007

CONTRIBUTIONS IN SCIENCE is a series of miscellaneous technical papers
in the fields of Biology, Geology and Anthropology, published at irregular intervals
by the Los Angeles County Museum of Natural History. 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 Robert J. Lavenberg, Los Angeles County Museum of Natural
History, 900 Exposition Boulevard, Los Angeles, California 90007.

Robert J. Lavenberg
Managing Editor

INSTRUCTIONS FOR AUTHORS

Manuscripts for the LOS ANGELES COUNTY MUSEUM, CONTRIBU-
TIONS 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. Al-
though priority will be given to manuscripts by staff members, or to papers dealing
largely with specimens in the Museum’s collections, 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 Edi-
torial Committee. They may also be subject to review by competent specialists out-
side 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.

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. Alterations or changes in the manuscript after galley proof will
be billed to the author. Unless specifically requested, page proof will not 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.

Dorothy M. Halmos
Editor

MAMMALIAN ORGANIZATIONAL SYSTEMS

By George F. Fisler1

Abstract: Confusion exists concerning the relationship be-
tween the seemingly separate concepts of territory and hierarchy
as used in describing mammalian organizational systems. This
paper is an attempt to synthesize knowledge of these concepts into
a single, continuous system of behavioral and ecological adapta-
tions based on two ultimate factors that are characteristics of
mammals, aggression and site attachment. Organizational systems
result from the interaction of these characters and are further re-
fined by proximate factors in mammalian life which include mor-
phological, behavioral, and ecological adaptations.

In order to emphasize diversity of organizational systems,
within and between species, and in order to clarify relationships
between the concepts of territory and hierarchy, a possible classi-
fication of systems is suggested. Based primarily on individual and
group organization, each of these two types is further elaborated
into subtypes which are then defined, discussed, and illustrated.

Introduction

The conspecific relationships and interactions between individual mam-
mals and the relationship of individuals to the land about them have long been
subjects of interest, speculation, and research. Observations of such inter-
actions have resulted in the development of two concepts of organization;
hierarchies of one type or another, and territories, also of several types. Al-
though earlier papers have pointed out the fact (Collias, 1944), it has become
increasingly clear recently that these two concepts are not mutually exclusive.
In the detailed studies of social behavior and population dynamics now being
undertaken, it has frequently become difficult to assign a particular hierarchy
type or territory type to a species. The tendency has been to treat the concepts
independently in spite of assertions over the years that the two systems are
quite related ( e.g ., Darling, 1937; Godfrey and Crowcroft, 1960; Davis, 1958,
1966). Also, the tendency has been to treat the two concepts as causes rather
than results, that mammals and their behavioral and spatial responses result
from the type of social system, rather than that the social system is a product
of the characteristics of the species and its environment. Consequently, there
is considerable misuse and overuse of the terms without recognition that the
concepts could be ill-defined and confused.

In 1957, Emlen cautioned against too rigid a definition of territory.
Bourliere (1954, 1964) maintained that there was a link between the two
ideas, hierarchy and territory, but that this link remained to be analyzed. Davis
(1958) suggested that territory and dominance could be poles of a continuum.

1Research Associate, Los Angeles County Museum of Natural History, and Depart-
ment of Biology, San Fernando Valley State College, Northridge, California 91324.

1

2

Contributions in Science

No. 167

Eisenberg (1965) took a significant step in the direction of correlating the
concepts of hierarchy and territory in his monograph on social organizations of
mammals. He also emphasized the importance of many environmental factors
which are involved in the determination of social organization of a species.
Brown (1966) stated that adjustments of the Burt (1943) concept concerning
home range and territory had become necessary. He considered that most
mammals were actually part of an organized community rather than being
singly organized on the basis of a utilized area (home range), part or all of
which may be defended against conspecifics (territory). The best clear corre-
lation of the two concepts in a general discussion was by Collias (1944) . More
recently Marler and Hamilton (1966) approached the problem by defining
territory as the external reference for dominance fixed in space (see Marler
and Hamilton for further references) , a definition closely following Carpenter’s
(1958) concept of territory as a behavioral system expressed in a spatial-
temporal frame of reference. Additional discussion of the problem by Marler
and Hamilton included the concept of “individual distance” (Hediger, 1950)
and the various types of dominance hierarchies. “Philopatry” (site attach-
ment) (Mayr, 1963) and mutual avoidance were mentioned, with the impli-
cation that the two characteristics may be involved in the maintenance of
home range. Earlier, Scott (1956) had discussed various ways of classifying
behavior patterns and social relationships but did not cover in detail the
problem as defined here.

Difficulty arises in attempting to define the term “territory.” Mammalo-
gists have traditionally followed the concept as outlined by Burt (1943, 1949),
who essentially adapted the earlier ideas of Howard (1920) and other orni-
thologists to mammals. Use of the definition of Noble (1939) that a territory
is “any defended area” has caused many workers to conceive of the social
systems of many mammals as systems of overt (and occasionally covert) de-
fense of a plot of ground, although some authors have realized that defense
may not always be in regard to space ( e.g ., Balph and Stokes, 1963). Noble
himself had a much broader concept of territory and social dominance than
this restrictive definition suggests. Live trapping studies of mammals abound
in the attempt to fit uncertain data into a “territorial” system. Since many
mammal species do not fit this definition, confusion occurs as to just what a
territory is and what the limits of the concept may be. Jewell (1966) has
recently discussed some problems with regard to the confusion of the concepts
of territory and home range, including “core area” of Kaufmann (1962), “foci
of activity” of Carpenter (1940), and “exclusive area” (undefended area) of
Pitelka (1959), and has suggested that a further term, the “monopolized
zone,” be used. Also, Willis (1967) in his discussion of territory has considered
the concept as one of “dominance reversal,” that is, dominance in space de-
creases from the “center” to the periphery of the range of one individual,
whereas dominance increases for an adjacent individual as the “center” of its
range is approached.

1969

Mammalian Organizational Systems

3

Hierarchies and territorial systems are not mutually exclusive concepts
relating on one hand to sociality and on the other hand to spatial utilization.
Rather, these two concepts are actually the possible resultant extremes in the
evolution of mammalian organizational systems, with many possible intermedi-
ate types of systems employing varying degrees of spatial territory and social
hierarchy formation. Territory and hierarchy are not different in kind but
rather in degree. Tinbergen (1957) and others have recognized that there are
territories of space which “move” with the traveling individual, that is, there is
defense of a space but this may change from hour to hour, week to week.
Furthermore, several workers have found that there are hierarchy systems
within territory systems (Brown, 1966; Crowcroft, 1955; Davis, 1958; Reimer
and Petras, 1967) and that at high population densities normally territorial
species may become hierarchical.

The purpose of this paper is to suggest a classification which may provide
a framework which will encompass the data accumulated and still accumu-
lating on the organizational systems of mammals from both spatial and be-
havioral concepts. Eisenberg (1965) has classified mammalian social systems
on the basis of two categories, solitary and communal, with the latter divisible
into subunits, the largest of which is the family band, and has extended this
with a detailed discussion of the concepts for rodents (Eisenberg, 1967). I
propose here to utilize current terminology of a broader scope, with some
additions, in an overlay of the work of Eisenberg and of the concepts of
hierarchy and territory as discussed in Collias (1944) and in Marler and
Hamilton (1966). This classification is, then, neither ecological nor behavioral
but rather, hopefully, a synthesis of the two areas into a unified concept of the
structure of mammalian organizational systems.

Ultimate Factors in Organizational System Formation

Hierarchies, territories, and the spectrum in between are different means
of expression of the same characteristics of mammals. Bourliere (1954, 1964)
has stated that hierarchy and territory formation are both “manifestations of
aggressiveness within the species,” and Eisenberg (1965) stated that territory,
individual distance, harem formation, and hierarchy formation were all im-
plemented by aggressive behavior. Tinbergen (1957) has expressed the idea
that territory is a result of two distinct tendencies which may or may not occur
together, namely, site attachment and intraspecific hostility. Eisenberg ( 1965),
although particularly stressing aggression, also mentioned site attachment as a
factor. Marler and Hamilton (1966) alluded to both aggression and site at-
tachment as being involved in animal spacing. I should like to extend these two
concepts, site attachment and aggression, separately and together, and state
that all mammalian organizational systems depend ultimately on these two
characteristics.

Denial of the existence of site attachment as a mammalian characteristic
is difficult. Innumerable studies are familiar to mammalogists, both trap-mark-

4

Contributions in Science

No. 167

release and visual observation, which attest to the frequency with which indi-
vidual mammals remain in the same area for long periods of time. This fact
has been conceptualized by Burt (1943) in the idea of “home range.” True,
there may be some home range shifting, but apparently few mammal species
are nomadic with no fixed home area. The advantages of and preferences for
such specific places have been discussed by Darling (1937). Furthermore, the
now well-known tendency for many mammals to home over surprisingly long
distances lends additional support to the idea that attachment to a specific site
is a fundamental character for most mammals. The “desire” for the security
of a familiar area is apparently very strong.

Aggressive tendencies are more difficult to assess, perhaps because they
are noticeable primarily as avoidance. It seems clear, however, that aggression
does exist in most mammals. Aggression is defined in many ways, but here
refers to the antagonistic reactions of individuals which are operative at high
intensity (attack) to those reactions just to, but not including, complete pas-
sivity. Defense, therefore, is included within this definition of aggression. Such
actions are directed at conspecifics for the attainment of some specified goal
such as a nest site, food, sex, or psychological well-being. This may be overt
(“face to face”) or covert (chemical, etc.). Discussion concerning aggression
does not appear to question whether it is present but whether it is innate or
learned (see Carthy and Ebling, 1964; Dobzhansky, 1967). It is not important
here to discuss this particular controversy but merely to point out that aggres-
sion is a part of the make-up of most mammals, whether innate or learned
during the maturation process.

Site attachment and aggression, then, form two basic powerful tendencies
which help dictate how a mammal will respond to its environment, even though
the environment may also be influencing these two characteristics in the indi-
vidual, through selection.

Proximate Factors in Organizational System Formation

The actual form assumed by an organizational system will depend on
several factors additional to site attachment and aggression. These systems must
be adaptive, that is, they should be the best system devisable in order to insure
survival for the species, and they must be sufficiently flexible to allow short-
term adjustments of the system as the environment dictates. Proximate factors
form the natural selective processes which mold a population into the proper
organizational system for any particular species under its current environ-
mental conditions.

The proximate factors listed in this paper undoubtedly are not all of those
involved, nor are they listed in any particular order of importance. Assessment
of the importance of these, and other factors, must be made for each individual
species studied. For further discussion of several of these factors, I refer the
reader to the work of Eisenberg (1965).

1. Basic morphology — The physical capabilities of a species in the utili-

1969

Mammalian Organizational Systems

5

zation of a habitat are involved in the type of social system developed. A mouse
does not cover as much ground in its daily routine as does a deer, so the likeli-
hood of similar organizational system development is remote. Furthermore,
the very conspicuousness of the deer as opposed to the mouse will have an
effect on the type of system that will develop.

2. Level of aggressive tendency — Some species are more aggressive than
others, although factors controlling the level of aggression are poorly known.

3. Level of site attachment — Attachment to a particular area may be very
strong in some species and individuals or it may be relatively weak.

4. Habitat occupied — Crook (1966) stressed habitat differences as being
responsible for social system differences between the baboons Thercopithecus
gelada and Papio cynocephalus. Also, many grassland dwellers have developed
systems different from those of forest dwellers of comparable size.

5. Surrounding mammal populations — Interspecific reactions and inter-
actions may influence the organizational system developed. A special case of
interspecific interaction, and perhaps the most important, is the predator-prey
relationship. The predator species requires a different type of organization
than does its prey and one may influence the type of system that the other
evolves (Estes, 1966). A further possible influence is that of biological rank
(Hediger, 1950) in which there may be a “hierarchy” among individuals of
different species, or where one species may be dominant over another (also see
MacMillen, 1964, and especially Calhoun, 1963).

6. Availability of nest, den, and refuge sites — If nest sites, etc., are
readily available and numerous, there will be little or no competition for them.
However, if the population density increases, or sites are destroyed, thereby
leaving such sites in short supply, there may be increased aggression, causing
a different organizational system to develop, at least until the situation is
relieved.

7. Availability and kinds of food — The comments in number 6 also apply
here. In addition, the organizational system may be influenced by food type.
Eisenberg (1965) has emphasized the similarity of the social systems of the
various anteaters (Myrmecophagidae, Manidae, Orycteropodidae) .

8. Reproductive requirements — Many conditions must be met here, but
the most important factors are, first, getting the sexes together, and, second,
caring for the young. There are numerous examples of organizational systems
which differ between the breeding and non-breeding seasons, and the method
of care of the young may vastly influence the organizational system (for ex-
ample, altricial versus precocial, and male participation in care).

9. Differential niche utilization by the sexes (or age groups) — Recent
evidence has indicated that there is a partitioning of the habitat by the sexes in
some birds (Selander, 1966). Future research will certainly show that such
partitioning also occurs in mammals, particularly in those with great sexual
size dimorphism.

10. Population density — There is now evidence that territorial systems

6

Contributions in Science

No. 167

may become hierarchical systems with increased population density, and vice
versa (Brown, 1966, and others). A hierarchy allows a greater number of
animals to occupy a given area than does a territorial system. Selection here
for flexibility of organizational systems is evident.

11. Climate — Difficult weather conditions and climates can alter the
system employed, either on a temporary basis (aggregation at food or water
sources during severe winters or droughts) or from a long-term evolutionary
standpoint.

The Effect of Proximate Factors
General examples

Two examples will illustrate how proximate factors impinge upon a
population in order to form a particular type of organizational system.

The northern fur seal ( Callorhinus ursinus ) is a species of mammal
adapted for an aquatic life, but it must return to land to breed. Site attachment
is high as is evidenced by return to the same localities, even to the natal site in
females, to breed and bear young (Kenyon, 1960). These are large, conspicu-
ous animals needing little cover while on land since land-based predators are
essentially non-existent. Food on land is not a factor as they are dependent
upon the sea. Aggressive level is high among bulls, less so among females. The
population density at the restricted breeding site is high, hence spatial territory
is not possible for all individuals. As a consequence, a system of male domi-
nance and harem formation has evolved. Because of the aggression of the
dominant bulls, younger and weaker bulls are forced to exist in bachelor herds.

The second example will be drawn from a completely different physical
type, a small mouse. Mobility is not great, that is, not much space can be
covered by the individual in a day, or even, for the most part, in a lifetime. The
habitat occupied contains other species competing for the same things, food,
nest sites, etc. Cover may be good, but aggregations as a normal way of life are
inadequate as predators could easily decimate the population (note the preda-
tion on lemmings during their “explosions”). Thus, any social system involving
harems or large aggregations would not be adaptive. Rather, it is more effective
to spread the individuals out over much of the available habitat, each defend-
ing its own small area, its nest site, or perhaps just itself or the area immediately
around it (psychic space).

Sex in relation to other proximate factors

Organizational systems may be reproductive, non-reproductive, or both.
It is well-known that many species of mammals achieve a specific type of
system during the breeding season, whereas they exhibit another during the
non-breeding period ( e.g ., many ungulates), or are apparently “unorganized”
during non-reproductive phases (e.g., some rodents, eared seals). Others main-
tain a similar sort of system whether breeding or non-breeding (e.g., many

1969

Mammalian Organizational Systems

7

primates, some rodents). Also, according to Eisenberg (1965), there is a
“social drive” distinct from sexuality and parental care. It is apparent, then,
that organizational systems are not based entirely on sex. Site attachment is a
year around phenomenon. Even individuals of migratory species tend to return
to the same sites at each end of the migratory path. Habitat selection, and other
proximate factors of organizational system formation, are also in effect all
year. Sex, however, is usually a seasonal phenomenon and the intricate systems
that must be developed for the perpetuation of the species through the more
frequent breeding of the better adapted individuals should be regarded as
organizational systems superimposed on the regular, normal system dictated
by other environmental factors. The sexual hierarchy or territory is apparently
a device evolved to insure the most efficient and bio-energetically economical
way of perpetuating the species under its current environmental conditions,
while at the same time meeting the basic requirements of the species with
regard to psychological and physical well-being. Sexual systems are super-
imposed upon the basic system of the species for that particular habitat (with
possible slight differences for variations of the habitat) , whether this be a loose
society (non-regimented in the sense of an obvious, restrictive system) or an
intricate system maintained all year through reproductive and non-reproduc-
tive periods.

The salt-marsh harvest mouse ( Reithrodontomys raviventris raviventris )
represents an example of a non-regimented system (Fisler, 1965). This form,
in the laboratory, exhibits little aggression or site attachment. Only pregnant
females build nests and then very weakly. There appears to be no real defense
of the nest. In the field, the same tendencies are evident (Fisler, 1965, and
unpublished). Whether in breeding season or not, these mice are highly
mutually tolerant of each other. Apparently under environmental conditions of
dense cover and frequent tidal inundation of their habitat, it is difficult to
establish a home site and thus it is disadvantageous for this species to develop
(or retain) great site attachment tendencies. Therefore, development of breed-
ing aggregations or territories would likewise be inadaptive. Rather, the most
effective way to insure survival under the peculiar marsh conditions was high
toleration, freedom of movement of individuals, no close relationships be-
tween adult individuals (copulation only), and little necessity for parental
care (young are self-sufficient in three or four weeks). The only social attach-
ment is the brief mother-young relationship.

On the other hand, a closely related upland species (presumably the
ancestor of the marsh form), the western harvest mouse ( R . megalotis ), with-
out tidal effects and with sparse cover, is a more excitable, aggressive animal
which does socially organize into male territories with one male dominant over
any other in the immediate vicinity (Fisler, 1965). Nests are well-built and
defended. But such organization is found only in breeding individuals, for
when non-breeding these mice are mutually tolerant. A reproductive hierarchi-
cal spatial territory (see beyond) is possible because the habitat has not im-

8

Contributions in Science

No. 167

posed restrictions of a nature that prevent this more intimate organization.

A second example can be drawn from the work of Darling (1937) on the
red deer ( Cervus elephas). In the non-breeding season, matriarchal herds
composed of females, fawns, and immatures of both sexes are formed. Adult
males are solitary or form small, loose, wide-ranging herds. Male and female
herd ranges may overlap, but not those of females, nor those of males. The
matriarchal herd occupies a group undefended home range (see beyond), and
the organization within each group is hierarchical. The male herds are not
specifically organized. Why? Female herds contain the young which must be
defended, and lacking organization, defense would be individualistic or non-
existent, with perhaps each female defending only her own young. A group
defense is more effective. Selectively, the male herds do not require communal
protection, as the loss of a few individuals is not important. Each male essen-
tially fends for himself, deriving what minimal protection he can obtain from
the loose aggregate with which he may associate. Also, groups probably tend
to compete more successfully for choice land than do individuals. Female herds
require better forage for proper development of the young, and stronger or-
ganization within each herd is dictated to maintain a sufficient area. Competi-
tion among male herds is nil, as enough suitable males will survive for breeding
under most conditions. Familiarity with one area seems to be desirable for
coordinated defense or flight and this can be accomplished better through group
attachment to a particular site.

A further question arises as to why adult males and females do not exist
in the same herds during the non-breeding season. Would it not be more
efficient for males to defend the herd instead of, or as well as, females? Again,
the answer to this apparently lies in other proximate factors impinging upon
the formation of the organizational system. There may be some niche separa-
tion between the two types of herds. Male herds may occupy less desirable
areas, and it may be advantageous for the species to “weed out” those males
that are unnecessary (extra) so that they do not compete directly for food
with fawns and females. No matter what the reason (s) may be, the fact re-
mains that these herds are separated during most of the year ( 10 months) and
only during the reproductive period do we find establishment of a different sys-
tem designed to facilitate the breeding of the stronger individuals, that is, an
organizational system in which the more dominant males control the most
females in harems. A shifting arena defensarium (see beyond) is formed and
maintained. However, within the group (harem), there still exists a female
hierarchy and a female still leads the group (including the male). The basic
female herd structure still exists, although a reproductive type of organiza-
tional system of only two months duration has been superimposed on the
usual type.

A third example can be taken from the primates. The social organization
of many baboons, macaques, and others, has recently been well-reported ( e.g .,
see Southwick, 1963; DeVore, 1965). In the rhesus macaque ( Macaca

1969

Mammalian Organizational Systems

9

mulatto) , bands are composed all year of all age and sex groups. The dominant
males form a linear hierarchy, as do the females, although the latter hierarchy
is not as clear, nor is the relationship between the two hierarchies clear. The
young may assume hierarchical status as adults dependent somewhat on the
status of the female parent (Koford, 1963a). In short, the band remains more
or less constant in content and individual status throughout the year. Previously
(Zuckerman, 1932, and others), it had been considered that bands of pri-
mates were held together because of year around sexual attraction. Yet recent
work indicates that many, if not most, primates do have breeding seasons
(Lancaster and Lee, 1965), just as do most other mammals. It becomes clear,
then, that sex is not the only factor responsible for keeping the organizational
system intact (Lancaster and Lee, 1965) . There must be factors other than sex
of more importance through the entire year which make a year around, rela-
tively stable system advantageous. Apparently, any change of the basic system
in the primate band for reproductive purposes is not advantageous, as indicated
by the fact that there are no other systems superimposed.

The most obvious reason for a continuing organizational system such as
exhibited by the rhesus bands would be the necessity for the care and defense
of the young. These are relatively dependent on the adult female for two or
three years, a period of time longer than in most mammals. Thus long-term
female groups, as in the red deer, are necessary. But why are the males also in
the group? Why are there not male groups, as in the red deer? Firstly, food may
not be a factor requiring separation. Most primates live in tropical or semi-
tropical areas without harsh winters (droughts or dry seasons, however, may
impose harsh food conditions). Males may not be a burden on the band when
they select the best foods for themselves. Secondly, males do function in de-
fense of the band, especially in the defense of the young. This is apparently
of greater selective value than having only subadult males and adult females
performing this task. The subadult male may be inadequate in defense, the
adult female less effective and hampered by the presence of a relatively im-
mobile and helpless youngster. The latter is not true for the deer, where the
young are mobile and less dependent on the female. Since the young rhesus are
dependent for long periods, there is no time of the year when the males are not
required for defense. A system developed thereby where all sex and age groups
remain together all year, eliminating the development of a separate system for
the breeding season. Orderliness and selection of the fitter males for breeding
was provided for in the development of the hierarchy system in use at all times.

Inter-group relations

Monkey bands exhibit a certain degree of organization between and
among each other. In the rhesus, some bands are dominant over others and
may move relatively freely over the area occupied by another group. There
may be various organizational aspects of groups which are similar to and
parallel with organizations based on individuals. These inter-group relation-

10

Contributions in Science

No. 167

ships vary with the primate species under consideration and many of them are
well described in the papers found in DeVore (1965). Suffice it to say that
development of these inter-group relationships in any mammal taxon is also
dependent on the ultimate and proximate causes of organizational system
development. The effect is presumably through the effect of individuals, how-
ever, rather than due to group selection (Wynne-Edwards, 1962).

Summary

The mouse, Reithrodontomys raviventris, is small, inconspicuous, and a
prey species. Its organizational system is greatly influenced by these factors in
addition to sex and habitat. The red deer is a large, conspicuous, prey species.
Its size allows a greater freedom of movement; it now has few, if any, preda-
tors. (Its North American counterpart, Cervus canadensis, has the same type
of organizational system, and predators (Murie, 1951)). It cannot be as in-
conspicuous as the mouse; an aggregating type of organizational system is
allowable, as aggregations will not greatly change its relation to predator
species through increased probability of attack. The predator could find it
easily anyway, even if not grouped. The ease of finding the group over the
individual has no effect, as group defense makes up for this shortcoming. A
mouse aggregation would only destroy the advantage an individual has of being
inconspicuous. The rhesus monkey is of intermediate size and conspicuous-
ness, and its organizational system is complicated by the relative helplessness
of the young and the length of time required to raise them. The monkey is large
enough so that aggregations are not sufficiently disadvantageous to preclude
development of sociality. Food habits are not restricting so as to separate bands
by sex. Defense and the long dependence of the young favor a system with the
greatest amount of predator defense. Sociality and other advantages of this
organization occurred after or with the initial evolution of the group system,
not prior to the development of such a system. It should be noted further that
only the largest primates, those in least danger from predators, have developed
highly flexible and even non-hierarchical or non-territorial group systems
(chimpanzee, Pan troglodytes ; gorilla. Gorilla gorilla ; see Table 11).

Conclusions

From the above discussion, it is apparent that studies of organizational
systems of mammals require extremely detailed data on ecology and social
relations if one is to discern the origin and development of each system and
all its ramifications for individual species. Careful attention to details of or-
ganization at all times of the year is necessary. In compiling the tables in this
paper, many good studies could not be used because of the lack of detailed
information. Eisenberg (1965) has previously stressed this fact and has listed
many relationships which should be studied.

Furthermore, the organizational system evolved may change as the proxi-

1969

Mammalian Organizational Systems

11

mate factors fluctuate. It would be adaptive for a population to utilize different
or slightly differing systems under changing environmental conditions. These
major or minor system changes through fluctuations of proximate factors result
in the myriad shades of differences that one encounters in the study of mam-
malian organizational systems. Dynamic, flexible, adaptive systems such as
these defy classification in the classical sense of “any defended area,” and also
defy relegation to a simple hierarchical system of a static nature. The system of
classification used by Eisenberg ( 1965) , while very useful and probably correct
in the main, does not provide any practical system for succinct discussion of
spatial and social relationships. The discussion by Calhoun (1963), pointing
out possible factors involved in the social use of space, particularly with regard
to group size, does not furnish a classification in a practical sense.

A Classification of Mammalian Organizational Systems

An organizational system in the context of this paper refers to any type
of organization that is present in any species. By definition, then, all species
have such a system even though this is not always “sociar in the normal con-
text of frequent interaction between and among individuals. Solitary species
have their own particular system, even as do highly “socialized” group and
colonial species. In other words, a system in this paper goes beyond the rela-
tions of individuals in close proximity, even beyond the transitory relations of
a pair of mammals in copulation, or the mother-young relationship, which may
be the only sociality (in the usual sense) of some species. These systems repre-
sent the manner in which species are organized in space and behavior at any
particular time of the year or sexual cycle.

Organizational systems cannot be viewed from a static concept in which
one species is considered territorial, another occupies a core area, and a third
species is hierarchical. They are flexible, dynamic systems of adjustment of the
populations to the organization that is best adapted to the major proximate
factors, perhaps even working to the detriment of some factors if the overall
adaptation to the environment is favorable. Therefore, many species will have
different types of organizational systems through the course of the annual
cycle. For example, it is misleading to speak of a species of deer as being
harem-forming unless this is displayed all year. De Vos, et cil., (1967) have
recently reviewed the social behavior of North American cervids during the
reproductive period and have clearly pointed out the differing social organiza-
tions through the yearly cycle. Leuthold (1966) has discussed the variations
and evolution of the types of territorial systems in the Uganda kob ( Adenota
kob), pointing out various factors (proximate) affecting the organizational
structure under differing behavioral and environmental conditions. One must
distinguish between the differing organizational systems throughout the year
for any species of mammal studied.

As in many classifications, the distinctions between categories here are not
always discrete. Indeed, in dynamic systems this cannot be so. For example, a

12

Contributions in Science

No. 167

mammal holding a classic spatial territory (Table 1 ) is in all likelihood also
holding a nidic territory (Table 2) . In placing species in their categories, I have
tried to place them in the category which seems spatially broadest, even though
they may also fit another, less encompassing territorial class. So even though
a mammal species may hold a permanent nidic territory, if the broader organi-
zation shows a basic exclusive core monopolization, it will be so classified.
Investigators should spell these details out in their studies.

A further problem is that of the concept of home range. The classic defi-
nition of Burt (1943) , essentially area traversed in day to day activities, though
not necessarily defended, is still acceptable. It is apparent now that most species
of mammals do have home ranges, preferred areas of occupancy. Very few
species are truly nomadic, traversing any area without regard to particular sites
but remaining only in the proper habitat. All species in this classification do
have preferred areas, but these may shift from time to time. The utilization of
this home range by the individual, and any shifting of the area occupied, de-
pend on the ultimate and proximate factors listed earlier in this paper, with site
attachment being the primary reason for the existence of the home range. How
an individual within a mammalian species uses this home range in relation to
space and other individuals, whether this individual defends all of the home
range, a part of it, or none of it, is part of the organizational system of that
species.

Lastly, the concept that has been termed individual distance (Hediger,
1950), or personal space (Marler and Hamilton, 1966), must be considered.
The former term was introduced for interspecific reactions but could also be
used on an intraspecific basis. Hediger’s terms, and several additional similar
terms, have recently been discussed by Hall ( 1966) . There is a greater or lesser
distance around many mammals within which they ordinarily will not allow
the approach of another individual. This characteristic seems to be responsible
for the formation of organizational systems called here “personal space dis-
persion” and “group psychic space.” Although many species, if not most, have
a personal, probably defended, space (or self), this does not usually supersede
other orders of organization. At times, it may be difficult to separate personal
space (a behavioral concept) from a spatial concept. For example, an indi-
vidual defending only a single homesite (permanent nidic territory) may be
defending just that, space, or it may be defending only itself, located within the
homesite. One certainly gains the impression that a species which defends a
changing homesite (shifting nidic territory) may well be defending only itself,
its personal space, with no real spatial connotations. Only further detailed
study will resolve this situation. A “group psychic space” involves defense of
the group and its current location without reference to fixed spatial boundaries,
thus extending the concept beyond the individual alone.

The classification proposed here does not pretend to be complete and
definitive. Probably no such classification can be. However, it is hoped that this
grouping ( 1 ) will serve as a guideline and can be modified as our rather incom-

1969

Mammalian Organizational Systems

13

plete knowledge of these systems is amplified with new work, and (2) will place
emphasis on the dynamism of organizational systems. I have tried to use terms
currently in use rather than invent new ones, although at times certain classical
definitions may be modified. Certain new terms were also found necessary.
Some species may not fit precisely into the categories as outlined here. This is
to be expected in a dynamic system. Furthermore, many species will be listed
under several categories, reflecting changes of systems over yearly and sexual
cycles. Also, a few species placed in one category at this time may have to be
moved to another as new data dictate. Our knowledge of the details, and even
generalities, of social behavior is still imperfect (Eisenberg, 1965).

This system of classification has one basic dichotomy; organizational
systems based on the individual, and organizational systems based on the group.
I have made no attempt to categorize by age groups, and the dichotomy here is
based on adult systems. Individual organization of a species occurs when the
primary structure rests with the activities and actions of single individuals, or
all individuals are allowed independent action with essentially no dependence
on group action. Group organization occurs when the primary organizational
structure of the species depends on the concerted efforts of more than one
individual so that the group may act as an integrated whole and may take on
properties of the individual (see Eisenberg, 1965, for a further elucidation of
grouped versus aggregated individuals). Many species exhibit more than one
type of organization, frequently at the same time. Examples will be found in
Tables 1 through 11, and there is a detailed example for one species given in
the legend for Figure 1 .

I. Organizational systems based on individuals:

A. Spatial territory (Table 1); defense of a given plot of ground within, or

even including the entire, home range.

1. Exclusive — defense of an area excluding all conspecifics. Only excep-
tion is a brief period for breeding. May be a shifting area.

2. Classic — area permanently defended from conspecifics of the same sex.

3. Shifting classic — as in IA2 except that area defended may change over
time. With further evidence, it is quite likely that IA2 species may
belong in this category. Includes slight alterations to major shifts and
movement to a completely new area.

4. Hierarchical — defense of an area including several individuals (of both
sexes) but with one animal (male) dominant. Remainder defend sub-
territories but are inferior to the dominant who has access to all the
territory.

B. Nidic territory (Table 2); defense of the nest or homesite only (probably

primarily restricted to females, particularly perinatal individuals).

1. Permanent — defense of a permanent nest or homesite.

2. Shifting — defense of the nest(s) or homesite (s) currently used by the
individual.

14

Contributions in Science

No. 167

C. Arena territory (Table 3) ; defense of a specific plot of ground for copula-
tion or harem formation, or both. Restricted mostly to males as a special
reproductive system.

1. Individual — defense of an area (usually small) by an individual which
allows females to enter, but individual females do not remain.

2. Haremic — defense of an area (usually large) which females occupy
relatively permanently.

D. Shifting arena defensarium (Table 4) ; defense of living animals which
move about, therefore not defense of a specific plot of ground. Restricted
mostly to males.

1 . Haremic — defense of the harem, or the space which the harem currently
occupies.

2. Individual — defense of the individual, or the space which the individual
currently occupies.

E. Core monopolization; no overt defense of a plot of ground but use of an
area of concentrated activity within the larger range of the individual.
Many species probably exhibit this system but data as yet are not clear.
Further work should provide definitive examples for this category.

1. Trespass — occasional conspecific intrusion.

2. Exclusive (monopolized area of Jewell, 1966) — no conspecific intru-
sion.

F. Personal space dispersion (Table 5); includes essentially solitary species
where avoidance of conspecifics is the rule; thus individuals are dispersed
although they may move over the same ground. This is conspecific personal
space at its greatest development. It may also include some aggregating
forms where no other form of organizational system is utilized. Many
species here are not overtly hostile to conspecifics (little overt aggression)
but may exhibit some site attachment.

1 . Solitary — individuals that live alone most of their lives.

2, Aggregate — individuals that aggregate or live in close proximity to one
another.

G. Hierarchy (Table 6); individuals involved with diminishing rights of
“possession” within a group; therefore associated with groups of indi-
viduals. Hierarchies, requiring two or more indivdiuals, are a result of de-
creased aggression, increased sociality, lessened individual site attachment,
that is, increased group attachment. As such, this class forms a transition
between individual and group systems and has characteristics of both
individual and group organizations.

1. Linear- — a straight line hierarchy where A is dominant over B, B over
C, etc. It may be a rigid system or a more flexible system of dominance
by frequency rather than absolute right.

2. Overlapping — a system where A is dominant over B, B over C, but C
over A, etc.

1969

Mammalian Organizational Systems

15

3. Ruling consortium — several of the highest ranking individuals support
one another in domination of the other individuals.

H. Undefended home range (Table 7); use of all areas freely by all indi-
viduals. No aggression apparent but there may be some preference for
certain areas by individuals. Mostly aggregates.

II. Organizational systems based on groups:

A. Spatial territory (Table 8) ; group defense of a plot of ground. Contains all
of the subtypes listed under individual spatial territory.

B. Core monopolization (Table 9); as in individual core monopolization in-
cluding both subtypes.

C. Group psychic (“personal”) space; defense of the area which the group
currently occupies rather than a permanent plot of ground. A “moving
territory.” Group composed of all ages and both sexes. Example, howler
monkey ( Alouatta palliata ), Carpenter, 1965, possibly.

D. Hierarchy (Table 10); groups maintain levels of dominance among one
another.

1. Linear — a straight line hierarchy as in individual hierarchies.

2. Overlapping — dominance dependent more on circumstances than upon
dominance rights of one group over another.

E. Undefended home range (Table 11); no apparent aggression between
groups.

The preceding categorization, as emphasized earlier, is based on site
attachment and aggression. Through this classification of organizational sys-
tems there is a decrease in individual aggression from IA through IH and
continuing on into group structure, the latter made possible by decreasing
individual aggression. Also, within group organization, there is a decrease in
group aggression from IIA through IIE. Similarly, there is a decrease in spatial
attachment from IA through IG and from IIA through IID. Spatial attachment
in IH and IIE may be present but is not of over-riding importance in the
organizational structure. A graphic summary of the organizational systems
and factors involved in their formation is presented in Figure 1.

There is some question whether IB (nidic territory) is really a part of I A
(spatial territory). Since, however, it is such a specialized interaction (or the
original territorial condition?) involving defense of nest or den alone, it
deserves separate rank and importance. Many female mammals may exhibit
only this type of organization, hence it should be recognized on this basis alone.

A further problem concerns the type of relationship involved with a
female in defense of her young. Certainly this is a part of any organizational
system. However, since this appears to be such a universal mammalian char-
acteristic, this relationship probably is simply an extension of categories already
listed, such as nidic, or an extension of the personal space of the female to
include the young.

16

Contributions in Science

No. 167

c

o

>s

o

o

M

o

<D

"o

JZ

o»

w

CL

o

■o E

O

>»

0) o

tt>

c

</>

>>

-a »-

o

a.

sz

C m

o

E

cl

o

L.

<J> c

♦-

4>

3

o

® o

o

I*.

O

k_

a>

-C

CL

O

v_

c

(0

o

a>

3

Figure 1. Inter-relationships of mammalian organizational systems and factors deter-
mining these systems. Horizontal distance between vertical lines to the organizational
systems indicates degree of relationship between systems. Additional organizational

home range

1969

Mammalian Organizational Systems

17

In selecting examples for the tables, I have chosen those which, on current
data, illustrate the principles involved. Further references may be found in
Eisenberg (1965). In essence, this classification recognizes the fact (which
has been recognized by earlier authors) that hierarchy and territory are differ-
ent ways of solving adjustment of populations to their environment, abiotic and
biotic, but having the same base. I hope that it will provide a basis for recog-
nition of the fact that most mammals do not fit into one convenient definition
of territory, but, rather, either this term must provide considerably more lati-
tude, or new terms must be coined for the varying systems of organization. By
perusing the literature, it is evident that we no longer can abide by the concepts
as they were first applied to mammals, useful as these were at that time. Indeed,
perhaps a closer look should be taken at classic definitions of territory and
hierarchy in other vertebrate groups, particularly birds, as there are dominance
orders related to the territory occupied (discussed in Colquhoun, 1942; Brown,
1963, 1964; Willis, 1967).

The organizational system developed by any mammalian species depends
on many factors and not necessarily on its phylogenetic history. The only
phylogenetic trends through the order Mammalia for which there is evidence
are ( 1 ) that those mammals with seemingly more complex brains tend toward
group organizational systems (primates and ungulates), and (2) that there is
a general size increase from the smaller species holding spatial territories to the
larger species holding other, succeeding systems (as outlined in this paper).
Since early studies concerned mostly small mammals, perhaps this accounts
for the over-emphasis on territory as a spatial concept only.

In conclusion, in order to discuss organizational systems with understand-
ing and meaningfulness, and in order to make valid comparisons, we are now
in need of a system which will provide a basis for realization of the intricate
balance set up between spatial and behavioral conditions, and that ultimate and
proximate factors as outlined previously in this paper are responsible for the
current form of any mammalian organizational system.

types may be added as data warrant. A = decreasing individual site attachment. B —
decreasing individual overt aggression. C = decreasing group site attachment. D =
decreasing group overt aggression.

The position of the category “undefended home range” is unsure. Degree of site
attachment is also uncertain for this category. Also, it should be kept in mind that any
species can exhibit more than one organizational type through the year, and that dif-
ferent sex and age groups frequently have different organization. Any change in the
proximate factors may also effect a change in the organizational type. For example,
female red deer ( Cervus elephas ) exhibit an undefended home range throughout the
year; males are similar (but separate) during the non-breeding stages, but some
change to a shifting arena defensarium during breeding.

18

Contributions in Science

No. 167

E

y x

X (U

43 <t->

.C

a .c

03

43 73
Cl* C
G

^ j!

||

C3

£ c

O o3

•£ Ie

2 w

3 73

33 .a

CpPP

03 4)

_» (-H

cd

&J0 H
G

43

’G 3

£ xf
G ..
X
43

43 CJO

’S c

03 12

35 13

4) 4)

43

4) „

Oh 2Px>

03 G e

5 §

» II

cd

a
a

cd

a

4)

a

o

73(NX)

^Z
.a w

03 C
03 *-h

cd 73

75 4)
4> n i

VO

VO .

2 s

OV

c ^

rG +S

O 4)

2 G

.a c

4) g

E Q

4)

c

•$ EC

° „

<N <N

S

43

43

s

> ■

i 2 5*
£ 2 15

33 O

C

4)

73

I E

73
C C

o

03 03

g a

o <3

J-l >

e e

03

2

c3

m

VD

P4

VD

m

VD

2

O

r«H

xj-

GV

C\

VO

vo

OV

ov

C\

m

T3

On

xf

vn

VD

go'

vo

«o

ov

JH

,—l

Q\

VD

c3

ov

ov

g"

c3

73

bh

u

4)

G~

O

03

73~

VD

Gs

OV

4H

O

QJ

Ph

o'

>-l

4)

C/J

J-i

4»

o

73

O

i~i

a

£

o

rG

G

43

03

o

^G

O

c3

O

j-T

2

c"

o

4)

£

o

3-1

73

G

c3

o

GO

Oh

cd

43

O

X

B

Z

E

E

PQ

o

(2

E

<u E E
c M ^
o

2 E E E P 9-

(— I I— I HH HH ^

z a

't M M (S

?s

43

-a

43

H 43

a <3

§ 8

C«D

a

a

"a

So

o. c^. c^.

(N (N (N

05

.43

*43

43

E

< o

E G

a

i4

a

05

O

>

a

a

a

Ph C4

E a
8 §

'I P

a

o

C4. ^
O 05

0. 05

.3 ^
^ §

C4,

o

H

*

s |

§ z

a a

.a g

CO

CO §

.43

U

43

O ~a

Sh.

£ a

a o

E 2
< ^
IS

W

CO

Z

a

Co

— ’

° o

E

§

Co

a -a
E**h hi

05

05 a

a a.

.12 O
e a
a

h 4>
43 ^

05 O5

a a

4j Vj

05 05

a a

a a

43 4)

ft, E

7- 7- 't

"a

Co

43

a

&5

g

a

2

« o

S5 ^3

05

a

a

a

£ 05

a

65 "5
05 a
» a

VD VD
VD VD
G\ C\

4) U

M £
4) 4)

o o
E J

4) 4)

G G

O O

Z X

z

z

z

z

E

z

z z

z z

PQ

ffl

E

o-

PQ

Z

E o- Z

E

E

E

E

E

pq"

n e

E E

o

Of

Of

Of

Of

c^»

o-

Of

Of

Of Of

Of

Of

<fo

*o

o-

Of o- <o

«o

<?o

-

«o

«o

<N (N

3? >»

2 Jn.

I "S

!k

ftn E

a a

"§ i

X §

05

05 |

i s

a* S.<3

05 05 E

a a O
5 a >
o o X
a a ^

43 43

a

5 |

43 a

-s

43 43

o? o?

a a

o

1969

Mammalian Organizational Systems

19

G

<D

G

a .

o-,.-

„ TD

i <U
<U

a !_ i

b 40
03 G
a> o

« G

Oh

>HI

G Z
GO -

. c/j
^ C

« ^
HH gj

o <D

>^PQ

o

•r x

0

45

JD

r-( *-H

0)

Vh

0)

O

£ ■§

40

czi

G

G

’S

on

G

M

<D

22

G "

G O

•£3 MH

iS.2

•s >,

o3

cl* f-

yj G
OJ oo
G
O

0
<u
Gh'G
cw c3

g s

aJ ™

2 x
S P-1

1 ^

s-S

O ~
GO (N

no

„ ON

40 o3
m m T?

\X) \£) N

ON ON ^

t£) di} G
• • o

£ £

o

s-T M
<U N
—i on

22 So

h-| J PP PP Z Ph P4

^ cu X HJ X

G C HH H

<zz<<<&

zzz z

cd pa pa pq pa pa m

CH- Of Of

Of fo fo Of Of Of

o

J-l

2 O' O' O' O'

N (U (N O' o- ' — i (N ' — i

^ 05

<U -G

,-G 03 o

So

03

5"S

G O
G =o

S 3

T3 5^
’So G

II

s 3

SX G

O CO

S ^
G G

o

G O
H "G
G ~

G
>a

•S

’S S © o ^ <5
$ <u i ^ ^
£ E £ £ * §

Table 3

Some mammalian species exhibiting an arena territory (type IC). Subtypes are 1, indi-
vidual; 2, haremic. Explanations may be found in the text. B = breeding, N = nonbreeding.

20

Contributions in Science

No. 167

V-)

X

ON

t-~

£

D

£

m

io

ON

NO

ON

r-

^t-

On

(N

r-

lo

ON

X

X

x

NO

o

X

On

a

o

ON

ON

o

X

13

o

o

Lh

<d

&o

_c

ON

r*H

X

Ch

O

c/T

co

j-T

f-T

tn

X

£

T3

CD

m-h

CD

Q

u

J-i

Ctf

ctf

c3

o

O

co

z

o

o

PQ

O

Q

o

W

2 -a
° C

X

X co

X

X

X

X

ON

On

1

tj 73

X

<D 2

D

co <3

co

^ S2

£

O o

O

Q > Q

r- r-

On On

co co
D D
co co

w w

hh w S

HH HH h—H

H ^ M HH

c^.

W W

PQ 0Q PQ PQ PQ PQ

<3 cj
D CL)

z ^

PQ < <

PQ PQ PQ PQ

a

a.

x

-a

a

LO

(N n CN

<N

<N

(N -h

<N

O

a

a.

03

a

-a

a,

CD

a

a

is

S3 Q

IS

S Z
^ Z

-a -a ^
a a o
o

N N O

a x

S .O

S: |

§<

fc £

a o
a to
a oj

a «

Et!

a

a

< a

W ^

Q

C|

2 s

>H Cg

53

a

.a

Sg

a

a

§ s

1-1

« s

S a
a a
a o
k| O

a

"a

!N.

a

a

-a

a

a

a

Co

1 i

"a 'a

Is? K?

a a

O CJ

1969

Mammalian Organizational Systems

21

£ I!

a

8 ^

^ c
'S ;3

5 8

g,«
- <u

s ^

G <0

C -5
a ^

c .5

<D

« 'g

■o §

a o
G n-<
0)

0)
CS ,D

W>

G o3

c/2

c

«J O

.S a

4-; 03

‘-2 'Eh

'_£ X

x W

«

co r3
<L> 2

S T3

a?

.E

JS <N

c

£ o

G <u

« S

S ^3

O „
GO r-H

T3 rn

pH V*Q ^

>n

0) -
S

G ,3 o

5 W o

VO

On

03 r-

ffi ^

£ z

PQ PQ

~g

GO <43
nj G>

□ O
M ^

< •$

2^

Pu

U

< O CQ

00

in

On

>

O

M

*G

a

d

PQ

"§

•!<
Goj
so ✓H

§ R

<43 U
“ <

6:

CO )0

^ S m
^ vo
w> «

G >

;G o +_r

£ hJ G

Q J

H

X DC

02
"a
£

^3

In?

.§ u

S <

Q §

^ ^ a 2 £

Sg^«3

PL <

§ ‘S.Q

Co

13

-G

<43

"S

§-83

PQ

o o

Of

Of

Of

Of Of

Of

Of

Of

Of

+ +

+

+

+

+

«o «o

«o

«o

a

5s.

a

<43

So

K

<3

(<

to

.<43

<43

o

5s.

0/0

Some mammalian species exhibiting personal space dispersion (type IF). Subtypes are 1
solitary; 2, aggregate. Explanations may be found in the text. B - breeding, N = non

22

Contributions in Science

No. 167

T3

G

c3

Jh
0)

. £

I*

E

c3

H-)

\o \o
o\ o\

c c
o o

1/3 C/3

15 'u

z z

c3

E

G

c3

6J0

<D

'C

co

VG

ON

c/T^
<u 'cO

^ o\

O

£a

“ on

C

c3

O
G3 ^

Oh N
GT cz)

c3

PQ P^

a

0

Z

z

z z

z

z z

w

ft

to

PQ

PQ

PQ PQ

PQ~

PQ PQ

H

ft

Of

Of

Of Of

c^-

«0

Of Of

*o

*0

«o

«o <?o

Of

«o <?o

&D

C

T3

cu

ft.

-s:

< »

^ Co

S g
w §

c 00

o ^

Z 3

o h

c^-

<u

G

E E

1— 1 HH

c^-

<u

G

0)

G

0

z

00

0

z

O

z

i— i (N <N

.a

ft*
a

I <

§ E

.6, pq

El ^ h

SS2

e ^ e
^ ^ s

03

„a

n ^

t.3

ft

o
o

<

CLh ft.

a

ft.

<3

ft C
Co H

§.S

§ z

w

Q

w

a 2

ft

"§

3

-ft

o<

ft H

|S

(N <N oh

"5

ft. r _
o 52 ar

ft; ft HH

~ ft 7
ft P-.Z

P^

^ ^ <
u

Orycteropus afer 1 9,3? B, N None? Rahm, 1961

Mammalian Organizational Systems

23

1969

u to
a)

G -a
— 0)
^ <0

PQ

o

<u •
X

3
GO

^ .G

O 03
*-• G
<o 3
O-, O

43

~ G

ia\§

<3

G

3

&H

X

a

•43 3

Z *fi

*3 O
Tto cn

X c
« O

w O

^ toO

0 a

D .G

Dh^

co

rO

G - tob

23 TI.S

23 GO'S

<X>

<u

■ — <u
Gh

Dh-G
3 G
O
G

<u «

a § ii

<N Z

wo wo
VO VO
ON Ov

a>

0)

x. | |

co > >

VO <13 <13 VO
OV Q Q GN

— uo
- 'G *G „ vo Vh

G C G o

f-i 3 3 cj i— i

O ^ 3

o « -3 = £-g

Q PQ PQ < PQ W

zzzzzz

PQ PQ QQ PQ PQ PQ

Of Of Of Of Ch-

«o <?o *o

<o-

co co co to
^ ^ „ c^-

VO

in vo

WO OV

ov —

„ <D
toO'ip

G o

PQ

i— i o

jr g

< o

a £

c^-

z z

PQ PQ

Of
<cO VVD

00 ~

.5 |

Vj 3

3

^ ^ ■>

60

>>

N?

5*.

^3 G

3

'ij 7^ ii

.3

•S

3^

* § < 7;

C

S

3

«n

■s ^

^ s e £

10

so

Cr

3

GO

3

_>p,

'g

3

j-i

<L)

Q

K
>3

o o "g ~b

£.'£.•§ 3

3 -G '

z
c w

S Q
O

?o 3
U hi

24

Contributions in Science

No. 167

4>

Oh

o

cjo

4>

a

o

,£5

TJ

4>

T3

fl

4>
M— I
4>
T3
C3

^ 73

W =5
.J

s :s

■§

euo

c

•’S

■ai
u u
“ 2
.a £

O c

" O

^ a

c 11

c3

g W)

g.s

g'S

« 2

I"

C/D PP

Of

Of

Of

Of

Of

o-

Of

Of

Of

Of

Of

Of

Of

43

Of ^

+

Of

Of

NO NO

- NO

CO

«o

NO

NO

NO

NO

NO

NO

NO

NO

NO

NO

NO

^!-

,h- vo no

S °N

ON ^ ^

^_r 4) a)

S &OW)
£ 3 3
3 03 cS

QUU

IT , m ,_i

NO NO NO
ON ON ON

c c

o o

C/3 CZ)

43 4> 3

^ ^ ^

C3

m rj

vo

ON °V

6h C

I-I o

<U cn

-g O

4)

cz .x

ON

on Nn V)
r-H NO NO
^ On On
T3 <-h '-h
Ih „ „

Ctf J_ J_

O M M

w £ ffi E E

ON

l/N

On

C

O

C/3

!-l

o3

4)

Ph

£

4)

B

o

’o

£

t*

<3

pa

no
m no
On on

fc/J 4)
.£ £
t: ©

cnS 3—1

Q

4) 4)

a c

Ph [4 <u

HH HH c

pp pp pp w

M HH ^ c

<r ° °

Szz

oo5

t— 1 HH f-+

< < < < 5

3— I 3— 1 3-H 1— 1 A-i

<

1— 1

o

t-i f-i

U « K

O' o-

zzz

.S .£ S

sEsE ^

z

z

c^>

z

pp pa pa

z z <

z<zzz&

PQ

pp

PJ

O Q

Z

pa Z

*, o c

s ^

a

a

43

"a

<

E sq
< 2
E |

£.§
c/d a

pp '£ ^

<] hn ^

<3

-a

Cn,

4)

S3 a

a

's' ^ cr

^ PP §

to m Ci~
S □ »2

C4.H s
o p^ c^

b O g

a p4 4>

^ 5 ^

V3

a ^

H .2

1 ^

ls2

a 4>

2 «
>3

a

§1

^-s

O

a

o >>
s*. s
2 a

»H &P

•s g

h ^

NJ i*^,
»2 02
a a
a a

O)

a >

Ccc-£
a >

2 2
O O 02

s K K

5 O a

02 ~?3

rS PS o O

a a a ^ ^

o o o ’S "S

<V) ^V) ^

Os fts Oh * PC

02

a

a

a

a

Ss.

a W

a Ph

I Z

o z

02

J <

2 H

a >^
02 H
S O
•S <

-IS

M O O

a a

a

ao

a ^

•2 ^

G pz m G
<

«n oo
NO ^
On °V

J 43

vo §
VO On
ON

^ *

1 gi

2 -* :
o p

w PP

4)

Z £
pa < pa

02

a

.a

a

-2

a

4) &2

a as

§ |c
"a 2-^
K -2
a 4)

4)

^ a 4>

a

•Sb^ ?

a 2 -S3
a •=: >

PC O O

pa

oj .2

1969

Mammalian Organizational Systems

25

_? a

’B

S3 .9

t/3 '"O

cu G
CL 3

o

42 ^
3 0)

03

< a

f— I CO

jp

Qh-2

os

w c

03

»-( CL
O *

.t! P-1

« ed
+J o

CL E

co (D
CL, IS

2 ^

O

w>*S

.a s

3 g>

x .5

<U £

• L «3

o „
CL

(*> :r

fl ‘

ctf o

1°!

a <l

a >

PQ

w o
O X

C/D <D

W)

_C

X3

o

CD

VD

in C\
\0 i-H

2 3

- .2
—~j *s_,

03

O

'T

ON

Vh

<U

a

<D

CL

S-H

03

” ^ P3

K P u

Q O

c^*

n n <n

a

CL

jo

~

-C3

L

L

O

lO 00

|s

C 03
"1 *-i

pq

s '■—
^ £
s a

<5

o ^
'o ^

o.

PQ

o

B °

<u

O

c

c

o

SI

O

£

£

ZZZ

z z

£

PQ PQ PQ

PQ m

PQ

.2

-2

Of

o

Of

o

+

Of Of Of

Of Of

+ + +

+ +

«o

*o <?o *o

<?o

y

*£a

L

12

o h

2 § 0 a: 5 oo % <S
E «; o

Some mammalian species exhibiting group core monopolization (type IIB). Subtypes are
1, trespass; 2, exclusive. Explanations may be found in the text. B = breeding, N =
nonbreeding.

26

Contributions in Science

No. 167

T2 no

t-H

<d ~

S-H — I

c I ^

P3 _ ^ i/I

<u

o 13 o 13 ^
Q Q O

wn

wo

C\

&D

C

r)

ov

c

c

a

<4-1

O

a

T3

C

c3

X)

.O

3

VO §
NO ON

OV

o o

O O Q O O

SC SC

z 2; z z z z z

m « 03 ed oo 03' 03 Z

Of Of
<fo <0

Of

«o

s

Co

ci

-H (N 1-H

(N

<N

O

Cu

Cv.

Co

s

K

*3

.O

fS

s

•2

3

On

"§

Cu

CSO 53

«n

S
O

<

S|

*C)

53 W

>0

>1

3
S
.3

•S

O

^ ^ cs

S <! Vj
^ C* '£

§

S > 3

o --f 10

ss A

& o

< ~
hJ g
>H fc
H ^

o *

< Cu

9 §

O ^

go

<

Ovis canadensis 1 $ B, N

1969

Mammalian Organizational Systems

27

i-, bO

<d ”2
G ^3
.d D

' o

r VH

1-1 ,D

2 |

G

S II

'S t/j

co .s

. 73

<D

^N <D

9^

CD II
CL

>;pq

^ 0)

^5

oj 2d

g w

2 §

&0^

W) «

X 22

CD c

«, .2

•§|

0,-2

«3 Dh

l ^

73 so

S.3

1 ^

2 ws

G 7^

2 O

o r

co n

x>

co 73 m
^ qj\C
03 ON

ON

O G

G G

§5

o «

£ £
« PQ

o O

+ +

«o «o

Cd

03

§ !•

-2

s 2

GO S 2

Y?

!!•

CD

Dh

CD

to

G

o3

(-H

CD

a

o

dd

73

CD

73

G

CD

q-H

<D

73

G

O

n-i

oo

dd
X
<D too

czi .a
.2 73

*CJ <D
CL) CD
Dh_£

" G
G O
.2 c
1 II
a^

G ^
G _G

CD *5
G <D
G <D

O l-H
CO dd

_r o o -

72 G G <L)

75 CD CD 03
O rY rX 2

rn 'S

U co

u~)

NO

ON

m

l>

NO

NO

NO

„

NO

CO

NO

ON

t/2

73

ON

ON

ON

00 ^
l/"-\

PQ

O

to «
c £
2d o
8-1

Q

On 2
^ O

r*T ^

E o

03 o3
CD <D

< <

CH-

G ^ 00
-h dd 73

2 ^
o3 (D

dd 2d

co^oot

73 a

^ a

•3 >H

02

C3

"G,

O 73
~G -d

s 0

on

CO

0

0 O

Qj

a ^

E

'"Go

t>o<J

"G

0 s

"o

<D

G.

H

, <

0

^ O
^ O

02

>

^ §
<3 cu
•h "C3

CO

§

C3

O H

u.

<D

5

pu

<

O

28

Contributions in Science

No. 167

Acknowledgments

I am indebted to many persons for their comments and criticisms of the
material presented here. I am sure that there are areas where their ideas, those
of many authors, and my own ideas have so merged that credit becomes im-
possible. However, I am particularly indebted to Andrew Starrett and Jim W.
Dole of San Fernando Valley State College for their discussion, comment, and
reading of the manuscript, and to George A. Bartholomew, Jr., of the Uni-
versity of California, Los Angeles, for his reading of the manuscript. This
is not to imply that these people are in full agreement with the ideas presented
here. Responsibility for this synthesis remains with the author.

Summary

The concepts of territory and hierarchy should be considered as opposite
extremes of a continuum of organizational systems of mammals. These systems
are the result of two fundamental characteristics, aggression and site attach-
ment (ultimate factors), which are acted upon additionally (at the individual
level) by proximate factors such as basic morphology, levels of aggression
and site attachment, habitat occupied, interspecific populational effects, com-
petition of various types, population density, and climate. Examples of how
the ultimate and proximate factors impinge upon and adapt a population to its
environment are given.

A classification for mammalian organizational systems utilizing a basic
dichotomy of individual and grouped organizations subdivided into spatial
territory, nidic territory, arena territory, shifting arena defensarium, core
monopolization, personal and group space dispersion, hierarchy, and unde-
fended home range, is suggested. Data for support of this system are given. No
phylogenetic trends were evident within this system other than that those
species with seemingly more complex brains tend toward group organization,
and smaller species tend to be found at the territorial end of the organizational
system spectrum. It is hoped that this system will help clarify evident confusion
as to the relationships among mammalian organizational systems.

Literature Cited

Balph, D. F., and A. W. Stokes. 1963. On the ethology of a population of Uinta
ground squirrels. Amer. Midland Nat., 69:106-126.

Bannikov, A. G. 1958. Distribution geographique et biologie du cheval sauvage et
du chameau de Mongolie ( Equus przewalskii et Camelus bactrianus) . Mam-
malia, 22:152-160.

Bartholomew, G. A., Jr. 1953. Behavioral factors affecting social structure in the
Alaska fur seal. 8th N. A. Wildlife Conf., Trans., 18:481-502.

, and P. G. Hoel. 1953. Reproductive behavior of the Alaska fur seal, Cal-

lorhinus ursinus. J. Mammal., 34:417-436.

Bourliere, F. 1954. The natural history of mammals. Knopf, New York. 363 p.

. 1964. The natural history of mammals. Knopf, New York. 3rd ed., rev.

387 p.

1969

Mammalian Organizational Systems

29

Bradt, G. W. 1938. A study of beaver colonies in Michigan. J. Mammal., 19:139-162.

Broadbooks, H. E. 1965. Ecology and distribution of the pikas of Washington and
Alaska. Amer. Midland Nat., 73:299-335.

Bronson, F. H. 1964. Agonistic behaviour in woodchucks. Anim. Behaviour, 12:470-
478.

Brown, J. L. 1963. Aggressiveness, dominance and social organization in the Steller
jay. Condor, 65:460-484.

. 1964. The evolution of diversity in avian territorial systems. Wilson Bull.,

76:160-169.

Brown, L. E. 1966. Home range and movement of small mammals. Zool. Soc. Lon-
don, Symp., 18:111-142.

Buechner, H. K. 1960. The bighorn sheep in the United States, its past, present, and
future. Wildlife Monogr., 4:1-174.

. 1961. Territorial behavior in Uganda kob. Science, 133:698-699.

Burt, W. H. 1943. Territoriality and home range concepts as applied to mammals. J.
Mammal., 24:346-352.

. 1949. Territoriality. J. Mammal., 30:25-27.

Calhoun, J. B. 1963. The social use of space, pp. 1-187. In Physiological mammal-
ogy, vol. 1. W. V. Mayer and R. G. Van Gelder, eds. Academic Press, New York
and London. 381 p.

Cameron, A. W. 1967. Breeding behavior in a colony of western Atlantic gray seals.
Canadian J. Zook, 45 : 161-173.

Carpenter, C. R. 1940. A field study in Siam of the behavior and social relations of
the gibbon ( Hylobates lar ). Comp. Psychol. Monogr., 16: 1-212.

. 1958. Territoriality: a review of concepts and problems, pp. 224-250. In Be-
havior and evolution. A. Roe and G. G. Simpson, eds. Yale Univ. Press, New
Haven. 557 p.

. 1965. The howlers of Barro Colorado Island, pp. 250-291. In Primate be-
havior. I. DeVore, ed. Holt, Rinehart, and Winston, New York. 654 p.

Carthy, J. D., and F. J. Ebling, eds. 1964. The natural history of aggression. Aca-
demic Press, London and New York. 159 p.

Caughley, G. 1964. Social organization and daily activity of the red kangaroo and
the gray kangaroo. J. Mammal., 45:429-436.

Coe, M. J. 1962. Notes on the habits of the Mount Kenya hyrax ( Procavia johnstoni
mackinderi Thomas). Zool. Soc. London, Proc., 138:639-644.

Collias, N. E. 1944. Aggressive behavior among vertebrate animals. Physiol. Zool.,
17:83-123.

Colquhoun, M. K. 1942. Notes on the social behaviour of blue tits. British Birds,
35:234-240.

Crook, J. H. 1966. Gelada baboon herd structure and movement. A comparative re-
port. Zool. Soc. London, Symp., 18:237-258.

Crowcroft, P. 1955. Territoriality in wild house mice, Mus musculus L. J. Mam-
mal., 36: 299-301.

. 1957. The life of the shrew. Reinhardt, London. 166 p.

Darling, F. F. 1937. A herd of red deer, a study in animal behaviour. Clarendon
Press, Oxford. 215 p.

. 1947. The life history of the Atlantic grey seal. pp. 217-231. In his Natural

history in the highlands and islands. Collins, London. 306 p.

Davis, D. E. 1958. The role of density in aggressive behaviour of house mice. British
J. Anim. Behaviour, 6:207-210.

. 1966. Integral animal behavior. Macmillan, New York. 1 18 p.

Deraniyagala, P. E. P. 1955. Some extinct elephants, their relatives and the two liv-
ing species. Govt. Press, Ceylon. 161 p.

30

Contributions in Science

No. 167

DeVore, I., ed. 1965. Primate behavior: field studies of monkeys and apes. Holt,
Rinehart, and Winston, New York. 654 p.

, and K. R. L. Hall. 1965. Baboon ecology, pp. 20-52. In Primate behavior.

I. DeVore, ed. Holt, Rinehart, and Winston, New York. 654 p.

de Vos, A., P. Brokx, and V. Geist. 1967. A review of social behavior of the North
American cervids during the reproductive period. Amer. Midland Nat., 77:390-
417.

, and R. J. Dowsett. 1966. The behaviour and population structure of three

species of the genus Kobus. Mammalia, 30:30-55.

Dobzhansky, T. 1967. Review of “On aggression,” by K. Lorenz. Anim. Behaviour,
15:392-393.

Dunnet, G. M. 1964. A field study of local populations of the brush-tailed possum
Trichosurus vulpecula in eastern Australia. Zool. Soc. London, Proc., 142:665-
695.

Eisenberg, J. F. 1963a. The intraspecific social behavior of some cricetine rodents
of the genus Peromyscus. Amer. Midland Nat., 69:240-246.

. 1963b. The behavior of heteromyid rodents. Univ. Calif. Publ. Zook, 69:1-

100.

. 1965. The social organizations of mammals. In Kukenthal, W., and T. Krum-

bach, eds., Handbuch der Zoologie, VIII. 10(7) : 1-92.

. 1967. A comparative study in rodent ethology with emphasis on evolution of

social behavior, I. U. S. National Mus., Proc., 122:1-51.

Emlen, J. T. 1957. Defended area? A critique of the territory concept and of con-
ventional thinking. Ibis, 99:352.

Estes, R. D. 1966. Behaviour and life history of the wildebeest ( Connochaetes tau-
rinus Burchell). Nature, 212:999-1000.

. 1967. The comparative behavior of Grant’s and Thomson’s gazelles. J.

Mammal., 48:189-209.

Evans, F. C., and R. Holdenried. 1943. A population study of the Beechey ground
squirrel in central California. J. Mammal., 24:231-260.

Fisler, G. F. 1965. Adaptations and speciation in harvest mice of the marshes of San
Francisco Bay. Univ. Calif. Publ. Zool., 77:1-108.

Godfrey, G., and P. Crowcroft. 1960. The life of the mole ( Talpa europaea Lin-
naeus). Museum Press, London. 152 p.

Goodall, J. 1965. Chimpanzees of the Gombe Stream Reserve, pp. 425-473. In Pri-
mate behavior. I. DeVore, ed. Holt, Rinehart, and Winston, New York. 654 p.

Gordon, K. 1936. Territorial behavior and social dominance among Sciuridae. J.
Mammal., 17:171-172.

Grubb, P., and P. A. Jewell. 1966. Social grouping and home range in feral Soay
sheep. Zool. Soc. London, Symp., 18: 179-210.

Hall, E. T. 1966. The hidden dimension. Doubleday and Company, Inc., Garden
City, New York. 201 p.

Hall, K. R. L. 1965. Behaviour and ecology of the wild Patas monkey, Erythrocebus
patas, in Uganda. Zool. Soc. London, Proc., 148:15-87.

, and I. DeVore. 1965. Baboon social behavior, pp. 53-110. In Primate be-
havior. I. DeVore, ed. Holt, Rinehart, and Winston, New York. 654 p.

Hediger, H. 1950. Wild animals in captivity. Butterworths Scientific Publications,
London. 207 p.

, and H. Kummer. 1961. Das Verhalten der Schnabeligel (Tachyglossidae).

In Kukenthal, W., and T. Krumbach, eds., Handbuch der Zoologie, VIII. 10(8a) :
1-8.

Heinsohn, G. E. 1966. Ecology and reproduction of the Tasmanian bandicoots (Per-
ameles gunni and Isoodon obesulus). Univ. Calif. Publ. Zool., 80: 1-96.

LIoward, H. E. 1920. Territory in bird life. Murray, London. 308 p.

1969

Mammalian Organizational Systems

31

Howard, W. E. 1949. Dispersal, amount of inbreeding, and longevity in a local popu-
lation of prairie deermice on the George Reserve, southern Michigan. Michigan,
Univ., Lab. Vert. Biol., Contr., 43 : 1-50.

, and H. E. Childs, Jr. 1959. Ecology of pocket gophers with emphasis on

Thomomys hottae mewa. Hilgardia, 29:277-358.

Innis, A. C. 1958. The behaviour of the giraffe, Giraffa Camelopardalis, in the eastern
Transvaal. Zool. Soc. London, Proc., 131:245-278.

Jay, P. 1965. The common langur of north India, pp. 197-249. In Primate behavior.
I. DeVore, ed. Holt, Rinehart, and Winston, New York. 654 p.

Jewell, P. A. 1966. The concept of home range in mammals. Zool. Soc. London,
Symp., 18:85-109.

Kaufmann, J. H. 1962. Ecology and social behavior of the coati, Nasua narica, on
Barro Colorado Island, Panama. Univ. Calif. Publ. Zool., 60:95-222.

Kenyon, K. W. 1960. Territorial behavior and homing in the Alaska fur seal. Mam-
malia, 24:431-444.

Kilham, L. 1958. Territorial behavior in pikas. J. Mammal., 39:307.

King, J. A. 1955. Social behavior, social organization, and population dynamics in a
black-tailed prairiedog town in the Black Hills of South Dakota. Michigan,
Univ., Lab. Vert. Biol., Contr., 67: 1-123.

Koford, C. B. 1957. The vicuna and the puna. Ecol. Monogr., 27 : 153-219.

. 1963a. Rank of mothers and sons in bands of rhesus monkeys. Science, 141:

356-357.

. 1963b. Group relations in an island colony of rhesus monkeys, pp. 136-152.

In Primate social behavior. C. H. Southwick, ed. Van Nostrand Comp., Inc.,
Princeton, N. J. 191 p.

Krieg, H., and U. Rahm. 1961. Das Verhalten der Ameisenbaren (Myrmecopha-
gidae). In Kukenthal, W., and T. Krumbach, eds., Handbuch der Zoologie,
VIII. 10( 12a) :2-19.

Kulzer, E. 1961. Uber die Biologie der Nil-Flughunde ( Rousettus aegyptiacus) .
Natur und Volk, 91:219-228.

Kummer, H., and F. Kurt. 1963. Social units of a free-living population of Hama-
dryas baboons. Folia Primatol., 1:4-19.

Lancaster, J. B., and R. B. Lee. 1965. The annual reproductive cycle in monkeys
and apes. pp. 486-513. In Primate behavior. I. DeVore, ed. Holt, Rinehart, and
Winston, New York. 654 p.

Lay, D. W. 1942. Ecology of the opossum in eastern Texas. J. Mammal., 23 : 147-159.

Layne, J. N. 1954. The biology of the red squirrel, Tamiasciurus hudsonicus loquax
(Bangs), in central New York. Ecol. Monogr., 24:227-267.

Lent, P. C. 1965. Rutting behaviour in a barren-ground caribou population. Anim.
Behaviour, 13:259-264.

Leuthold, W. 1966. Variations in territorial behavior of Uganda kob Adenota kob
thomasi (Neumann 1896). Behaviour, 27:215-257.

Linsdale, J. M. 1946. The California ground squirrel; a record of observations made
on the Hastings natural history reservation. Univ. Calif. Press, Berkeley, 475 p.

Lockie, J. D. 1966. Territory in small carnivores. Zool Soc. London, Symp., 18: 143-
165.

Lockley, R. M. 1961. Social structure and stress in the rabbit warren. J. Anim. Ecol.,
30:385-423.

Lowe, V. P. W. 1966. Observations on the dispersal of red deer on Rhum. Zool. Soc.
London, Symp., 18:211-228.

MacMillen, R. E. 1964. Population ecology, water relations, and social behavior of
a southern California semidesert rodent fauna. Univ. Calif. Publ. Zool., 71 : 1-59.

Marler, P. R., and W. J. Hamilton, III. 1966. Mechanisms of animal behavior. John
Wiley and Sons., Inc., New York. 771 p.

32

Contributions in Science

No. 167

Mayr, E. 1963. Animal species and evolution. Belknap Press, Cambridge, Mass.
797 p.

Murie, O. J. 1951. The elk of North America. Wildlife Management Institute, Wash-
ington, D. C., and Stackpole Co., Harrisburg, Pa. 376 p.

Nelson, J. E. 1965. Behaviour of Australian Pteropodidae (Megachiroptera). Anim.
Behaviour, 13:544-557.

Nicholson, A. J. 1941. The homes and social habits of the wood-mouse ( Peromys -
cus leucopus noveboracensis ) in southern Michigan. Amer. Midland Nat., 25:
196-223.

Noble, G. K. 1939. The role of dominance in the social life of birds. Auk, 56:263-
273.

Orr, R. T. 1967. The Galapagos sea lion. J. Mammal., 48:62-69.

Pearson, O. P. 1959. Biology of the subterranean rodents, Ctenomys in Peru. Lima,
Peru, Mus. Hist. Nat. “Javier Prado,” Mem., 9:1-56.

Pitelka, F. A. 1959. Numbers, breeding schedule, and territoriality in pectoral sand-
pipers of northern Alaska. Condor, 61:233-264.

Rahm, U. 1961. Das Verhalten der Erdferkel (Tubulidentata). In Kukenthal, W.,
and T. Krumbach, eds., Handbuch der Zoologie, VIII. 10(23) : 1-8.

Reimer, J. D., and M. L. Petras. 1967. Breeding structure of the house mouse, Mus
musculus, in a population cage. J. Mammal., 48:88-99.

Reynolds, V., and F. Reynolds. 1965. Chimpanzees of the Budongo Forest, pp. 368-
424. In Primate behavior. I. DeVore, ed. Holt, Rinehart, and Winston, New
York. 654 p.

Ryszkowski, L. 1966. The space organization of nutria ( Myocastor coy pus) popula-
tions. Zool. Soc. London, Symp., 18:259-265.

Schaller, G. B. 1963. The mountain gorilla, ecology and behavior. Univ. Chicago
Press, Chicago. 43 1 p.

Schenkel, R. 1966. Play, exploration and territoriality in the wild lion. Zool. Soc.
London, Symp., 18:11-22.

Scott, J. P. 1956. The analysis of social organization in animals. Ecology, 37:213-

221.

Selander, R. K. 1966. Sexual dimorphism and differential niche utilization in birds.
Condor, 68:113-151.

Southwick, C. H., ed. 1963. Primate social behavior. Van Nostrand Comp., Inc.,
Princeton, N. J. 191 p.

Tinbergen, N. 1957. The functions of territory. Bird Study, 4: 14-27.

Ullrich, W. 1961. Zur Biologie und Soziologie der Colobusaffen ( Colobus guereza
caudatus Thomas 1885). Zool. Garten, (n.s.) 25:305-368.

Willis, E. O. 1967. The behavior of bicolored antbirds. Univ. Calif. Publ. Zool., 79:
1-127.

Wolfe, J. L. 1966. Agonistic behavior and dominance relationships of the eastern
chipmunk, Tamias striatus. Amer. Midland Nat., 76:190-200.

Wynne-Ed wards, V. C. 1962. Animal dispersion in relation to social behaviour.
Hafner, New York; Oliver and Boyd, Edinburgh and London. 653 p.

Yerger, R. W. 1953. Home range, territoriality, and populations of the chipmunk in
central New York. J. Mammal., 34:448-458.

Zuckerman, S. 1932. The social life of monkeys and apes. Harcourt, Brace, New
York; Routledge and Kegan Paul, Ltd., London. 357 p.

Accepted for publication March 11, 1969.

LOS

ANGELES

COUNTY

MUSEUM

CONTRIBUTIONS
IN SCIENCE

£ 2L

JMBER 168

June 30, 1969

TAXONOMIC REVIEW OF THE MEXICAN SKINKS OF
THE EUMECES BREVIROSTRIS GROUP

By James R. Dixon

Los Angeles County Museum of Natural History • Exposition Park
Los Angeles, California 90007

CONTRIBUTIONS IN SCIENCE is a series of miscellaneous technical papers
in the fields of Biology, Geology and Anthropology, published at irregular intervals
by the Los Angeles County Museum of Natural History. 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 Robert J. Lavenberg, Los Angeles County Museum of Natural
History, 900 Exposition Boulevard, Los Angeles, California 90007.

Robert J. Lavenberg
Managing Editor

INSTRUCTIONS FOR AUTHORS

Manuscripts for the LOS ANGELES COUNTY MUSEUM, CONTRIBU-
TIONS 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. Al-
though priority will be given to manuscripts by staff members, or to papers dealing
largely with specimens in the Museum’s collections, 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 Edi-
torial Committee. They may also be subject to review by competent specialists out-
side 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.

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. Alterations or changes in the manuscript after galley proof will
be billed to the author. Unless specifically requested, page proof will not 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.

Dorothy M. Halmos
Editor

TAXONOMIC REVIEW OF THE MEXICAN SKINKS OF
THE EUMECES BREVIROSTRIS GROUP

By James R. Dixon1

Abstract: The nominal species Eumeces dicei and E. in-
dubitus are shown to be subspecies of E. brevirostris. Eumeces
copei is removed from the E. anthracinus group and placed in the
E. brevirostris group. A critical analysis of squamation and color
pattern is presented for the group. Keys, photographs, and distri-
bution maps are given for each form. A possible explanation for
the present distributional patterns of the populations of E. brevi-
rostris is presented.

Introduction

While conducting an ecological survey of the herpetofauna of the Navado
de Colima in southwestern Jalisco, Mexico, I collected a series of skinks be-
longing to the Eumeces brevirostris group but could not allocate them to
species. An examination of specimens comprising the group indicated that
discrepancies existed in certain characters used to define the species. Such
characters as the length of the dorsolateral light line, interparietal enclosed or
not, seventh supralabial in contact with the upper secondary temporal scale or
not, number of supraoculars, supraoculars in contact with the frontal, trans-
verse dorsal and longitudinal dorsal scale rows, and the presence or absence of
a lateral light line, proved to be more variable than reported by other workers
(Taylor, 1935; Smith and Taylor, 1950; Tanner, 1958). A systematic analysis
seemed necessary for a redefinition of the species of the E. brevirostris group.

Characteristics of the EUMECES BREVIROSTRIS Group

Scales : The group is characterized by having the scale following the
postgenial wider than long; one postmental scale, and postnasal scales absent;
seventh supralabial scale usually in contact with upper secondary temporal
scale, occasional contact on one side only or absent in a few individuals of
Eumeces brevirostris, E. dugesi, E. colimensis, and in most specimens of
E. copei ; three anterior supraocular scales in contact with frontal on each side,
except in E. dugesi , with two; four supraocular scales on each side, except in
E. dugesi with three; interparietal scale either enclosed by the parietal scales or
abutting with nuchal scales; frontonasal scale either in contact with frontal or
separated from it by prefrontals or an asygous scale; primary temporal scale,
if present, in contact with lower secondary temporal scale; primary temporal

'Research Associate in Herpetology, Los Angeles County Museum of Natural His-
tory, and Associate Professor of Wildlife Science, Texas A & M University, College
Station, Texas.

1

2

Contributions in Science

No. 168

scale frequently fused to upper secondary temporal or absent in some species;
two tertiary temporal scales, lower scale often hidden by auricular scales;
transverse dorsal scale rows 49 to 68; longitudinal dorsal scale rows 20 to 28;
lamellae beneath the fourth row 10 to 17.

Color patterns : Color pattern of the various species of the Eumeces
brevirostris group is relatively stable but shows some loss of the boldness of
pattern with age. Older individuals tend to lose the black or brown pigment,
with a darkening of the white to cream yellow light lines on the head and
dorsum. The color pattern is usually consistent for any given population but,
in some cases, a cline may be found in the width, length, and distinctiveness of
some of the lines in adjoining populations. No light median lines or bifurcating
light lines occur on the head or body.

The color pattern of the various populations of the E. brevirostris group
consists of six major longitudinal lines as follows: (1) dorsolateral light line-
varying length, usually beginning on the rostral, passing posteriorly along the
outer edge of the supraoculars, nuchals, and on longitudinal scale rows two,
three, and sometimes four; (2) lateral light line — varying length, beginning on
the supralabials, passing posteriorly to the ear, arm, hind limb, or tail, usually
on longitudinal scale rows four, five, six, and seven, or a combination of any of
these in sequence; (3) primary lateral dark line — beginning near the nostril,
passing posteriorly along side of head, neck, body, and sometimes onto tail,
usually covering longitudinal scale rows three, four, five, and six, or a combi-
nation of any of these in sequence; (4) upper secondary dark line (not always
present)— -varying length along the inner border of the dorsolateral light line,
usually beginning on the frontonasal, passing posteriorly along the middle of
the supraoculars and along longitudinal scale rows one, two, or three; (5)
lower secondary dark line (not always present) — beginning on the neck on
longitudinal scale rows five, six, or seven, sometimes present over the arm and
along the side of the body to the hind limb and sometimes onto the tail, usually
present on longitudinal scale rows five, six, seven, and eight, or any combina-
tion of these in sequence at midbody; (6) dorsal dark lines (interrupted occa-
sionally, often absent) — various lengths, usually extending from the nuchals to
above the hind limbs, usually located between two longitudinal scale rows, such
as a median line between the paravertebral rows and between longitudinal
scale rows two and three ( E . copei ) or passing down the center of each
paravertebral row (as in some E. brevirostris pineus ).

The presence, absence, length and position of each line on a particular
longitudinal scale row is important in diagnosing species of the group. For this
reason the method of counting is given below.

Longitudinal scale rows are counted from the paravertebral (first) row,
laterally. Transverse scale rows are counted from the first nuchal to above the
anus. Frequently, lines shift position from one row to another, usually between
the neck and midbody; for example, a dorsolateral light line on the lower half
of the second longitudinal scale row for four transverse scales may shift to the

1969

Mexican Skinks of E. BREVIROSTR1S Group

3

middle one-third of the third longitudinal scale row for 54 transverse scale
rows. If a line is wider than one scale row, e.g., the lateral dark line, it may be
found on the lower half of the third longitudinal scale row through the upper
half of the fifth on the neck, shifting to the lower half of the fourth longitudinal
scale row through the upper half of the sixth at midbody.

Taxonomic Arrangement of the Skink
Group EUMECES BREVIROSTRIS
I propose the following changes in the most recently published taxonomic
arrangements of the species of the Eumeces brevirostris group (Smith and
Taylor, 1950; Tanner, 1958).

Previous status

E. brevirostris brevirostris (Gunther)
E. brevirostris bilineatus Tanner
E. dicei dicei Ruthven and Gaige

E. dicei pineus Axtell

E. indubitus Taylor

E. colimensis Taylor
E. copei Taylor
E. dugesii Thominot
E. ochoterenae Taylor

Current allocation

= E. brevirostris brevirostris

— E. brevirostris bilineatus

— E. brevirostris dicei

New Combination
= E. brevirostris pineus
New Combination
= E. brevirostris indubitus
New Combination
= E. colimensis

— E. copei

— E. dugesii

— E. ochoterenae

Taylor (1936) indicated that E. copei belonged with the five-lined forms
of the anthracinus group, but was the least typical of the group. A primary
character of the brevirostris series group is the presence of a scale that is wider
than long following the postgenial scale. Eumeces copei has this condition,
while members of the anthracinus group have the scale following the postgenial
longer than wide. The dorsolateral light lines arise on the rostral in the brevi-
rostris group and in E. copei, but on the last supraocular in the anthracinus
group. A median dorsal light line is absent in the brevirostris group and E.
copei, but is present in the anthracinus group. As the principal characters of
E. copei ally the species to the brevirostris group, it is removed from the
anthracinus group.

The characters used to define the species and subspecies of E. dicei over-
lap to a considerable extent with features of populations of E. brevirostris from
the Sierra Madre Occidental ( bilineatus ), Oaxacan Highlands and Sierra
Madre del Sur (brevirostris) . The two subspecies of E. dicei are similar to
E. brevirostris bilineatus in most characters of squamation but differ in the
absence of an enclosed interparietal scale and in minor features of color pat-

4

Contributions in Science

No. 168

tern. The interparietal scale is enclosed in all specimens of bilineatus examined,
12 per cent of brevirostris, not in dicei, and in two per cent of pineus. In this
character dicei shows a closer relationship to the nominate form of brevirostris
found to the south of dicei than to bilineatus to the west.

The primary temporal scale is absent in all dicei, in 93 per cent of pineus,
in 60 per cent of bilineatus and in three per cent of brevirostris, indicating a
close relationship in eastern populations of brevirostris in this character.

The number of dorsal, longitudinal, and transverse scale rows, fourth toe
lamellae, superciliary scales, and postsubocular scales in dicei all overlap
among individuals of populations of brevirostris from western and southern
Mexico (Tables 1 & 2).

The dorsolateral light line is about one scale wide in brevirostris dicei
and is somewhat obscure at the shoulder. The line is about two scales wide and
usually distinct to the hind limb in pineus and bilineatus and about two scales
wide but variable in length in populations of brevirostris from central and
southern Mexico. The lateral light line is present in southern populations of
brevirostris but usually obscure in dicei, pineus, and some individuals of
bilineatus. The lateral dark field is about three scales wide in dicei, pineus,
bilineatus, and one and a half to two and a half scale rows wide in southern
and central populations of brevirostris. Eumeces b. dicei and E. b. pineus are
somewhat lighter in color than western and southern populations of
brevirostris.

The presence of four dark, dorsal longitudinal lines on the middle one-
third of the first and second scale row on each side serves to distinguish dicei
and pineus from all populations of brevirostris. The samples of brevirostris
have four to seven dark, dorsal longitudinal lines, but the lines are found
between the first, second, third, and infrequently the fourth scale rows. The
lines are occasionally absent or obscure in dicei and brevirostris.

The above data indicate that dicei and pineus share a combination of a
number of characters found in western, southern, and central populations of
brevirostris and are considered as members of E. brevirostris Rassenkreis.

Eumeces indubitus is essentially similar to E. brevirostris in all characters
of squamation. The only useful character for separating populations of E.
brevirostris and E. indubitus is the absence of a lateral light line from the
ear to the hind limb. The lateral light line is replaced on the side of the neck in
E. indubitus by a series of scales that have white anterior edges and centers
with black posterior borders, on longitudinal scale rows five, six, seven, eight,
and nine, or any combination of these rows in sequence. In addition, all samples
of E. indubitus with the above character have the interparietal enclosed
posteriorly by the parietals in 92 per cent or more of the specimens. However,
a population of E. brevirostris from the vicinity of Rio Frio, Mexico (state),
has the interparietal enclosed in 38 per cent of the sample, but has a lateral
light line on the neck on longitudinal scale rows six and seven. Where the
ranges of E. indubitus and E. brevirostris approach one another (between

1969

Mexican Skinks of E. BREVIROSTRIS Group

5

Cuernavaca and Tepoztlan, Morelos; a distance of 1 1 miles), some individuals
have enclosed interparietals and a faint lateral light line on the neck or a non-
enclosed interparietal without a lateral light line on the neck. The shifting of
characters between the two populations along a narrow zone of contact belies
the contention that they represent two distinct species, and 1 consider E. in-
dubitus conspecific with E. brevirostris.

Key to the Members of the EUMECES BREVIROSTRIS Group
The key to the species and subspecies of the E. brevirostris group must be
used with care. Skinks are notorious for being conservative, so conservative
that most scale characters utilized in the key overlap to some degree with these
features in every other species of the group. If doubt remains as to the identity
of the specimen at hand, the color pattern description, photographs, and maps
will supplement the key.

Key to the Eumeces Brevirostris Group

la. Three supraoculars dugesii

lb. Four supraoculars (occasionally three) 2

2a. Dorsolateral light line restricted to third scale row from middle of neck to

hind limb copei

2b. Dorsolateral light line on second and third or second, third and fourth

scale rows from middle of neck to midbody or hind limb. .. 3

3a. Longitudinal scale rows 26 to 28; fourth toe lamellae 1 5 to 17; toes overlap

when limbs adpressed along side of body colimensis

3b. Longitudinal scale rows 20 to 26; fourth toe lamellae 10 to 15; toes sep-
arated by five or more body scale when adpressed along side of body 4

4a. Transverse dorsal scale rows 49 to 57 (x = 53.7), maximum snout-vent
length 52 mm $ $ , 55 mm 9 $ ; primary temporal contacting lower sec-
ondary temporal in 45 per cent of population; dorsolateral light lines
separated from one another by two or two and two half-scale rows at

midbody ochoterenae

4b. Transverse dorsal scale rows 50 to 68 (x = 59.0), maximum snout-vent
length 70 mm $ $ , 73 mm 9 9 ; primary temporal contacting lower sec-
ondary temporal in less than one per cent of population; dorsolateral light
lines separated by three and two half-scale rows or four scale rows at

midbody brevirostris

Key to the Subspecies of Eumeces Brevirostris

la. Interparietal usually enclosed by parietals 6

lb. Interparietal usually not enclosed by parietals 2

2a. Primary temporal usually present 4

2b. Primary temporal usually absent 3

3a. Broad lateral dark stripe on fourth scale row of body (Nuevo Leon,

Tamaulipas b. dicei

6

Contributions in Science

No. 168

3b. Broad lateral dark stripe on third scale row of body (Nuevo Leon, Coa-

huila, Tamaulipas) b . pineus

4a. Lateral light line present from ear to hind limb (Guerrero, Oaxaca)

b. brevirostris

4b. Lateral light line obscure or absent 5

5a. Lateral light line from ear to arm on scale rows six and seven, bordered

below by thin black line (Mexico, Morelos, Puebla, Veracruz)

b. brevirostris

5b. No definite light line from ear to arm bordered below by thin black line
(longitudinal scale rows six, seven, eight, and nine may have light cen-
tered, dark scales) 6

6a. Primary temporal absent in 60 per cent of sample; no lateral light line on
neck; neck scales whitish from scale row five or six, grading into light belly
color; dorsolateral light line may extend to tail or beyond (Chihuahua,

Durango) b. bilineatus

6b. Primary temporal present in 96 per cent or more of sample; lateral light
line of neck consists of a series of light centered, dark scales on scale rows
five, six, seven, eight, and nine, or any combination of these in sequence;
dorsolateral light line usually faint at shoulder and beyond (Morelos,
Mexico, Jalisco) b. indubitus

Description of Species and Subspecies

Eumeces brevirostris brevirostris (Gunther, 1860)

Maximum snout-vent length 70 mm for $ $ , 73 mm $ 9 ; color pattern
consists of a dorsolateral light line usually beginning on rostral, passing poste-
riorly along the outer one-third of supraoculars, nuchals, lower one-third of the
first and upper one-third of the second longitudinal scale row, shifting to the
lower one-third of the second and most of the third longitudinal scale row at
middle of neck and/or anterior one-third of body, fading posteriorly or distinct
to the hind limb; dorsal dark lines may be distinct and continuous, dashed,
spotted, obscure, or absent, but if the former, usually present between the
paravertebral rows and between longitudinal scale rows one and two, two and
three, and sometimes three and four posteriorly; upper secondary dark line
begins on the prefrontals, passing posteriorly along middle one-third of nuchal,
lower edge of longitudinal scale row one and upper one-third of two to the level
of the arm or fading between arm and midbody; dorsal area between dark lines
light bronze to dark brown; primary lateral dark line (stripe) usually begins on
upper edge of nostril, passing through side of head to level of upper edge of
supralabials, posteriorly along the upper edge of post suboculars and above
ear, on lower edge of longitudinal scale row two through upper edge of six on
neck, on lower one-third of third through upper half of fifth row at midbody;
lateral light line may be long (Oaxaca, Guerrero, Figs, la and b) or short
(Mexico, Morelos, Puebla, Veracruz, Distrito Federal, Figs, lc and d), but

1969

Mexican Skinks of E. BREV1ROSTRIS Group

7

always present from the supralabials to arm on fifth and sixth or sixth and
seventh longitudinal scale rows of neck and in southern populations from the
arm to hind limb on fifth, fifth and sixth, or sixth longitudinal scale rows at
midbody; lateral light line bordered below by lower secondary dark line from
the lower edge of the supralabials through the neck and to the hind limb in
southern populations.

Miss Alice G. C. Grandison ( pers . comm.) kindly supplied me with in-
formation on the holotype of E. brevirostris (Fig. 2) in the British Museum
(Natural History): “There is a dorsolateral light line on the lateral half of
scale row two and median half of scale row three, and lateral to this line there
is a broad dark band which in width extends from the lateral half of the third
scale row to the median half of the sixth row. The light line referred to extends
from the tip of the snout along the lateral edge of the supraoculars, the lateral
half of the parietal and the nuchal to become a dorsal lateral stripe. The line is
dark edged medially, the edge occupying a small blob in approximately the
central area of the second scale row. I confirm that there is evidence of a
lateral light line from the upper lip, over the shoulder, to the hind limb.” This
color pattern is typical for individuals of brevirostris from Oaxaca (Fig. 2) .

Squamation : The number of supraoculars contacting the frontal is usually
three on each side but occasionally may be two on each side or two on one
side and three on the other; number of supraoculars usually four on each side
but occasionally may have three on each side (one specimen of 102) or three
on one side and four on the other (two specimens). This condition appears to
be a fusion of one supraocular with another rather than the loss of one. One
specimen from Oaxaca has the fourth supraocular of one side transversely
divided, producing five supraoculars on that side.

The amount of contact of the seventh supralabial with the upper second-
ary temporal is relatively constant but the number of scales in the superciliary
series and the enclosure of the interparietal are variable (Table 1). The
presence of a primary temporal is relatively constant, but one specimen has the
primary temporal fused to the upper secondary temporal on both sides and
another individual has the upper secondary temporal vertically divided on
one side.

The mean number of dorsal transverse scale rows is nearly constant for
samples for Veracruz, Puebla, and Mexico-Morelos, while those from Oaxaca
and Guerrero are lower, but a statistical analysis of the five samples reveals a
considerable overlap of variation between samples. The mean number of
longitudinal scale rows around the body is lower in the Oaxaca sample, while
the remaining four samples are relatively uniform. The mean number of
lamellae beneath the fourth toe are identical in the Oaxaca and Guerrero
samples and increases slightly to the north and west in the remaining samples
(Table 2).

A statistical analysis of the squamation characteristics in populations of
b. brevirostris and b. indubitus indicates no differences (Table 2). The color

8

Contributions in Science

No. 168

Figure 1. A dorsolateral view of the color pattern of Eumeces brevirostris brevirostris.
A. Tejocotes, 7800 ft, Oaxaca; B. 2 mi SW Omilteme, 7900 ft, Guerrero; C. near
E. Seco, 7100 ft, Puebla; D. 23.9 mi NW Toluca, 7600 ft, Mexico. Note the shortening
of the lateral light line in C and D.

1969

Mexican Skinks of E. BREVIROSTRIS Group

9

Figure 2. Holotype of Eumeces brevirostris Gunther, from Oaxaca, Mexico. By per-
mission of the Trustees of the British Museum ( Natural Flistory ).

10

Contributions in Science

No. 168

pattern of individuals intermediate between the two populations is discussed in
the preceding section.

Localities, museum depository, number of specimens examined, and
pertinent literature records: GUERRERO ; Omilteme, FMNH, 11, UIMNH,
5; 2 mi SW Omilteme, TCWC, 23; 5 mi W Mazatlan, TCWC, 2; 4 mi W
Mazatlan, TCWC, 3; 3 mi W Omilteme, MVZ, 1. MEXICO ; Rio Frio, KU,
12, UMMZ, 7; Rio Frio and road from Puebla to Mexico, UMMZ, 4; 23.9 mi
NW Toluca, MVZ, 1 ; Ajusco Mountains, between Mexico City and Cuerna-
vaca, FMNH, 1. MORELOS', Laguna de Zempoala, UMMZ, 2; near Huajint-
lan, km 133, FMNH, 1; Tepozatlan, FMNH, 1. OAXACA', Rancho Mira
Cumbre, UMMZ, 1; 32.5 mi W Oaxaca, UMMZ, 1; 24 km NNE Oaxaca,
UMMZ, 2; 27 km NNE Oaxaca, UMMZ, 2; Oaxaca, UIMNH, 2, FMNH, 1;
La Parada, USNM, 1; 3 km W San Vincente Lochixio, KU, 1; 18 km N, 33
km W Oaxaca, TCWC, 1; between Pluma Hidalgo and Copalitos, AMNH,
2; Pluma Hidalgo, AMNH, 1; near continental divide, 3 mi N Tierra Colo-
rada, AMNH, 4; 5 mi S Llano de las Flores, Sierra Juarez, AMNH, 3; 2 mi E
Ixtlan de Juarez, AMNH, 2; El Punto, AMNH, 11; 1 mi S El Punto, near
Cerro San Felipe, AMNH, 1; Cerro San Felipe, UIMNH, 13; north slope of
ridge E Cerro San Felipe, AMNH, 8; Tejocotes, AMNH, 2; 6 mi N Cumbre
del Estudiante, AMNH, 2; 3.5 mi N Cumbre del Estudiante, AMNH, 5; south-
west ridge of Zempoaltepetl, AMNH, 1; 11 mi S Sola de Vegas, AMNH, 1;
northeast of San Andres Chicahuaxtla, AMNH, 14, UIMNH, 2; 9.8 mi NE
San Juan del Estado, AMNH, 1; Lachixorax, above Aseradero, UIMNH, 1;

8 mi SE Nochixtlan, UIMNH, 5; 9 mi SE Nochixtlan, AMNH, 1; 13 mi NE
Nochixtlan, UIMNH, 1; San Jose Lachiguiuri (Taylor, 1943); Tehuantepec
(Smith and Taylor, 1950). PUEBLA ; Pajaro Verde, UIMNH, 1; 24 km W
Texmelucan, USNM, 3; Laguna San Bernardino, FMNH, 4, UIMNH, 1; near
El Seco, km 205, UIMNH, 4, FMNH, 4: near Acultzingo, FMNH, 1 ; 4 mi NE
Esperanza, Sierra Negra, UMMZ, 1; near Rio Frio, UIMNH, 1 1; east of Rio
Frio, FMNH, 4, UIMNH, 1; 5 mi E Rio Frio, MVZ, 1 ; 6 mi E Rio Frio,
MVZ, 1. VERACRUZ', 3 km W El Limon, KU, 1; San Antonio Limon, To-
talco, UIMNH, 3, FMNH, 2; Totalco, FMNH, 3; near Totalco, UIMNH, 8;
FMNH, 12; Perote, USNM, 1; Orizaba, USNM, 2; Laguna (Smith and Taylor,
1950).

Eumeces brevirostris brevirostris X brevirostris indubitus : MORELOS ;
km 58, road to Cuernavaca, FMNH, 1 ; 5 km N Tres Marias, UIMNH, 1 ; 5 mi
E Cuernavaca, MVZ, 1.

Eumeces brevirostris indubitus Taylor, 1933

The maximum snout-vent length for this subspecies is 69 mm $ $, 71 mm

9 9 ; color pattern consists of a dorsolateral light line beginning on rostral,
passing posteriorly along the outer half of supraoculars, median one-third of
nuchal, outer two-thirds of second longitudinal scale row, descending to outer
one-third of second and upper two-thirds of third row at shoulder, fading

1969

Mexican Skinks of E. BREVIROSTRIS Group

I 1

posteriorly and obscure at midbody; dorsal dark lines usually obscure but, if
distinct, similar to the pattern found in the nominate subspecies; upper second-
ary dark line begins on prefrontals and terminates between head and shoulder,
usually found on upper one-third of second longitudinal scale row but some-
times including the outer edge of the first row; primary lateral dark line (stripe)
is brownish black and begins on or near nostril, covering side of head from
level of supralabials and sometimes including most of latter, passing posteriorly
through ear, on lower one-third through upper half of sixth row on neck (upper
edge of seventh, eighth, and ninth rows often black), on lower three-fourths
of fourth through sixth scale row at midbody; lateral light line represented by
narrow white line on supralabials, by a series of light centered, dark edged
scales on sixth and seventh, sometimes eighth and ninth rows on neck, and
usually absent on the side of the body; lower secondary dark line usually united
with primary lateral dark line, i.e., absent (Fig. 4a).

Ten individuals of 41 from Morelos and Mexico have a faint indication
of a lateral light line, represented by a series of light centered, dark scales on
the fifth and sometimes the sixth longitudinal scale row at midbody, lending
further support to the conspecific status of indubitus and brevirostris.

The color pattern of the sample from Jalisco is similar to the Morelos and
Mexico sample, but usually lacks the lateral light line on the supralabials (Fig.
4b). The upper secondary dark line is longer, extending posteriorly to midbody
in most specimens. The lateral light line is represented by light centered, dark
seals on the seventh and eighth scale rows on the neck rather than the fifth,
sixth, seventh, eighth, or ninth, or any combination in sequence. Juveniles and
sub-adults have similar color patterns but usually more distinct lines and
contrasting colors.

Squamation : Variation in squamation among three samples of b. indubi-
tus (Mexico, Morelos, and Jalisco) is presented in Tables 1 and 2. Minor
differences occur in the number of longitudinal scale rows and lamellae be-
neath the fourth toe between the Jalisco samples and those farther east. The
ratio of juveniles and adults of the Jalisco sample having 26 longitudinal scale
rows are 39 and 47 per cent, respectively, out of a sample of 51 individuals.
The mean number of transverse dorsal scale rows varies less than one scale
between all three samples. The number of supraoculars are four on each side
in all three samples; number of supraoculars contacting the frontal are usually
three on each side, but occasionally two contact the frontal on each side (7
per cent) or two on one side and three on the other (5 per cent) .

The head scales of one individual from Talpa, Jalisco, are aberrant, having
the upper secondary temporal vertically divided on both sides, forming what
appears to be two primary temporals in tandem. This, in effect, allows a con-
tact between the anterior half of the split upper secondary temporal and the
lower secondary temporal. Fusion of the primary temporal with the upper
secondary temporal occurs on one side in two specimens and on both sides
in a third.

12

Contributions in Science

No. 168

A single specimen from Puerto Hondo, near Zitacuaro, Michoacan, is
listed by Taylor (1936) and Duellman ( 1961 ) as E. indubitus. This individual,
FMNH 106577, is a specimen of E. dugesi, apparently mis-identified by Taylor
and Duellman because of its small size (juvenile, 36 mm in snout-vent length).
The specimen clearly has three supraoculars on each side and the color pattern
of E. dugesii.

Localities: JALISCO ; 7 mi W Atenquique, LACM, 30, MSU, 1, KU, 2;
6.4 mi W Atenquique, KU, 3; 8 km SE Jasmin, UMMZ, 3; 0.5 mi SE El Isolte,
UMMZ, 1 ; 0.5 mi N Garcia, UMMZ, 1 ; 10 mi W Cuidad Guzman, UMMZ, 1 ;
3 km NE Talpa, KU, 1; Talpa, Mascota, FMNH, 1; La Cumbre de los Arras-
trados, BM(NH), 5; Sierra de Jaunocatlan, La Laguna (Tanner, 1958).
MEXICO ; near Asuncion, 40 mi NE Zitaquaro, USNM, 1, UIMNH, 4,
FMNH, 9; Tenancingo, FMNH, 1. MORELOS', 5 km S Tres Marias (Tres
Cumbres), FMNH, 1, UIMNH, 2; km 61, near Cuernavaca, FMNH, 3,
UIMNH, 2; km 63, near Cuernavaca, FMNH, 4, UIMNH, 2 (type locality);
km 58, near Cuernavaca, FMNH, 3, UIMNH, 2; 7.5 mi N Cuernavaca, MVZ,
1; Cuernavaca, UIMNH, 1, USNM, 1; Zempoala, UIMNH, 1.

Eumeces brevirostris bilineatus Tanner, 1958
The maximum snout-vent length for b. bilineatus is 60 mm for $ $ and
$ $ ; color pattern consists of dorsolateral light line beginning on rostral, pass-
ing posteriorly along outer one-third of supraoculars, median one-third of
nuchals, most of second longitudinal row on anterior neck, descending to
lower half of second and upper half of third row to anterior one-third of body,
descending to third row for remainder of body; dorsolateral light line may in-
volve fourth scale row in some specimens either as a series of light centered,
dark scales, or upper half of fourth light colored from arm to hind limb; dor-
solateral light line variable in length, may be distinct to arm, midbody, or tail,
obscure in some old adults; upper secondary dark line begins on rostral or
internasal, passing posteriorly along inner one-third of supraoculars, nuchals,
and upper half of second scale row for 8 to 15 scales beyond shoulder; dark
dorsal lines usually obscure or absent but, if present, found between each
scale row of the dorsum between primary lateral dark line; primary lateral dark
line begins at nostril, covering side of head from edge of supralabials to super-
ciliary series, passing posteriorly through upper half of ear, on lower half of
third through upper half of sixth row of neck, and on all or part of fourth
through upper half of sixth row at midbody; upper edge of primary lateral
dark line may be partly divided beyond arm (Fig. 4c) on lower edge of third
and upper one-third of fourth row, a condition often found in brevirostris
pineus’, lateral light line usually obscure or absent but, if present, found on
sixth and seventh rows of neck or lower half of seventh row of neck and lower
half of sixth row at midbody (Fig. 4c).

Squamation : Variation in two samples of b. bilineatus (Durango and
Chihuahua) is given in Tables 1 and 2. The presence or absence of the primary

1969

Mexican Skinks of E. BREVIROSTRIS Group

13

Figure 3. Map of the distribution of Eumeces brevirostris brevirostris, E. brevirostris
indubitus, E. copei, and E. ochoterenae in central and southern Mexico.

temporal is variable, occurring on both sides in 33 per cent, one side only in
53 per cent of a sample of 48 specimens. A comparison of characters between
the two samples indicates that they are similar, except for the number of scales
in the superciliary series. There appears to be a reduction in the number of
superciliary scales from south to north in the various populations of breviros-
tris, with the Oaxaca sample having seven scales on each side in 53 per cent,
and the Chihuahua sample having none with seven but 63 per cent with six on
each side. A similar reduction is found in the number of post suboculars. The
enclosed interparietal, number of supraoculars, and seventh supralabial —
upper secondary temporal contact is constant for both samples. There are usu-
ally three supraoculars contacting the frontal on each side, but occasionally
only two on each side (in one) and two on one side, three on the other (in two) .

Localities: CHIHUAHUA ; Mojarachic, UMMZ, 1; 15 mi S, 6 mi E
Creel, KU, 2; 2 mi W Samachique, KU, 3; 7 mi SE El Vergel, MVZ, 1 ; 3 mi
NE Rio Verde (town), MVZ, 1 ; 7 mi SW Lagunita, MVZ, 1. DURANGO',

14

Contributions in Science

No. 168

2.5 mi W San Luis, UMMZ, 1 ; 33 mi W El Salto, 1 mi W Buenas Aires, CSLB,
7; 9 mi SW El Salto, KU, 3; 10 mi SW El Salto, KU, 2; 15 mi SW El Salto, KU,
1; Coyotes, FMNH, 3; near Hacienda Coyotes, UMMZ, 1; 6 mi WSW Los
Adjuntas, MVZ, 13; Laguna del Progresso, UMMZ, 4; Cuidad, BM(NH), 2;
loquiro, Tarahumari, BM(NH), 1; Highway 40, 137 mi E junction of High-
ways 40 and 15, AMNH, 1; Cueva (Webb and Baker, 1962); 1 mi ESE
Cajones (Webb and Baker, 1962).

Eumeces brevirostris pineus Axtell, 1960

The maximum snout-vent length for b. pineus is 60 mm $ $ , 66 mm $ $ ;
color pattern consists of a dorsolateral light line beginning on the rostral or
internasal, passing posteriorly along outer one-third of supraoculars, nuchals,
upper half of second and lower two-thirds of third scale row to hind limb; upper
one-third of fourth scale row with series of whitish dots from the neck to hind
limb, bordered above by thin black line along outer edge of third scale row;
dorsal dark lines usually present on median one-third of first and second scale
rows, with dark line of the second row representing the upper secondary dark
line; dark dorsal lines may be present as a series of dashes or spots, but usually
distinct from rear of head to level of hind limb; primary lateral dark line
begins at nostril, covering side of head above supralabials, passing posteriorly
over ear, on lower half of third through upper one-third of sixth of neck, lower
one-third of third through upper one-third of sixth row at midbody; lateral
light line usually absent but, if present (one specimen), represented as a series
of light centered, dark scales on sixth and seventh scale rows of neck and
seventh row at midbody; lower secondary dark line usually absent; head
brownish with dorsal area between dorsolateral light lines light bronze (Fig.
4d) .

Squamation : Variation in seven characters of squamation is given in
Tables 1 and 2. Eumeces b. pineus differs from b. indubitus and b. brevirostris
in the absence of a primary temporal scale in 93 per cent of the sample ex-
amined. However, Eumeces b. bilineatus exhibits a loss of the primary tem-
poral in 60 per cent and E. b. dicei 100 per cent loss in the samples examined.

The number of supraoculars, post suboculars, seventh supralabial —
upper secondary temporal contact is almost constant in b. pineus. The number
of supraoculars contacting the frontal is usually three on each side but occa-
sionally only two contact the frontal on each side (in two specimens), or two
on one side and three on the other (in one).

Localities: COAEIUILA; 13 mi E San Antonio de las Alazanas, KU, 18,
FMNH, 3, AMNH, 1; 7 mi E San Antonio de las Alazanas, AMNH, 2.
NUEVO LEON\ Ojo de Agua, near Galeana, FMNH, 1 ; Cerro Potosi, FMNH,
4, AMNH, 2; Pablella Galeana, FMNH, 4, UIMNH, 5.

Localities for intergrades between b. pineus x b. dicei : TAMAULIPAS ;
Rancho Ceilo, LACM, 1 ; El Chique, UMMZ, 1 ; La Joya de Salas, UMMZ, 8;

1969

Mexican Skinks of E. BREVIROSTRIS Group

15

Figure 4. A dorsolateral view of the color patterns of some subspecies of Eumeces
brevirostris. A. indubitus, 61 km, near Cuernavaca, Morelos; B. indubitus, 9 mi W
Atenquique, Jalisco; C. bilineatus. Chihuahua; D. pineus , Ojo de Agua, Nuevo Leon;
E. died. Santiago, Nuevo Leon.

16

Contributions in Science

No. 168

5 mi NW Gomez Farias; 7 mi WNW Gomez Farias (Martin, 1958; Axtell,
1960).

Eumeces brevirostris dicei Ruthven and Gaige, 1933

The maximum snout-vent length for Eumeces b. dicei is 60 mm $ $ ,
61 mm $ 9; color pattern (Fig. 4e) similar to E. b. pineus, with following
exceptions: dorsolateral light line thin, usually confined to lower edge of
second and upper two-thirds of third scale row from neck to arm, fading
posteriorly into light bronze color of dorsum; primary lateral dark line begins
at nostril, usually covering side of head but brownish in color, passing over
ear, on lower edge of third through upper half of sixth scale row of neck,
lower half of fourth through upper three-fourths of sixth scale row at midbody,
but may vary half a scale row in either direction in some specimens.

Axtell (1960) examined five specimens of E. b. dicei from Tamaulipas
and indicated that the dark dorsal lines were usually present between the
dorsolateral light lines as a series of linear dark pigment spots. A series of six
specimens from Nuevo Leon lack the dark pigment between the dorsolateral
light lines except for the upper secondary dark line that occurs from the first
supraocular to the parietal in one, to the level of the arm in two and only on
the supraoculars in one.

Some of the characters used by Axtell (1960) to separate the two sub-
species are variable in the series at hand. The average snout-vent length given
by Axtell for dicei and pineus was 50.7 and 59.3 mm, respectively. The present
samples reveal an average snout-vent length for dicei and pineus as 61.0 and
60.2 mm, respectively. There is considerably more overlap in the variation of
position of the primary lateral dark band in each of the two samples than was
noted by Axtell. However, the frontonasal-frontal contact and length of limbs
serve as distinguishing features for each population.

Based upon intermediate conditions in coloration, the Santiago, Nuevo
Leon, sample indicates that this area may be one of secondary intergradation.
Unfortunately, most scale characters overlap between the two subspecies and
only a statistical trend may indicate which population is at hand. Eumeces b.
pineus usually has 22 longitudinal scale rows and 24 per cent have contact
between the frontonasal and the frontal; E. b. dicei usually has 24 longitudinal
scale rows and 90 per cent have frontonasal-frontal contact. The Santiago
sample is closer to dicei than to pineus in the latter characters. If this area is a
zone of secondary intergradation, then the suppression of pineus characters
proposed by Axtell (1960) in areas of secondary intergradation also occurs in
the Santiago sample.

Squamation : Variation in squamation of 1 1 specimens is given in Tables
1 and 2 (see Axtell, 1960, for detailed comparison of dicei and pineus).

Localities: NUEVO LEON; Santiago, Hacienda Vista Hermosa, FMNH,
6. TAMAULIPAS; 3 mi S Acuna, UMMZ, 2; near Acuna, UMMZ, 1; 12 mi
NNW Chamal, UMMZ, 1; Marmolejo (Ruthven and Gaige, 1933).

1969

Mexican Skinks of E. BREVIROSTRIS Group

17

Eumeces dugesii Thominot, 1883

The maximum snout-vent length for this species is 66 mm $ $ , 69 mm
9 $ ; color pattern consists of dorsolateral light line beginning on rostral, pass-
ing posteriorly along outer half of supraoculars, outer one-third of nuchals,
outer half of second longitudinal scale row for three to nine scales, descending
to outer one-third of second and inner one-third of third for three to six
scales, descending to upper half of third row for 9 to 16 scales, fading posterior-
ly to arm (occasionally descending to fourth row in some specimens and infre-
quently extending to the hind limb); mid-dorsal area light bronze to copper;
upper secondary dark line begins on first supraocular, passing posteriorly
through median one-fourth of nuchals, terminating on fourth transverse row
of scales, faintly indicated on third longitudinal scale row at midbody (Fig.
8b) ; primary lateral dark line covers side of head from supraoculars to edge of
supralabials, lower half of second through upper half of sixth longitudinal scale
rows of neck, lower half of third through upper half of sixth row at midbody;
lateral light line from first supralabial through ear to above arm, on lower half
of sixth and upper half of seventh longitudinal scale row of neck, becoming a
series of light spots on lower half of sixth longitudinal scale row at midbody;
lower secondary dark line on lower half of seventh longitudinal scale row of
neck, all of seventh at midbody; posterior edges of infra- and supralabials with
vertical or slightly oblique black bars; limbs dark brown to black above, belly
yellowish below, sometimes bluish gray to light olive.

Squamatiotv. Variation in squamation for 27 specimens of E. dugesii is
given in Tables 1 and 2. The most unusual character of E. dugesii is the pres-
ence of three supraoculars. The loss of the first supraocular appears to be due
to a fusion of the enlarged first superciliary with the first supraocular. There is
wide variation in the number of scales in the superciliary series (Table 1 ) . The
primary temporal is absent on both sides in one specimen, absent on one side
in two. A small azygous scale is present between the frontonasal and the frontal
in two specimens. The number of supraoculars contacting the frontal are two
on each side without exception.

Localities: JALISCO ; 3 mi WSW Mazamitla, KU, 1; Nevado de Colima
(Smith and Taylor, 1950). MICHOACAN-, 2 mi W San Jose, TCWC, 1;
Rancho San Jose, USNM, 1; 12 mi S Carapan, KU, 5; Puerto Hondo, FMNH,
1 ; 9 mi W Zacapu, near Cerro Tecolote, UIMNH, 1 ; El Soledad, Tancitaro,
FMNH, 2; 11 mi S Patzcuaro, AMNH, 4; 1.7 mi W San Jose de la Cumbre,
UMMZ, 1; 7.2 mi N Cheran, UMMZ, 1; 4.5 mi WNW Cheran, UMMZ, 4;
2.3 mi N Opopeo, UMMZ, 2; Uruapan, Parque Nacional, UMMZ, 1; near
Carapan (Taylor, 1943); Tangancicuaro, Patamban, Apatzingan (Smith and
Taylor, 1950).

Eumeces copei Taylor, 1933

The maximum snout-vent length for E. copei is 66 mm $ $,13 mm
9 9 ; color pattern consists of a dorsolateral light line beginning on rostral,

18

Contributions in Science

No. 168

passing posteriorly along outer third of supraoculars, nuchals, most of second
longitudinal scale row for one to three transverse rows, descending to middle
third of third longitudinal scale row to tail; upper secondary dark line begins
on outer edge of rostral, passing posteriorly along inner half of supraoculars,
nuchals, upper edge of second longitudinal scale row for two to ten transverse
scales, descending to upper one-third of third longitudinal scale row to the tail;
dorsal dark lines present between paravertebral rows and one each between
longitudinal scale rows one and two; dorsal coloration between upper second-
ary dark lines bronze to dark brown; primary lateral dark line covers side of
head from supraoculars to supralabials, on lower half of third through upper
half of sixth longitudinal scale row of neck, shifting to lower one-third of third
through upper half of fifth longitudinal scale row at midbody; lateral light line
present from lower edge of first supralabial to tail, on lower two-thirds of sixth
through upper half of seventh longitudinal scale row of neck, lower half of
fifth through upper tip of sixth longitudinal scale row at midbody (occasionally
present only on the fifth (in 17 specimens), sixth (in seven), sixth and seventh
(in one) and fifth, sixth, and seventh (in three) longitudinal scale rows at mid-
body) ; lower secondary dark line begins on lower edge of last infralabial, lower
one-third of seventh through upper tip of eighth longitudinal scale row of
neck, lower half of sixth through upper third of seventh longitudinal scale row
at midbody (Fig. 8a).

Squamation : Variation in squamation of 125 specimens of E. copei is
given in Tables 1 and 2. Three specimens have only three supraoculars on each
side and one has three on one side, four on the other. The lower number of
supraoculars in some specimens of E. copei overlaps with the number of supra-
oculars present in E. dugesii. However, the dorsolateral light line is always
present only on the median third of the third longitudinal scale row from the
neck to the tail in copei and is never restricted to the third scale row in dugesii.

Taylor (1933) stated that the seventh supralabial always contacts the
upper secondary temporal in copei, thereby separating the primary temporal
from the lower secondary temporal. In the series of copei at hand, the primary
temporal scale contacts the lower secondary temporal on one side in ten speci-
mens, on both sides in eight specimens, consisting of approximately 10 per
cent of the total sample examined (125 specimens).

The frontonasal does not contact the frontal in 22 specimens, 18 of which
have the prefrontals in contact with each other; the remaining four have the
frontonasal transversely divided, forming a small azygous scale between the
frontonasal and the frontal.

Usually three supraoculars contact the frontal but occasionally only two
contact the frontal on each side (in three specimens) or two on one side and
three on the other (six specimens).

Localities: DISTRITO FEDERAL ; Monte Alegre, Sierra Ajusco,
AMNH, 1; Santa Rosa, UMMZ, 1 ; 1 mi SW Santa Lucia, AMNH, 1. MEXI-
CO, Nevado de Toluca, AMNH, 34; Rio Frio, TCWC, 3, UIMNH, 13,

1969

Mexican Skinks of E. BREVIROSTRIS Group

19

USNM, 1 ; 5 km W Rio Frio, UIMNH, 3, USNM, 9; 13 km W Toluca, UMMZ,
1 ; 10 mi N, 6 mi E Valle de Bravo, KU, 3; 24 km NW Toluca, MVZ, 7; Villa
Victoria, UIMNH, 2; 8 mi W Villa Victoria, UIMNH, 1; Toluca, USNM, 2;
10 mi SE Asuncion, USNM, 1; 15 km W Toluca, USNM, 13; Tenancingo
(Taylor, 1936); 3 mi W Zinacantepec (Taylor, 1943). MICHOACAN\ Tan-
citaro, FMNH, 3; 9 mi W Zacapu, UIMNH, 1. MORELOS : Lagunas de Zem-
poala, TCWC, 9; UIMNH, 31, AMNH, 21, USNM, 19; 2 mi W Huitzilac,
TCWC, 5; 3 km NW Tes Cumbres, TCWC, 2; 5 km N Tres Cumbres, TCWC,
17, KU, 2; 5 km S Tres Marias, UIMNH, 1; 6 km NW Huitzilac, KU, 2.
VERACRUZ ; Potrero Viejo. UIMNH, 1. GENERAL ; Overland Route,
USNM, 1.

Eumeces ochoterenae Taylor, 1933

The maximum snout-vent length for this species is 52 mm $ $ , 55 mm
$ $ ; ochoterenae is the smallest species in the krevirostris group, averaging
48.3 mm $ $ and 49.4 mm $ 2 in snout-vent length; color pattern consists of
a dorsolateral light line beginning on rostral, passing posteriorly along outer
half of supraoculars, nuchals, outer two-thirds of second longitudinal scale
row for two to six transverse scales, descending to lower half of second
through upper half of third longitudinal scale row to tail; upper secondary
dark line usually from nuchals to tail along upper half of second longitudinal
scale row, but often faint from midbody to tail; dorsal dark lines are absent,
color between upper secondary dark lines brown, bronze, or dark brown; pri-
mary lateral dark line covers side of head from supralabials to supraoculars, on
lower half of third through upper half of sixth longitudinal scale row of neck,
on lower half of third through upper two-thirds of fifth longitudinal scale row
at midbody; lateral light line faintly indicated on most specimens, consists of a
series of light colored spots on scales of sixth and seventh or seventh and
eighth longitudinal scale rows of neck and as a series of light centered, dark
edged scales on the sixth longitudinal scale row at midbody (Fig. 8c); venter
yellowish, tail blue.

Squamation : Variation in squamation of 39 specimens of E. ochoterenae
is given in Tables 1 and 2. The interparietal is enclosed in two of 39 specimens.
The seventh supralabial contacts the upper secondary temporal on both sides
in 15 specimens, one side only in 13 specimens. Number of supraoculars con-
tacting the frontal is usually three on each side but occasionally two contact the
frontal on each side (two specimens) or two on one side and three on the other
(three specimens). The primary temporal is absent on both sides in two
specimens and on one side in two specimens. The longitudinal scale rows
usually number 22 around the body, but 20 rows occur once and 24 rows occur
eight times.

Localities: GUERRERO ; Acahuizotla, TCWC, 1; Agua del Obispo,
FMNH, 6, UIMNH, 2, TCWC, 1; near Agua del Obispo, FMNH, 3; 1 mi N
Agua del Obispo, UMMZ, 1; near Chilpancingo, FMNH, 2; 12 mi S Chil-

20

Contributions in Science

No. 168

pancingo, FMNH, 2, UMMZ, 1 ; 3 to 4 mi S Chilpancingo, FMNH, 10,
UIMNH, 4; Chilpancingo, UIMNH, 1; 3 mi W Chilpancingo, TCWC, 1; 7 mi
E Chilpancingo, 4 mi N Chilpancingo (Taylor, 1943); between Rincon and
Cajones, UIMNH, 1, USNM, 1, AMNH, 1; Chilapa, USNM, 1; Mazatlan,
UIMNH, 1.

Eumeces colimensis Taylor, 1935

The maximum snout-vent length of E. colimensis is 65 mm $ $ (holo-
type), unknown in $ 8 ; color pattern of holotype consists of a dorsolateral
light line from rostral to posterior third of body, on outer half of supraoculars,
inner tip of upper secondary temporal, median third of outer half of nuchals,
most of second longitudinal scale row for two transverse scales, outer third of
second and all of third longitudinal scale row for seven transverse scales,
descending to lower half of third and most of fourth longitudinal scale row
for 28 transverse scales, and in addition to the above, upper third of fifth
longitudinal scale row for 19 transverse scales; upper secondary dark line
extends from frontonasal to shoulder on upper half of second longitudinal
scale row to middle of neck, descending to upper tip of third longitudinal scale
row to midbody; anterior third of each dorsal scale row between dorsolateral
light lines dark brown but not forming definite dorsal dark lines (Fig. 8d);
lateral light line on seventh and eighth longitudinal scale rows of neck repre-
sented at midbody by a series of light spots on the posterior half of the seventh
longitudinal scale row; primary lateral dark line from nostril, covering side of
head from edge of supralabials to supraoculars, on lower half of third through
upper half of sixth longitudinal scale row of neck, lower three-fourths of fifth
through anterior half of seventh longitudinal scale row at midbody; venter
whitish.

Two additional juvenile specimens from Michoacan are similar to the
holotype in the position of the light and dark lines of the head and body, but
the lateral light line is faint in one and obscure in the other and is not evident at
midbody in either specimen.

Squamation : Variation in squamation of three specimens of E. colimensis
is given in Tables 1 and 2. The two juvenile specimens from Michoacan differ
from the holotype in having the number of scales in the superciliary series
eight on each side in one, eight on one side and nine on the other in one (holo-
type with six on each side), 26 scale rows around body rather than 28; seventh
supralabial not in contact with upper secondary temporal rather than in con-
tact; primary temporal present on both sides rather than fused to the upper
secondary temporal. However, the variation in the above characters falls
within the range of variation for other species of the brevirostris group where
larger samples are available.

The limbs are long in all three specimens, toes overlapping when adpressed
along the sides of the body; toes relatively long with a corresponding increased
number of lamellae beneath the fourth toe.

1969

Mexican Skinks of E. BREVIROSTRIS Group

21

One additional specimen of Eumeces (KU 44733) from 1 mi E Santa
Lucia, Sinaloa, cannot be placed with E. colimensis with certainty, although
Webb (1959) recognized the individual as being E. colimensis. The major dif-
ferences between the Sinaloan specimen and E. colimensis are: (1) presence
of a well defined lateral light line from supralabials to tail (Fig. 8e) rather than
obscure; (2) dorsolateral light line on lower half of second and upper two-
thirds of third longitudinal scale row to tail, rather than descending to third,
fourth and fifth, as in colimensis; (3) presence of a small primary temporal,
with a large seventh supralabial contacting upper secondary temporal. The
primary temporal is fused with the upper secondary temporal in the holotype,
allowing a contact between the seventh supralabial and upper secondary
temporal, but the seventh supralabial is small, only slightly larger than the
sixth supralabial.

The Sinaloan specimen may be related to E. callicephalus, especially those
populations lacking the postnasal, with bifurcating lines on the head in old
adults and having a single postmental.

Localities: COLIMA ; Colima, FMNH, 1. MICHOACAN\ 0.5 mi NE
Coalcoman, UMMZ, 1; 1 mi ESE Estopilas de Salitre, UMMZ, 1. SINALOA ;
1 mi E Santa Lucia, KU, 1.

Discussion

Evolution of E. Brevirostris

All populations of Eumeces brevirostris are found in pine-fir or pine-oak
woodland communities between elevations of 1,800 and 11,000 feet in the
mountain ranges of Mexico. Orogeny during Pliocene and Pleistocene times
(volcanic upheaval and associated rise and fall of land mass along fault lines)
and climatic shifts probably accounts for the disruption of the once continuous
geographic range of the original parental stock of E. brevirostris. Since these
times five recognized subspecies have become established.

The northern populations ( bilineatus of northwestern Mexico, dicei and
pineus of northeastern Mexico) are more similar to each other in having a
reduced number of head scales and the loss or fusion of the primary temporal
to the upper secondary temporal scale, than to southern and west central popu-
lations of E. b. brevirostris and b. indubitus. Two possible explanations for a
similar reduction of head scales in two widely separated populations of north-
ern brevirostris are: (1) the northern populations exhibited some head scale
reduction prior to the climatic and physical changes produced during initial
glaciation or before, these environmental changes restricted the two popula-
tions to their present habitats on each side of the Mexican Plateau; (2) parallel
evolution in the reduction of head scales has taken place in these populations
after they were separated during glacial minima.

Martin (1958) indicated that there may have been an oak-savanna corri-
dor across the plateau during the Late Pleistocene that has been eliminated in
the arid interior at the present time. He listed several plants and animals that

22

Contributions in Science

No. 168

occur on either side of the plateau to support this idea. However, one also may
assume that the entire plateau may have been oak-savannas during glacial
maxima and the similar eastern and western plateau plants and animals are
relicts of once wide-spread transplateau populations.

The absence of an enclosed interparietal, marked similarity of color pat-
tern, and increase in number of head scales in brevirostris populations from
Coahuila to eastern Oaxaca indicate that a connection probably existed be-
tween the present allopatric populations of dicei and pineus in Tamaulipas and
brevirostris in Veracruz and Puebla, perhaps during glacial maxima. A thor-
ough search of the small isolated mountain ranges between Puebla and Tam-
aulipas may reveal relict populations of this species along the eastern escarp-
ment of the Mexican Plateau.

A possible route of gene exchange between populations of brevirostris
from southern to northwestern Mexico may have been via the Transverse
Volcanic Range across the southern edge of the plateau. Populations of
brevirostris with a lateral light line and some individuals of the populations with
an enclosed interparietal are found in central Puebla and western Veracruz.
Populations with individuals showing a decrease in the length of the lateral
light line and an increased presence of an enclosed interparietal are found in
eastern Mexico-Morelos. Populations with an enclosed interparietal in most
individuals and the lateral light line obscure occur in Jalisco and northwest-
ward into Durango and Chihuahua.

The Transverse Volcanic Range seems to have been isolated from the
Sierra Madre Oriental and occupied by a single population of brevirostris
indubitus for some period of time. A more uniform environment may have
allowed a re-invasion of the E. brevirostris indubitus population from the
Sierra Madre Oriental, following a glacial maximum. The suggested point of
contact between the two formerly separated populations of indubitus and
brevirostris probably occurred in the eastern part of the Transverse Volcanic
Range (Puebla and Mexico). A shifting and infrequent intergradation of
characters in this area suggests secondary contact and lends support to this
hypothesis.

A large hiatus exists in the distribution of Eumeces b. indubitus from
southern Jalisco to east central Mexico. The gap may be the result of insuffi-
cient collecting and probably does not represent an actual allopatric situation,
since the habitat of this form is continuous across southern Jalisco, Michoa-
can, Mexico, and northern Puebla.

An additional hiatus exists between populations of b. indubitus and b.
bilineatus from northwestern Jalisco to southern Durango. The valley of the
Rio Santiago is an effective barrier to genetic exchange between northwestern
Jalisco and southern Nayarit. Additional collecting in the small isolated moun-
tain ranges of northeastern Jalisco and northwestern Guanajuato may reveal
disjunct populations of E. b. bilineatus and indicate that a corridor may have
formerly connected E. brevirostris populations of the Transverse Volcanic

1969

Mexican Skinks of E. BREVIROSTRIS Group

23

Figure 5. Map of the distribution of Eumeces brevirostris indubitus, E. colimensis,
E. copei, and E. dugesii in west central Mexico.

Range and the Sierra Madre Occidental rather than through the mountains of
northwestern Jalisco and southern Nayarit.

Species Relationships

The significant scale and color pattern characters that separate the five
species of the brevirostris group are combinations of interparietal enclosed or
not enclosed; amount of contact between seventh supralabial and upper sec-
ondary temporal; number of supraoculars; number of longitudinal and trans-
verse scale rows; lamellae beneath fourth toe; snout- vent length; length of
limbs; presence and placement of a lateral light line; presence of a dorsolateral
light line on one, two, or three longitudinal scale rows.

The color pattern of two to four longitudinal light lines (lateral two lines
often incomplete) and squamation of the five species of the brevirostris group
give a clue to the primitive member of the group. The color pattern of four
complete, well-defined light lines in E. copei or an intermediate condition of
two complete and two incomplete light lines in E. dugesii could be the primitive
condition. Eumeces copei has a color pattern of four complete well-defined
light lines; all other species show some loss of the lateral light lines. The dor-
solateral light line is completely enclosed by the upper secondary and primary
dark lines in copei, less so in the other four species. The lateral light line is en-
closed by the primary dark line and the lower secondary dark line in copei and
southern populations of brevirostris. It may or may not be enclosed in dugesii
and is usually obscure in colimensis, ochoterenae, and central and northern
populations of brevirostris. The length of the dorsolateral light line is variable
in all species except copei and ochoterenae. Thus copei, showing a retention of

24

Contributions in Science

No. 168

Figure 6. Map of the distribution of Eumeces brevirostris bilineatus and E. sp. in Chi-
huahua, Durango, and Sinaloa, Mexico.

1969

Mexican Skinks of E. BREVIROSTRIS Group

25

all four lines, may have the primitive color pattern. However, dugesii is not
only intermediate in color pattern, it is also intermediate in squamation when
compared to the other four species.

The number of lamellae beneath the fourth toe is lowest in ochoterenae,
followed in order by dugesii , brevirostris, copei, and colimensis. The number of
transverse scale rows is lowest in ochoterenae, followed by colimensis, dugesii,
copei, and brevirostris. The number of longitudinal scale rows varies from 20
to 26 in brevirostris, overlapping the range of variation found in the longitudi-
nal scale rows of the other species. The number of superciliary scales, totaled
for both sides, is usually 10 to 12 in northern populations of brevirostris, 13 to
14 in southern populations of brevirostris and all populations of copei and
ochoterenae, 12 to 14 in dugesii, and 12 to 17 in colimensis. The number of
post suboculars totaled for both sides is six to eight in all species of the group.
The number of supraoculars is usually four on each side for all species except
dugesii, which has three. The primary temporal is usually present in all species,
except for northern populations of brevirostris. The least contact between the
frontonasal and the frontal is found in ochoterenae, followed by colimensis,
brevirostris, dugesii, and copei. Thus E. dugesii appears to be intermediate be-
tween other species of the group and may be the most primitive member, even
though it has only three supraoculars. Eumeces copei and E. ochoterenae
appear to be the most specialized members of the group in having scale char-
acters usually not found in the other three species and in having the opposite
extremes in color pattern.

Acknowledgments

For the loan of material utilized in this study I am grateful to Charles M.
Bogert, American Museum of Natural History (AMNH); Rollin H. Baker,
Michigan State University (MSU); Richard J. Baldauf, Texas Cooperative
Wildlife Collection, Texas A & M University (TCWC) ; William E. Duellman,
University of Kansas Museum of Natural History (KU) ; Alice G. C. Grandi-
son, British Museum of Natural History (BM [NH]); Robert F. Inger and
Hymen Marx, Field Museum of Natural History (FMNH); Richard B.
Loomis, California State College, Long Beach (CSLB); James A. Peters,
United States National Museum (USNM); Hobart M. Smith, University of
Illinois (UIMNH); Robert C. Stebbins, Museum of Vertebrate Zoology, Uni-
versity of California (MVZ); Donald W. Tinkle, University of Michigan
Museum of Zoology (UMMZ) .

I especially thank Charles M. Bogert for a copy of his field map for
localities in Oaxaca; Alice G. C. Grandison for her detailed description of the
holotype of E. brevirostris ; Jay M. Savage and Roy W. McDiarmid for their
critical reading of the manuscript.

SUMARIO

Las especies nominales Eumeces dicei y E. indubitus muestran ser sub-spe-

26

Contributions in Science

No. 168

Figure 7. Map of the distribution of Eumeces brevirostris dicei and E. b. pineus in
northeastern Mexico.

cies de E. brevirostris. Eumeces copei ha sido removida del grupo E. anthraci-
nus y colocada en el grupo E. brevirostris. Un analisis critico de las escamas y
colorido ha sido hecho para este grupo. Claves, fotografias y mapas de distrib-
ucion se presentan para cada forma. Tambien se discute una posible explicacion
de la distribucion de grupos de E. brevirostris.

1969

Mexican Skinks of E. BREVIROSTRIS Group

27

Figure 8. Dorsolateral view of the color patterns of four species of the Eumeces brevi-
rostris group. A. E. copei, Zempoala, Morelos; B. E. dugesii, 12 mi S Carapan, Mi-
choacan; C. E. ochoterenae, Agua del Obispo, Guerrero; D. E. colimensis (holotype),
Colima, Colima; E. Eumeces sp. {colimensis) , 1 mi E Santa Lucia, Sinaloa.

Variation in seven characters in the E. brevirostris group: A-E, E. brevirostris brevirostris from Oaxaca, Guerrero, Vera-
cruz, Puebla, Mexico-Morelos, respectively; F-H, E. brevirostris indubitus from Morelos, Mexico, Jalisco, respectively;
I-J, E. brevirostris bilineatus from Durango and Chihuahua, respectively; K, E. brevirostris pineus; L, E. brevirostris dicei;
M, E. dugesiv, N, E. copei\ O, E. ochoterenae\ P, E. colimensis.

28

Contributions in Science

No. 168

diduivs

d

»n

d

©

d

to

5— H

o

00

d

!

d

On

CO

ui uaqumpq

o

co

co

co

d

•o

'=t

Tt

d

d

CO

luasdid jDuodwaj

On

©

o

o

VO

OV

ov

o

^t

d

o

o

d

VO

dAVUlud IjllM %

On

on

O

o o

o\

ov

cv

ON

2

ON

vo

\VlUOA{ ‘IUOD

JVSVUOlUOjf IflJM. %

59

67

vo

O co
Tt wo

40

m

o\

d

ro

00

00

00

24

90

ON

00

92

O

CO

CO

jDAodwai C/z?

-puooas uaddn %iuoo

66

00

on

O

o

66

L6

8

o

o

86

o

o

o

o

o

o

o

o

o

o

o

55

CO

CO

■qvi -dns qii qjiM %

T"H

T— H

^ 1

,—l

--H

T— 1

pasopua pidund

VO

d

co

25

d

co

92

o

o

ON

o

o

o

o

d

o

VO

ON

uo

g

-J9JU1 qijM %

4

3 ° £

c\ on *•»

vo

o

©

o

©

o

o

o

o

o

00

o

d

o

©

•2 s ^

i

on

o

o

o

o

o

o

o

o

o

Ov

o

ON

O

o

Q .2 s

t-H

t-H

^ « o

d

d

t-H

!§ -5 5

co

^ S 3

co

T-H

o

d

^ 8 69

i

co

o

wo

o

00

d

.

4

,““l

•S «

O

o

o o

o

d

rc

m

lo

>o

d

v~)

d

CO

o

"s

£ £

i

WO

co

vo

ON

on

d

OV

■T

d

d

T— 1

t-H

«r>

©

^j-

00

d

wo

on

>o

d

d

d

d

»o

d

§ 3

co

d

d

co

T— 1

d

1

d

d

.2 ts

co

o

co

wo

t-H

vn

m

rc

vo

d

t-H

00

©

£ o
3 l3~

i

co

d

'T

m

vo

VO

On

ON

d

in

d

^ *d

•js: o

co

d

»5

i

d

£ 5

o

-9 2-2

xt-

d

CO

CO

00

00

t~H

Tt-

VO

d

'nI-

d

CO

K

00

CO

O

3 g

•5-3

00

1

d

00

ON

d

d

vo

On

T— H

d

d

-Ci o

d

co

O

d

o

rc

00

o

ov

Tt

d

ON

d

VO

g t'

d

wo

d

VO

d

vo

t-H

CO

d

co

co

r-H

o

d

d

VO

d

d

VT)

On

CO

.§ 'g

VO

d

d

d

<N

(N

d

d

d

t-H

d

3 Cl

vo

O

G\

O

d

o

m

m

vo

d

!

vo

CO

> S
-c *

i

vo

t-H

d

(N

d

VO

d

IT)

CO

d

CO

VO

d

d

d

o

ON

▼— H

CO

CO

h

d

H

4

m

d

4

<

B

C

D

E

F

G

X

z

o

fl-

Table 2

Variation in three characters of squamation of the E. brevirostris group.

1969

Mexican Skinks of E. BREVIROSTRIS Group

29

ov

Os

'st

00

>n

VO

n

r

n

r-

r

'St

r-'

ri

a

o

o

▼-H

1

T— 1

CT

> — i

’-H

1

1

n

t— i

o

i — i

1

t/3

o'

o

I

d

d

o

I

o

o

o

o

I

1

o

i

o

i

o

o

1

d

l

1

i

+

i

+

i

+

+

i

+

i

+

i

+

+

i

+

i

+

i

+

i

+

i

+

i

+

su

3

>n

co

oo

o

'St

'St

o

o

00

1

o

o

o

oo

00

1

3

U

CO

OS

so

T"

o

00

00

rj

n

q

1

Os

r>

t"

r~

|

o

o

d

q

o

o

q

q

d

q

o

o

d

o

o

V.

Su

~Ci

-2

z

<

w

q

q

as

ri

rn

ri

rj

ri

q

ri

,d-

cn

rr

q

Os

ri

00

ri

vq

«n

ri

-

O

vd

g

3

O

S

>

W

»n

'st

•st

in

'St

'St

in

vo

m

—a

in

'St

CO

'St

n

r-

o

1

’—t

— <

—

’ —

— t

—

— H

1

z

o

o

d

o

d

d

d

d

o

d

d

d

o

d

o

wo

Pi

,_l

— H

’ — 1

• —

• —

o

*st

VO

r-

r-

vo

>n

vo

o

r-

r-

o

r-

vo

Vi

H

' — i

i— 1

’—I

i

1

r—l

rr

n

’—i

n

rl

o

r-H

1

3

w

o

i

o

o

i

o

o

i

o

i

1

O

o

1

d

i

o

i

o

i

o

i

d

I

o

i

1

i

+

i

+

i

+

+

i

+

i

+

1

d

+ +

i

+

i

+

i

+

i

+

i

+

3

o

o

o

o

o

o

i

o

o

o

o

o

o

o

o

1

So

3

-2

£

g

Q

co

o

o

os

o

o

o

oo

d

1

-

oo

o

vo

d

00

o

Os

d

o

00

d

o

1

-2

Z

<3

os

•st

q

n

q

VO

o

OO

q

oo

q

vq

rl

m

•St

Os

o

"p

Vi

cd

CO

cd

rn

cn

'St

'Sf

cn

rn

ri

CO

CO

ri

ri

vd

V.

su

O}

2

n

rl

(N

n

n

n

n

ri

rl

rl

n

n

n

n

n

rl

-3

3

W

^t

'st

'st

'St

in

'St

'St

vo

vo

'-f

't

'St

'St

'St

'St

00

3

a

rl

ri

rl

(N

n

rl

ri

rl

n

n

rl

rl

n

rl

n

rl

z

o

(N

ri

r'l

ri

ri

•Ct-

ri

ri

ri

d

ri

ri

d

d

vd

<3

Pi

(N

rl

n

n

(N

n

rl

n

ri

rl

rl

n

ri

n

ri

rl

'St

f"

vo

o

r-

vo

os

vo

n

o

'St

oo

vo

W

rl

co

co

CO

'St

»n

in

CO

Os

m

co

VO

>n

1

ri

1

CO

o

o

o

d

o

o

d

d

o

d

o

o

o

d

o

1

>. 3

z

o

1

1

1

1

1

1

1

1

+

1

+

1

+

1

+

"3

+

+

+

+

+

+

+

+

1

+

1

+

1

+

1

V.

o

o

o

o

o

o

o

o

o

o

o

o

o

o

o

i

o

SU

3

§

•st

ri

co

ri

ri

ri

rj

ri

o

cn

q

ri

q

ri

ri

o

rl

ri

Os

00

ri

o

ri

vq

1

Cj

O

oa

5*.

Sl>

<L

oa

Z

<

m

in

d

co

00

n

o

co

o

cr

d

vq

ov

in

os

o

q

00

q

00

00

Os

o

Os

d

vq

os

q

CO

r-

d

>

tc

S

>n

>n

VO

vo

VO

in

in

VO

in

>n

>n

vo

»n

>n

m

in

>

3

r*

w

co

n

in

in

'St-

in

ri

00

in

os

in

in

CO

•n

r-

m

NJ

s*.

a

VO

VO

VO

VO

VO

vq

vq

vq

vq

»n

vq

vo

vo

vq

m

in

z

O

d

d

vo

VO

>n

»n

vo

d

vo

>n

d

CO

«n

os

d

pi

>n

in

•n

»n

•n

•n

•n

in

in

>n

«n

in,

•n

>n

'St

>n

9ldlUVS

rl

in

rl

o

r-

in

O

00

rl

r-

>n

Os

CO

o

ct

co

co

m

n

T—l

in

ct

'St

— i

n

ri

CO

ui JdqiuriN

.<■3

£

t»w>

<n

s«a

05

su

oa

3

3

3

O

S

SV>

*3

-2

ccj

O

c3

X

o

i-i

a

u

(O

N

o

M

O

03

a3

3

cj

<U

l-i

o

s

6

o

’x

3

"3

.3

-d

C/3

o

3

i-i

o

o

'x

O

o

C/3

3

SV;

.3

d

d

O

ox;

c

CO

»-!

cO

3

i-S

3

-C

b. pineus
b. dicei

Cvc

3

"3

3

2
S o

CL-3

o ^

sj O

3

Su

3

O

&5

cd

o

<U

<D

&a

o

<u

3—i

oa

=3

3

Co

oa

&a

&a

!>a

SU

3

o

o

>

Cl

SU

3

s

oj
l— s

3

0 u

<d>

3

Vj

O

su

Vj

U

SO

sv»

Sb

Sl)

3

3

3

?!

3

3

3

3

3

2

2

2

3

2

3

d

d

d

d d

d d d

d

30

Contributions in Science

No. 168

Literature Cited

Axtell, Ralph W. 1960. A new subspecies of Eumeces dicei from the Sierra Madre
of northeastern Mexico. Copeia, 1960:19-26.

Duellman, William E. 1961. The amphibians and reptiles of Michoacan, Mexico.
Univ. Kan., Mus. Nat. Hist., Publ., 15:1-148.

Gunther, Albert. 1860. On new reptiles and fishes from Mexico. Zool. Soc. Lon-
don, Proc., 1860:316-322.

Martin, Paul S. 1958. A biogeography of reptiles and amphibians in the Gomez
Farias region, Tamaulipas, Mexico. Univ. Mich., Mus. Zool., Misc. Publ., 101:
1-102.

Ruthven, Alexander G., and Helen T. Gaige. 1933. A new skink from Mexico.
Univ. Mich., Mus. Zool., Occ. Pap., 260: 1-3.

Smith, Hobart M., and Edward H. Taylor. 1950. An annotated checklist and key
to the reptiles of Mexico, exclusive of the snakes. U. S. Natl. Mus., Bull., 199:
1-253.

Tanner, Wilmer W. 1958. Two new skinks from Durango, Mexico. Great Basin
Nat., 18(2): 57-62.

Taylor, Edward H. 1933. Two new Mexican skinks of the genus Eumeces. Biol. Soc.
Wash., Proc., 46:129-138.

. 1935. A new skink from Mexico. Field Mus. Nat. Hist., Zool. Ser., 20:77-80.

. 1936. A taxonomic study of the cosmopolitan scincoid lizards of the genus

Eumeces, with an account of the distribution and relationships of its species.
Univ. Kan. Sci. Bull., 23:1-643.

. 1936. The rediscovery of the lizard Eumeces altamirani (Duges) with notes

on two other Mexican species of the genus. Biol. Soc. Wash., Proc., 49:55-58.

. 1943. Mexican lizards of the genus Eumeces, with comments on the recent

literature on the genus. Univ. Kan. Sci. Bull., 29:269-300.

Thominot, M. A. 1883. Note sur un reptile d’espece nouvelle provenant du Mexique
et appartenant au genre Eumeces (Plestiodon) . Soc. Philom., Paris, Bull., (7),
7( 1882-1883) : 138-139.

Webb, Robert G. 1959. Eumeces colimensis (Sauria, Scincidae), in Sinaloa, Mexico.
Southwest. Nat., 4:42.

Webb, Robert G., and Rollin H. Baker. 1962. Terrestrial vertebrates of the Pueblo
Nuevo area of southwestern Durango, Mexico. Amer. Midi. Nat., 68:325-333.

Accepted for publication May 5, 1968

LOS

I ANGELES
COUNTY
MUSEUM

UMBER 169

CONTRIBUTIONS
llltd IN SCIENCE

June 30, 1969

A NEW STREAM BREEDING FROG FROM OAXACA, MEXICO
(ANURA, HYLIDAE)

By Ian R. Straughan and John W. Wright

i

|

Los Angeles County Museum of Natural History • Exposition Park
Los Angeles, California 90007

CONTRIBUTIONS IN SCIENCE is a series of miscellaneous technical papers
in the fields of Biology, Geology and Anthropology, published at irregular intervals
by the Los Angeles County Museum of Natural History. 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 Robert J. Lavenberg, Los Angeles County Museum of Natural
History, 900 Exposition Boulevard, Los Angeles, California 90007.

Robert J. Lavenberg
Managing Editor

INSTRUCTIONS FOR AUTHORS

Manuscripts for the LOS ANGELES COUNTY MUSEUM, CONTRIBU-
TIONS 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. Al-
though priority will be given to manuscripts by staff members, or to papers dealing
largely with specimens in the Museum’s collections, 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 Edi-
torial Committee. They may also be subject to review by competent specialists out-
side 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.

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. Alterations or changes in the manuscript after galley proof will
be billed to the author. Unless specifically requested, page proof will not 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.

Dorothy M. Halmos
Editor

A NEW STREAM BREEDING FROG FROM OAXACA, MEXICO
(ANURA, HYLIDAE)

By Ian R. Straughan1 and John W, Wright2

Abstract: A new species of hylic! frog is described from a
series of adults, juveniles and larvae from the Sierra Madre del
Sur of Oaxaca, Mexico. This new species is assigned to the Hyla
bistincta species group and differs most distinctively from other
members of the group in the total absence of webbing between fin-
gers and in its unique color and color pattern.

The hylic! frog fauna of Mexico and Central America is characterized by
numerous species with relatively limited geographic ranges. Eight stream
breeding species of the Hyla bistincta species group have been described from
a relatively small area of southern Mexico between the transverse volcanic belt
and the Isthmus of Tehuantepec. In southwestern Oaxaca we collected a series
of frogs, referable to this group, from a high mountain stream in the Sierra
Madre del Sur, within 50 km of localities for two other species of the group
(H. bistincta and H. pentheter, Duellman, 1964, and Adler, 1965). As this
distinctive frog represents a previously unknown species, we propose

Hyla bogertae sp. nov.

Figs. 1-5

Holotype: LACM 44400; adult female (Fig. 1) collected from a tribu-
tary of the Rio Atoyac, below Vivero El Tapanal at 8,700 ft. (2652 m), one
mile (1.6 km) south of La Cofradia, district of Sola de Vega, Oaxaca, Mexico.
Collected by I. R. Straughan and J. W. Wright, September 28, 1968.

Allotype: LACM 44401; adult male (Fig. 5); collection data as for
holotype.

Paratypes: LACM 44402-14; two adult females (Fig. 2), nine juveniles
(Fig. 3), nine larvae (Fig. 4), and one recently metamorphosed juvenile
(reared from stage 30 larva in laboratory). All are topoparatypes.

Diagnosis: Hyla bogertae is allocated to the Hyla bistincta species group
which is distinguished from other frogs from Mexico by the following com-
bination of characters: absence of quadrato-jugal element, vocal slits, and
webbing between fingers; presence of both finger and toe discs, a non-protrud-
ing prepollex; toes fully webbed except fourth, which is webbed to distal
subarticular tubercle.

1 Research Associate, Section of Herpetology, Los Angeles County Museum of Natu-
ral History, and Department of Biological Sciences, University of Southern Califor-
nia, Los Angeles, California 90007.

2Curator of Herpetology, Los Angeles County Museum of Natural History.

1

2

Contributions in Science

No. 169

Figure 1. Hyla bogertae sp. nov. Holotype (LACM 44400). Left, dorsal view; right,
ventral view.

As a member of the bistincta group, it is distinguished from the remaining
members as follows: from H. bistincta Cope by complete absence of webbing
between fingers, relatively smaller tympanum, weaker tarsal fold, weaker
fringe of web along penultimate phalanx of fourth toe, and absence of
elongated cloacal flap; from H. charadricola Duellman by less extensive web-
bing on fourth toe, complete absence of webbing between fingers, and absence
of axillary membrane; from H. chryses Adler by more extensive webbing on
fourth toe, absence of axillary membrane, presence of distinctly darker flanks
with yellow spots; from H. robustifemora Taylor [?=H. crassa (Brocchi)] by
complete absence of webbing between fingers, less extensive webbing on fourth
toe, absence of strong tarsal fold, and presence of a single bony element in the
prepollex; from H. pachyderma Taylor by obvious tympanum, less extensive
webbing on fourth toe, absence of strong tarsal fold, absence of spines on
nuptial pads of breeding male; from H. pent he ter Adler by absence of strong
tarsal fold, absence of elongated cloacal flap, and more extensive webbing on
fourth toe; from H. robertsorum Taylor by complete absence of webbing
between fingers, absence of spines on nuptial pads of breeding male, venter
immaculate white without dark and light spots; and from H. siopela Duellman
by absence of rostral keel, complete absence of webbing between fingers,
presence of distinctly darker flanks with yellow spots. The diagnoses within the
H. bistincta group are summarized in Table 1.

Description of Holotype: Adult female (Fig. 1), snout-vent length 50.1
mm. Habitus stout, limbs strong, hind limbs relatively long — tibio-tarsal
articulation reaches to eye. Snout rounded, truncate in profile. Canthus

1969

Oaxacan Stream Breeding Frog

3

Figure 2. Hyla bogertae sp. nov. Two adult female paratypes. Left, LACM 44403, at
extreme of darkening; right, LACM 44402 at opposite extreme of color change.

rounded, loreal area concave, nostril not protruding, closer to tip of snout
than to eye. Tympanum small, rounded, upper margin hidden in supratym-
panic fold, one-third to one-half diameter of eye, separated from eye by a
distance of 1.5 times own diameter. Tongue rounded, barely notched poste-
riorly, free for one-fifth of length behind and with free fringe along anterio-
lateral margins. Vomerine teeth on small rounded ridges, anterior margins
between choanae and posterior margin behind level of posterior edge of
choanae. Forearm thick. Wrist fold weak. Fingers entirely without webbing,
in order of length 1<2<4<3. Discs moderate, twice width of narrowest
part of finger. Disc of third finger completely covering tympanum. Subarticular
tubercules minute, flat and rounded. When tibiae adpressed to femora, heels
barely overlap. Tarsal fold not continuous, marked by a line of barely raised
white crenatures. Inner metatarasal tubercle small and oval, outer indistinct.
Palmar tubercles minute, rounded, in a single row along metacarpals. Super-
numerary tubercles minute, rounded and in single rows along proximal part of
digits. Toes almost fully webbed, in order of length 1<2<3 = 5<4. Webbing
to discs of all toes except fourth, which has penultimate phalanx free with a
weak fringe to disc. A narrow flap of skin on inner edge of first toe folded
towards ventral surface. Cloacal opening above level of middle of thigh.
Cloacal flap transverse. Skin smooth to finely shagreen on all dorsal surfaces,
granular on venter.

Dorsal surface of body and limbs olive green (gun metal gray in alcohol)
with extensive silver to pale bronze reticulation, largely maintained in alcohol.
Light bar above level of cloaca continuous with dorsal reticulation on left

ic features of members of the bistincta

4

Contributions in Science

No. 169

Oh

ft

O

S— i

W)

o

<u

Oh

CO

Vh

ft

£

£

ft

m

CD O5

O S
rs O

& ft.

S a

<o

X

■ ^

^ c

X. a

a sc
2 -ft
ft

CD C3

ft

g-* *

h1

ft -ft
ft ^

CD ft

3 §

k|

CD

CO

ft

Mft.

•ft (U
> £4)

^ Vh CO

£ £ o
^*8 «

W) ~

A 2

> ft
-ft O
O co

A

V

<N

co

»- ft

2^ a

ft O C3
ft ft

co ft D
^3 C d
■ft 44 o

'a S ft

c -22

<L> 2 £
Q.-C _
0.0
<4-1 -*-*

O « JJ

<u 03 &P
co C C

ft ft •£
42 %-■ Cft<

CD

Co

ft

-ft

cn

a £

O I co
ft

.-ft o 5

^ <W

Z £ o

> ft co
-2 2-2
ft "ft ft
S— i co

T3

ft

£ 03 x

s-i G ft

tJ o «
ccS -ft .c
ft ft «5

A

V

m

\

(N

^ g

ft ft «

gj

O W W) o
ft ft ^
co C
ft O
42 ’ft

CJQ

ft
O 'ft

S-h

ft

^ «+H

" ° -5 S3

2 £ O g

S c £ ft

S « $ 2

CO

ft O

« Oh
ft on

W) ft

A ^
£ 2
£4)-0

-2 £
"ft • «
Oh £

Oh

ur>

ftt-

m

£ 5

s’i:

Vh

ft

c

ft

£

"£

chryses all base of ante- weak ca. 1/2 present transverse not distinct from

rudimentary penultimate dorsum

phalanx

pachyderma all middle of strong concealed absent transverse dark with pale

rudimentary penultimate yellow reticulations

phalanx

1969

Oaxacan Stream Breeding Frog

5

x

§ c

M .2

2

^ 3
^ .2 ^

X QJ
~ > X

- I E

tS 2

« i.S

T3

QJ

£ 3 42

O W)~
J-i c o3
>-i n <d

03 J2 43
C <L> on

<N

a>

§ 3

V(-l 03

° E *

u x c
T3 3 2

73 c &
'p <U X
C Q- CL

3 6

o3

E

c/5

4: 2
~ o

> cu

> CO

C £
£ o

O X
i-< <u
X

<U

(N

■ x
3 §

si

<4-1

o

<u o3
co C
03 C

x x

&

tj-

3 <u

C G

<u o

E c

X <N
CJ £ —

i|

X <D
£

<U ~

X g
o on co

3 £ O
3 O Oh
73 <+-l CO

A

V

m

C/J

c

o

5 =

^ § g

2 <N

3 -r2

ri d *—

5

o ^

5 «

^ %

•x q

0)

a

£

2~
’S

E

03

<L>

o ”g
E §

•X II

e

Uh c<3
O <U
X X

co co

<N

3

§ 3

<4H C?

° S X
jo X C
-a 3 3

73 c o3
■G <L> X

c a- o-

J-H

03

X

"So

a

<u

3 ^ rJ

ri —

G "^r

<N C

6

Contributions in Science

No. 169

Figure 3. Three juvenile paratypes of Hyla bogertae sp. nov.

side, broken with dot on right. Posterior surface of thigh darker olive, with two
cream (white in alcohol) spots below cloaca at beginning of ventral granularity.
Ventral surface of body white with yellow wash along flanks. Limbs mainly
darker with yellow in groin area, around heels, upper arm, and elbow. Throat
darker olive with large yellow spots and minute creamy pustules.

Description of Allotype: Adult male (Fig. 5), snout-vent length 45.1 mm.
Non-spinose, black nuptial pad on prepollex, with a separate small oval pad
on penultimate phalanx of first finger. Bony element of prepollex single, curved
and conical. Wrist fold entirely lacking.

Color darker than in holotype with essentially the same pattern, but
slightly less development; throat color darker and more extensive than in
holotype. In all other characters allotype agrees with holotype.

Description of Paratypes: Two adult females, slightly smaller than the
holotype, agree closely with the holotype except for color details (see Fig. 2).
Coloration in the type series including juveniles varies in the degree of reticula-
tion and changes in life with the degree of hydration and temperature (Figs.
2 and 3). The variation ranges from almost completely silver with scattered
small dark spots to almost completely dark olive with indistinct reticulation,
and outstanding dark flanks with bright yellow spots. Spots below cloaca may
be elongated or broken into smaller units. Spots on throat may be smaller than
in holotype, imparting a speckled effect. Measurements of type series are
tabulated below (Table 2).

Description of Tadpoles: At stage 30 (Gosner, 1960) the tadpoles have a
typical mountain stream habitus (Fig. 4, middle) — streamlined bodies, long
muscular tails with low tail fins and a large suctorial mouth. Anal tube right of
tail fin. Spiracle high on left side, just visible from below. Eyes slightly closer
to margin of head than to midline. Nostrils within line projected from eyes

1969

Oaxacan Stream Breeding Frog

7

Figure 4. Hyla bogertae sp. nov. Top, larva at stage 25, lateral view; middle, larva
at stage 30, lateral view; bottom, mouth parts of larva at stage 30. The bar below
each illustration is 1 cm.

8

Contributions in Science

No. 169

to tip of snout. Tail nearly twice as long as body. Tail muscles heavy but taper-
ing acutely. Mouth (Fig. 4, bottom) surrounded by two rows of oral papillae,
with one additional row dorsal and two rows ventral. Clusters of papillae in
angle of mouth. One complete tooth row and one divided dorsally; three com-
plete ventrally, middle longest. Head and body heavily pigmented dorsally,
tapering to an anteriorly directed dark wedge from eyes to between nostrils.
Paler laterally with scattered melanophores and golden chromatophores. Tail
muscles pale, myotomes distinct, marked off by lines of melanophores. Tail
fins translucent with large clusters of melanophores. Total body and tail length
from 52 to 57 mm in four specimens preserved at stage 30. At stage 25 (Fig. 4,
top) tadpoles are from 30 to 32 mm (3 specimens) in total length, closely
resembling stage 30 tadpoles, except that the mouth parts are represented by
a single row of papillae enclosing mouth. The number of tooth rows are not
reduced but are not fully developed in the number of teeth. Snout-vent length
at metamorphosis (one specimen in laboratory) 20.2 mm.

Habitat: All individuals were collected from a system of small and medium
sized streams flowing down steep slopes through pine-fir forest. Adults and
juveniles were encountered sitting on rocks or piles of detritus deposited by
flood. The main stream consisted of small pools (two to three m wide) with
sandy bottoms partially covered with leaf litter, and small water falls and
rapids (one-half to one and one-half m wide). When disturbed, the frogs
jumped into the water and remained submerged for a short period before
re-emerging. Water temperature at time of capture of frogs was 14° C. Tad-
poles were found in the larger pools resting on the bottom in quieter water.

Distribution: Known only from the type locality.

Relationships: Hyla bogertae is the tenth species to be allocated to the
bistincta group. Duellman reviewed the five known members of the group in
1964. He examined the single known specimen of both H. crassa (female from
“Mexico”) and H. robustifemora (male) and concluded that they were con-
specific, although he noted some clear differences. We have not examined H.
crassa, but in view of the difficulties of specifically associating females with
males within this group and the fact that H. crassa and H. robustifemora differ
in significant ways, we conclude that H. crassa should be regarded as a nomen
dubium until such time as it can be clearly associated with an extant population.

Duellman (1968) commented on the relationships within the bistincta
group. He divided the group into three series, and suggested that the four
species in his crassa subgroup (hereinafter referred to as the robustifemora
subgroup) — crassa {^^robustifemora) , pachyderma, robertsorum and siopela
— might be subspecies of a single species. As defined, H. bogertae is allocated
to this subgroup, and represents its most distinctive member. Four of the mem-
bers of this subgroup ( bogertae , robustifemora, pachyderma and siopela ) have
been described from at best a small series of specimens from single widely
separated localities, and one ( robertsorum ) is known only from a series of

1969

Oaxacan Stream Breeding Frog

9

five proximal localities relatively close to the locality of siopela. Hyla siopela
is so superficially similar to robertsorum that they could be considered to
belong to the same species population; however, in addition to the rostral
keel (difficult to ascertain in preserved specimens), radiograms reveal that the
bony prepollex element in siopela is more than twice as large as in robertsorum.
More extensive collecting in intermediate areas will be necessary before an
objective analysis of status of the members of this subgroup can be made.

Duellman grouped the remaining four species into two groups, bistincta
and pentheter, characterized by having long cloacal sheaths, and charadricola
and chryses, characterized by having axillary membranes and lacking nuptial
excrescences. We present below a key which emphasizes the diagnostic
characters discussed.

Key to the Members of the Hyla bistincta Species Group

la. Toes fully webbed (to disc of 4th toe) 2

lb. Toes not fully webbed (penultimate phalanx of 4th toe

or more free of web) 3

2a. Wrist fold present 9

2b. Wrist fold absent charadricola

3a. Tympanum completely concealed pachyderma

3b. Tympanum distinct 4

4a. Axillary membrane obvious chryses

4b. No axillary membrane 5

5a. Prominent, narrow, elongate cloacal flap or sheath 6

5b. Cloacal flap broad, transverse, not protruding posteriorly 7

6a. Black dorso-lateral band from tip of snout to

groin (may be broken along flanks) pentheter

6b. No dark dorso-lateral band bistincta

7a. Prominent rostral keel present siopela

7b. No rostral keel 8

8a. Venter brown with creamy white flecks robertsorum

8b. Venter immaculate white bogertae

9a. Tympanum distinct robustifemora

9b. Tympanum hidden crassa

Remarks: Although H. bogertae is here allocated to the Hyla bistincta
species group as defined by Duellman ( 1964) and broadened by Adler ( 1965)
and Duellman (1968) because of the absence of a quadrato-jugal element and
the presence of a non-projecting prepollex, there remains some question as to
its proper generic allocation. Seemingly the only difference between the
bistincta group and members of the genus Plectrohyla is in the degree of
development of the ossified prepollex element (or elements) . The large number

10

Contributions in Science

No. 169

Figure 5. A. Hyla bogertae sp. nov. Allotype (LACM 44401), radiograph of left
hand with drawing of prepollex based on composite X-rays.

B. Hyla robustifemora holotype (UIMNH 25050), radiograph of left hand with
drawing of prepollex based on composite X-rays.

of similarities in habitus of adults and tadpoles, breeding habits and habitat,
argues for the inclusion of the entire H. bistincta species group with the genus
Plectrohyla. The question as to whether this assemblage (as well as other
similar groups of stream breeding frogs in Mexico and Central America) is a
phylogenetic unit or the result of parallel development is currently under
investigation.

Specimens Examined

Hyla bistincta Cope. MEXICO: Guerrero: between “Puerto Chico” and
“Asoleadero,” ca. 63-65 km (rd) SW “Casa Verde” (20 mi WNW of Chilpan-
cingo), 2440-2500m (UMMZ 125376). Jalisco: Sierra de Autlan, ±15 mi
SE of Autlan, 7000' (UMMZ 102076). Michoacan: Uruapan: Los Conejos,
4 mi WSW of Uruapan (UMMZ 94238-40); Cerro San Andres, W slope,

1969

Oaxacan Stream j

Breeding

Frog

11

Table 2

Measurements of type series of Hyla bogertae in mm

LACM NO.

Sex

Snout-vent Tibia
length length

Head

length

Head

width

Eye Tympanum
diam. diam.

44400a

9

50.1

27.0

14.4

18.0

5.4

1.8

44401"

$

45.1

22.8

11.8

16.5

4.7

1.9

44402

9

49.3

25.8

14.6

17.3

5.0

2.0

44403

9

43.3

24.0

13.0

16.6

5.0

1.9

44404

juv.

32.5

15.1

8.7

10.2

3.6

1.2

44405

juv.

31.7

15.5

9.1

10.6

3.2

1.1

44406

juv.

29.9

15.1

8.4

9.7

3.4

1.3

44407

juv.

29.0

15.1

8.5

9.7

3.3

1.2

44408

juv.

28.0

13.2

8.3

9.6

3.2

1.1

44409

juv.

28.7

13.2

8.2

9.5

3.1

1.0

44410

juv.

25.7

12.1

7.3

8.8

2.6

indistinct

44411

juv.

25.7

12.1

7.1

9.0

2.8

indistinct

44412

juv.

26.5

11.6

7.1

9.0

2.7

0.8

44413°

juv.

20.2

9.8

5.5

7.2

2.6

indistinct

aHoIotype.

^Allotype.

cMetamorphosed in laboratory ( 1-1-1969) from tadpole collected at stage 30.

±11 mi WNW of Ciudad Hidalgo, 7800' (UMMZ 102075) ; Parque Nacional,
5500' (UMMZ 115233, 121515). Veracruz: Cumbres near Acultzingo on
Puebla side in Veracruz (FMNH 105482-83).

Hyla bogertae sp. nov. MEXICO : Oaxaca: tributary of the Rio Atoyac,
below Vivero El Tapanal at 8,700 ft (2652m), one mile (1.6 km) south of
La Cofradia, district of Sola de Vega (LACM 44400-14).

Hyla charadricola Duel! man. MEXICO: Puebla: 3.6 mi W Acaxochitlan
on Pachuca-Tuxpan Road (UMMZ 104032); 9 mi W of Huachinango,
±7400' (UMMZ 118166); 7.3 mi SW of Huachinango (UMMZ 121567).
Hidalgo: Lago Tejocotal, 7 mi E of Acaxochitlan, 7300' (UMMZ 118165).

Hyla chryses Adler. MEXICO: Guerrero: between “Puerto Chico” and
“Asoleadero” ca. 63-65 km (rd) “Casa Verde” (20 mi WNW of Chilpan-
cingo), 2440-2500m (UMMZ 125373, 125375).

Hyla pachy derma Taylor. MEXICO: Veracruz: Pan de Olla (USNM
115026-29).

Hyla pentheter Adler. MEXICO: Oaxaca: ca. 23 mi (rd) N of San
Gabriel Mixtepec, 1700m (UMMZ 125377-80).

Hyla robertsorum Taylor. MEXICO: Hidalgo: El Chico National Park
(FMNH 100124, 75786), (UMMZ 92462); El Chico Parque Nacional,
2700m (UMMZ 106401, 106443).

Hyla robustifemora Taylor. MEXICO: Oaxaca: Cerro San Felipe
(UIMNH 25050).

12

Contributions in Science

No. 169

Hyla siopela Duellman. MEXICO: Veracruz: from a small stream on
the west slope of Cofre de Perote, elevation 2500-2550m (UIMNH 57693,
57697, 58207-08).

Acknowledgments

We wish to express our sincere appreciation to Martha M. Bogert for her
gracious hospitality during our stay in Oaxaca, for her enthusiastic field com-
panionship and for the time and effort she provided in maintaining the frogs
alive while we conducted field work with her husband in Chiapas. It is with
pleasure that we name this new frog in her honor not only for her relation-
ship with us, but for her numerous, essentially anonymous, contributions to
herpetology.

Specimens used for comparative purposes were kindly made available to
us by the following individuals and institutions: Hymen Marx, Field Museum
of Natural History (FMNH); James A. Peters, United States National
Museum (USNM); Dorothy M. Smith, University of Illinois Museum of
Natural History (UIMNH); and Charles F. Walker, University of Michigan,
Museum of Zoology (UMMZ).

We also extend our appreciation to Mr. William Presch and Mr. John
Meyer for kindly preparing radiograms of material examined, to Mrs. Evie
Templeton for illustrating the tadpoles, and to Mr. James Honey for photo-
graphic assistance.

SUMARIO

Una nueva especie de rana, Hyla bogertae es aqui descrita basada en series
de adultos, juveniles y estados larvarios colectados en la Sierra Madre al Sur de
Oaxaca, Mexico. Esta nueva especie ha sido asignada al grupo especifico de
Hyla bistincta; la diferencia mas marcada, de otros miembros de este grupo,
es la ausencia total de membranas entre los dedos y la distribution de su
coloration que es unica.

Literature Cited

Adler, K. 1965. Three new frogs of the genus Hyla from the Sierra Madre del Sur of
Mexico. Univ. Michigan, Mus. Zook, Occ. Pap., 642: 1-18, pi. 1.

Duellman, W. E. 1964. A review of the frogs of the Hyla bistincta group. Univ.
Kansas, Mus. Nat. Hist., Publ., 15:469-491.

. 1968. Descriptions of new Hylid frogs from Mexico and Central America.

Univ. Kansas, Mus. Nat. Hist., Publ., 17:559-578, pis. 17-19.

Gosner, K. L. 1960. A simplified table for studying anuran embryos and larvae with
notes on identification. Herpetologica, 16:183-190.

Accepted for publication February 17, 1969.

LOS

ANGELES
COUNTY
MUSEUM
—
f UMBER 170

CONTRIBUTIONS
flMe IN SCIENCE

June 30, 1969

TAXONOMIC NOTES ON THE
MYRMECOCYSTUS MELLIGER COMPLEX
(Hymenoptera: Formicidae)

By Roy R. Snelling

Los Angeles County Museum of Natural History • Exposition Park
Los Angeles, California 90007

CONTRIBUTIONS IN SCIENCE is a series of miscellaneous technical papers
in the fields of Biology, Geology and Anthropology, published at irregular intervals
by the Los Angeles County Museum of Natural History. 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 Robert J. Lavenberg, Los Angeles County Museum of Natural
History, 900 Exposition Boulevard, Los Angeles, California 90007.

Robert J. Lavenberg
Managing Editor

INSTRUCTIONS FOR AUTHORS

Manuscripts for the LOS ANGELES COUNTY MUSEUM, CONTRIBU-
TIONS 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. Al-
though priority will be given to manuscripts by staff members, or to papers dealing
largely with specimens in the Museum’s collections, 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 Edi-
torial Committee. They may also be subject to review by competent specialists out-
side 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.

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. Alterations or changes in the manuscript after galley proof will
be billed to the author. Unless specifically requested, page proof will not 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.

Dorothy M. Halmos
Editor

TAXONOMIC NOTES ON THE
MYRMECOCYSTUS MELLIGER COMPLEX
(Hymenoptera: Formicidae)

By Roy R. Snelling1

Abstract: The Myrmecocystus melliger complex is defined
and the taxonomic status of the specific and varietal names in-
cluded is investigated. A key to the component species, for work-
ers and females, is given and these castes are illustrated. Species
included are: M. mendax, M. melliger (synonym: M. melliger
mendax comatus) , M. semirufus and M. placodops (synonym:

M. melliger orbiceps ); M. melliger testacea is removed from the
synonymy of M. semirufus and transferred to the mexicanus group
as a senior synonym of M. mojave (new synonym ) .

When Wheeler (1908) revised the genus Myrmecocystus, he recognized
but two protean species, M. melliger Forel and M . mexicanus Wesmael. Each
of these had attributed to it a number of subspecies and varieties. Additions by
Wheeler and others in subsequent years added three specific names and a welter
of additional subspecies and varieties, most of these assigned to M. melliger.

Probably the number of names proposed would not be so high had not
Wheeler propounded a peculiar theory with respect to these ants. He postu-
lated that certain forms were behaviorally incapable of forming repletes. The
result was that whenever he found a colony of a presumably nonreplete form-
ing species which contained repletes, he was faced with a dilemma; whether to
abandon his original idea or to describe the occupants of the colony as a new
form. Unfortunately, he chose the latter course of action, resulting in a number
of superfluous names founded on extremely dubious characters.

Creighton (1950) reevaluated many of these names and rightly placed
them in synonymy. The present paper is a further evaluation of one complex
in the genus, the M. melliger complex, and is published at this time in order that
the changes proposed here may be available to others conducting investigations
in this genus.

As constituted by Creighton (1950), the M. melliger complex includes
those species with the following combination of characteristics in the worker
caste: mandibles with seven teeth; eyes small, their greatest diameter equal to
the length of the first funicular segment; ocelli always present, prominent;
petiolar node thick from front to back, the crest blunt; erect hairs coarse and
numerous, abundant and conspicuous on cheeks, those on gaster arising from
punctures at the top of small conical papillae; pubescence dense, especially on
gaster, largely obscuring surface beneath; integument dull to moderately shin-
ing; larger species, 4.5-1 1.0 mm.

’Entomology Section, Los Angeles County Museum of Natural History.

1

2

Contributions in Science

No. 170

The component names assigned to this complex by Creighton were
arranged as follows:

M. comatus Wheeler, 1908
M. melliger melliger Forel, 1886
M. melliger calif ornicus Cole, 1936
M. melliger orbiceps Wheeler, 1908
M. menclax Wheeler, 1908

The subspecies M. m. californicus was known only from the type mate-
rial, which Creighton had not seen, and was only tentatively placed with
M. melliger. As I have shown elsewhere (Snelling, The identity of Myrme-
cocystus melliger mimicus var. californicus Cole, unpublished), this species
must be transferred to the genus Formica. Creighton predicted that M. m.
orbiceps would prove to be a synonym of the nominate form and in 1954
Creighton and Crandall were able to verify this. The only recent change was by
Gregg (1963), who proposed returning M. mendax to the status originally
given it by Wheeler, a subspecies of M. melliger.

My own investigation of this genus, based on type material of all forms,
has convinced me that drastic changes are necessary within this complex and I
propose to arrange the species as follows:

M. mendax Wheeler, 1908
M. melliger Forel, 1886
M. semirufus Emery, 1893
M. placodops Forel, 1908

Each of these component species is discussed in detail below.

Myrmecocystus mendax Wheeler

Myrmecocystus melliger var. semirufus Emery, 1893. Zoologische Jahr-
buch, Abt. fur Systematik, 7:667 (in part).

Myrmecocystus melliger subsp. mendax Wheeler, 1908. Bull. Amer. Mus.
Nat. Hist., 24:351, fig. 4, 9 $ $; Wheeler, 1912, Psyche, 19:173; Gregg,
1963, Univ. Colo. Press, pp. 645-648.

Myrmecocystus mendax Creighton, 1950. Bull. Mus. Comp. Zool., 104:
442, 445.

This species, originally described from Colorado, ranges westward as far
as California, through New Mexico and Arizona. It is closely related to M.
semirufus and even more closely to M. melliger. I believe, however, all are
distinct species, even though the differences which separate them are slight.
When Emery described M. melliger var. semirufus, he had before him a mixed
series of specimens from San Jacinto, California, and from Denver and Pueblo,
Colorado; each series represents a different species. The Denver specimens are
M. mendax, the Pueblo specimens are M. melliger and the series from San
Jacinto (which includes the type) are M. semirufus. From the related species,
M. mendax may be separated by the characters given in the key below.

1969 Notes on the MYRMECOCYSTUS MELLIGER Complex

3

Figure 1. Myrmecocystus spp., major workers; left, frontal view of head; right, lat-
eral view of head and alitrunk; M. melliger, cotype from Mexico; M . mendax , cotype
from Mt. Washington, Colo.; M. placodops , type from Mexico; M. semirufus , cotype
from San Jacinto, Calif.

4

Contributions in Science

No. 170

Myrmecocystus melliger Forel

Myrmecocystus melliger Forel, 1886. Annales Societe Entomologique de
Belgique, 30 : 202. ? .

Myrmecocystus melliger var. semirufus Emery, 1893. Zoologische Jahr-
buch, Abt. fur Systematik, 7:667 (in part).

Myrmecocystus melliger subsp. mendax var. comatus Wheeler, 1908.
Bull. Amer. Mus. Nat. Hist., 24:352, fig. 5, ? $ $. Wheeler, 1912, Psyche,
19:173. NEW SYNONYM.

Myrmecocystus comatus, Creighton, 1950. Bull. Mus. Comp. Zool.,
104:442; Gregg, 1963, Univ. Colo. Press, p. 643.

The types of M . melliger and M. comatus have been available to me and
have been compared with one another, and found to be the same. It is unfor-
tunate that M. comatus is the same as M. melliger since this necessitates apply-
ing a completely revised concept to an old name. How Wheeler made the error
he did is not difficult to understand, even though Forel had sent him specimens
from the original series of M. melliger. As pointed out above, Wheeler was
convinced that some species do not form repletes. During his extensive field
work in this genus, he never found repletes in the nests of the ant which he
described as M. melliger comatus. Obviously, in his concept, this could not be
the same as Forel’s M. melliger, since that was known to produce repletes. On
the other hand, another very similar ant commonly had repletes present in the
nest, and Wheeler was bound to equate this with M. melliger. In so doing, he
ignored the specimens of true M. melliger and relied instead on behavioral data
interpreted in accordance with an incorrect postulate. Further, his own field
data were inadequate, as I took a colony of M. comatus ( i.e ., M. melliger) in
the Jeff Davis Mountains, the type locality of M. comatus, which contained
several fully developed repletes.

Myrmecocystus semirufus Emery

Myrmecocystus melliger var. semirufus Emery, 1893. Zoologische Jahr-
buch, Abt. fur Systematik, 7:667 (in part).

Although Wheeler had a cotype of this species available to him, he mis-
identified the ant and applied this name to a different taxon. The correct
identity of M. semirufus has never been recognized until now. This species is
known only from semi-desert mountain areas in southern California. I have
seen specimens from Riverside, San Bernardino, Los Angeles and Inyo
Counties.

Although very closely allied to M . mendax, it differs consistently from that
species in the characters given below in the key. The two species occupy very
similar habitats in the Joshua Tree-Juniper Woodland association, but appear
to be allopatric. In California, M. mendax is known only from scattered desert
mountain ranges in the eastern Mojave Desert, M. semirufus from the San

1969 Notes on the MYRMECOCYSTUS MELLIGER Complex

5

PLACODOPS

SEMIRUFUS

Figure 2. Myrmecocystus spp., gynes, lateral view; M. melliger, from Davis Mts.,
Tex.; M. mendax, cotype from Mt. Washington, Colo.; M. placodops, cotype of M.
orbiceps from Bull Cr., Tex.; M. semirufus, from Llano, Calif.

6

Contributions in Science

No. 170

Jacinto, San Bernardino, San Gabriel and Sierra Nevada ranges flanking
desert areas.

The species previously considered to be M. semirufus by Wheeler and
subsequent authors must now be known as M. kennedyi Cole, 1936. In 1950
Creighton listed M. melliger var. testacea Emery as a synonym of M. semi-
rufus. I have examined cotypes of that form and find that it has nothing in
common with the true M. semirufus, nor is it related to M. kennedyi. This
species must be transferred to the mexicanus group as a valid species and a
senior synonym of M. mojave Wheeler, 1908, over which it has priority
(NEW SYNONYM).

Myrmecocystus placodops Forel

Myrmecocystus melliger var. placodops Forel, 1908. Bulletin Societe
Vaudois des Sciences Naturelles, ser. 5, 44:70. 9 .

Myrmecocystus melliger, Wheeler, 1908. Bull. Amer. Mus. Nat. Hist.,
24:348, fig. 2, 9 9; Wheeler, 1912. Psyche, 19:174-175; Creighton, 1950.
Bull. Mus. Comp. Zool., 104:442, 444-445. (Misidentification)

Myrmecocystus melliger subsp. orbiceps Wheeler, 1908. Bull. Amer. Mus.
Nat. Hist., 24:349, fig. 3. 9 9 ; Wheeler, 1912. Psyche, 19:173; Creighton,
1950. Bull. Mus. Comp. Zool., 104: 442, 445. NEW SYNONYM.

Forel’s name was published in March, 1908, and Wheeler’s appeared on
May 9 of the same year. Accordingly, it is necessary to use the name M. pla-
codops for this species. I have examined the unique type, now in the Forel
Collection at the Museum d’Histoire Naturelle, and compared it with cotypes
of M. melliger orbiceps. There is no doubt that these are conspecific. The type
of M. placodops is, unhappily, from an unknown locality in Mexico. It is un-
fortunate that the type locality is thus Mexico, since so little of the range of
this species lies in that country. In the United States, M. placodops is found
from central Texas to southern California. The Mexican distribution of this
species appears to be limited to areas immediately adjacent to the International
Border.

That portion of the key to Myrmecocystus species by Creighton (1950)
involving the M. melliger complex is inadequate to separate the various forms
involved. I have, therefore, prepared the following key to separate the worker
and female castes of the four species.

Key to Members of MYRMECOCYSTUS MELLIGER Complex

Workers

1. Worker maxima with head length equal to or exceeding maximum head
width, the outer margins of head, in full face view, subparallel or slightly
convex; frontal area with dense, distinct punctures; occiput, behind ocelli,
with dense, fine punctures; erect mesonotal hairs longer than those of

1969 Notes on the MYRMECOCYSTUS MELLIGER Complex

7

MELLIGER

MENDAX

SEMI RUFUS

Figure 3. Myrmecocystus spp.. gynes; left, frontal view of head; right, dorsal view of
mesoscutum.

8

Contributions in Science

No. 170

femora, erect hairs elsewhere exceptionally long, slender, flexuous (except
in M. semirufus , which has especially pronounced cephalic punctures).... 2

Worker maxima with maximum head width exceeding head length, the
outer margins of head, in full face view, moderately to strongly convex;
frontal area dull, roughened, with scattered weak punctures; occiput, behind
ocelli, dull, roughened, with inconspicuous fine punctures; erect hairs every-
where stiff, blunt, those of mesonotum shorter than those of femora

placodops Forel

2. Erect hairs of body very uneven in length, exceptionally long, slender,

pointed, those of mesonotum longer than those of femora; frontal punctures
somewhat obscured by roughened integument 3

Erect hairs of body uniform in length, short, stiff, blunt, those of mesono-
tum distinctly shorter than those of femora; frontal integument somewhat
shining, the punctures very distinctly defined semirufus Emery

3. Head of all sizes longer than broad; longest pronotal hairs longer than

greatest eye length melliger Forel

Head of largest workers as broad as long; longest pronotal hairs no more
than three-fourths as long as greatest eye length mendax Wheeler

Females

1. Mesoscutal and mesoscutellar discs finely punctate, the punctures dense

and evenly spaced, separated by about 1. 5-2.0 times a puncture diameter
near notaulices, with or without an impunctate median area, with fine,
appressed or subappressed white pubescence in addition to long, coarse,
erect yellowish hairs 2

Mesoscutal disc with sparse, coarse, setigerous punctures, the integument
shining, without appressed pale pubescence; mesoscutellum finely, closely
punctate, with scattered coarse punctures semirufus Emery

2. Mesoscutal disc with a median impunctate area of variable extent, though

median line may be punctate 3

Mesoscutal disc uniformly punctate throughout mendax Wheeler

3. Impunctate area of mesoscutal disc with a longitudinal median zone of fine,
close punctures; frontal lobes dull, with coarse, variably spaced punctures
which become finer and denser toward frontal line; vertex, between eyes

1969 Notes on the MY RMECOCY STU S MELLIGER Complex

9

and ocelli, roughened, with scattered, obscure punctures of variable size
placodops Forel

Mesoscutal disc with entire median area impunctate; frontal lobes somewhat
shining, with fine, dense, evenly spaced punctures; vertex between eyes and
ocelli with fine, distinct, close punctures melliger Forel

Type material of the species discussed above has been made available
through the courtesy of the following: C. Besuchet and C. Ferriere, Museum
d’Historie Naturelle, Geneva; W. L. Brown, Jr., Museum of Comparative
Zoology; J. G. Rozen, Jr., American Museum of Natural History; D. R. Smith,
United States National Museum. To each of these gentlemen, I offer my most
profound thanks for their cooperation. The figures were prepared by Ruth Ann
DeNicola under assistance rendered by Grant No. 4494, the Penrose Fund, of
the American Philosophical Society.

Literature Cited

Creighton, W. S. 1950. The ants of North America. Harvard, Mus. Comp. Zook,
Bull., 104:1-585.

Creighton, W. S., and R. H. Crandall. 1954. New data on the habits of Myrmeco-
cystus melliger Forel. C. C. N. Y., The Biol. Rev., 16:2-6.

Gregg, R. E. 1963. The ants of Colorado. Univ. Colo. Press, 792 pp.

Wheeler, W. M. 1908. Honey ants, with a revision of the American Myrmecocysti.
Amer. Mus. Nat. Hist., Bull., 24:345-397.

Accepted for publication February 24, 1969

LOS

ANGELES

COUNTY

MUSEUM

CONTRIBUTIONS
fS? IN SCIENCE

I UMBER 171

June 30, 1969

THE PHILIPPINE SUBGENUS HOPLOPROSOPIS OF HYLAEUS
( Hymenoptera : Collet] dae )

By Roy R. Smelling

Los Angeles County Museum of Natural History • Exposition Park
Los Angeles, California 90007

CONTRIBUTIONS IN SCIENCE is a series of miscellaneous technical papers
in the fields of Biology, Geology and Anthropology, published at irregular intervals
by the Los Angeles County Museum of Natural History. 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 Robert J. Lavenberg, Los Angeles County Museum of Natural
History, 900 Exposition Boulevard, Los Angeles, California 90007.

Robert J. Lavenberg
Managing Editor

INSTRUCTIONS FOR AUTHORS

Manuscripts for the LOS ANGELES COUNTY MUSEUM, CONTRIBU-
TIONS 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. Al-
though priority will be given to manuscripts by staff members, or to papers dealing
largely with specimens in the Museum’s collections, 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 Edi-
torial Committee. They may also be subject to review by competent specialists out-
side 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.

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. Alterations or changes in the manuscript after galley proof will
be billed to the author. Unless specifically requested, page proof will not 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.

Dorothy M. Halmos
Editor

THE PHILIPPINE SUBGENUS HOPLOPROSOPIS OF HYLAEUS
(Hymenoptera: Colletidae)

By Roy R. Snelling1

Abstract: The subgenus Hoploprosopis of the Philippine Is-
lands is redescribed and compared with the superficially similar
subgenus Metylaeus of the Ethiopian Region. The diagnostic char-
acteristics of Hoploprosopis are the antennal scape shorter than
interantennal distance; preoccipital carina present; scutellum and
metanotum conspicuously spinose; cephalic and mesosomal punc-
tation very coarse.

Hoploprosopis was described by Hedicke (1926) as a subgenus of Pro-
sopis, to accommodate a new species described from Lamao, Luzon, Philippine
Islands. Hedicke stressed the unusual presence of spines on the scutellum and
metanotum and the very coarse punctation as important diagnostic features.
These are the same characters emphasized by Bridwell (1919) for his genus
Metylaeus of the Ethiopian Region. Since Hedicke made no mention of Brid-
well’s genus, it may be assumed he was unaware of its existence. Popov ( 1939) ,
apparently working from descriptions alone, suggested that Hoploprosopis may
prove to be a synonym of Metylaeus.

At the present time I, together with G. I. Stage, am studying the hylaeine
bees of the Ethiopian Region, and it is pertinent to reconsider the status of
Hoploprosopis. Two males in the collections of the Los Angeles County Mu-
seum of Natural History (LACM) belong to the subgeneritype of Hoplopro-
sopis. Paratypes of Metylaeus cribratus Bridwell, the type species of Metylaeus,
were studied at the United States National Museum.

Subgenus Hoploprosopis Hedicke

Type species: Prosopis ( Hoploprosopis ) quadricornis Hedicke, 1926.
Monobasic and original designation.

Diagnosis'. Antennal scape shorter than interantennal distance; preoccip-
ital carina present; scutellum and metanotum with conspicuous caudally di-
rected, dorso-ventrally compressed, apically broadened, spines; cephalic and
mesosomal punctation very coarse.

Male\ Mandibles slender, bidentate apically, outer tooth longest; man-
dibular bases almost contiguous with lower eye margins; clypeus longer than
broad; clypeo-ocular distance slightly greater than clypeal width at base;
clypeo-antennal distance less than clypeal width at base, slightly more than
maximum diameter of antennal sockets; antenn-ocular distance equal to about
three-fifths the maximum diameter of antennal socket; interantennal distance
about two times an antennal socket diameter; supraclypeal area elevated be-
tween antennal sockets, with carinate margin, the carinae curving toward, and

Entomology Section, Los Angeles County Museum of Natural History.

1

2

Contributions in Science

No. 171

Figure 1. Hylaeus ( Hoploprosopis ) quadricornis, male. Frontal view of head. Fig-
ure by Ruth Ann DeNicola.

joining, antennal sockets on their lower margin; scape very short, its maximum
length (exclusive of basal condyle) less than interantennal distance, distinctly
shorter than following three segments combined; first flagellar segment distinct-
ly broader than long; flagellum stout, median segments about as broad as long;
genal areas narrow, about half as wide as eyes in lateral view; vertex com-
pressed, interocellar distance much greater than distance between ocelli and
posterior margin of vertex; preoccipital carina present. Carina on pronotal
collar complete, crossing pronotal lobes anteriorly; a subcarinate ridge from
pronotal humeri to neck; anterior, impunctate face of mesepisternum separated
from punctate lateral face by a distinct carina, especially well-defined below;
scutellum (Fig. 2) with defined horizontal and posterior faces, meeting at about
a right angle, with a flattened, caudally directed, slightly curved, apically broad-
ened spine on each side; axillae laterally marginate, slightly, arcuately pro-
duced; metanotum well below level of horizontal face of scutellum, with a
spine, similar to that of scutellum, on either side of mid-line; propodeum

1969

Hoploprosopsis of Hylaeus

3

Figure 2. Thoracic dorsa of H. ( Hoploprosopis ) quadricornis (above) and H.
( Metylaeus ) crib rat us (below). Figures by Ruth Ann DeNicola.

4

Contributions in Science

No. 171

Figure 3. H. ( Hoploprosopis ) quadricornis. Left, eighth ventrite; middle, ninth ven-
trite, ventral view; right, ninth ventrite, lateral view. Figures by Ruth Ann DeNicola.

Figure 4. H. ( Hoploprosopis ) quadricornis. Genital capsule, from left to right, ven-
tral, dorsal and lateral views. Figures by Ruth Ann DeNicola.

1969

Hoploprosopsis of Hylaeus

5

coarsely areolate above, lateral, oblique and transverse carinae present but ob-
scured by areolae; propodeal disc reticulate; lateral fovea of second tergite
punctiform; seventh sternite narrowly rounded apically, disc flat, without de-
pressed margin. Marginal cell of forewing about twice as long as distance from
its apex to apical wing margin, apical, free portion about twice as long as basal
portion; second submarginal cell higher than long, about one-third as long as
first. Terminalia: Figs. 3-5. Head and thorax very coarsely punctured; meta-
somal punctures much finer.

Female : Unknown.

Hoploprosopis has an undeniable superficial resemblance to the African
subgenus Metylaeus, but the very different terminalia will immediately distin-
guish between them. Useful external features are the scape length (exceeding
interantennal distance and length of the following three segments combined in
Metylaeus ), the caudally directed, flangelike occipital carina of Metylaeus
(simple in Hoploprosopis) , the different thoracic spination (spines laterally
placed on both segments and laterally compressed in Metylaeus , while they are
dorsoventrally compressed in Hoploprosopis) , the differing metasomal puncta-
tion (nearly as coarse as the thoracic punctation in Metylaeus) , and the dif-
ference in the shape of the second submarginal cell (quadrate in Metylaeus) .
By analogy with other groups of Hylaeus , it is predicted that most of the above
enumerated features will apply to females as well.

I have compared Hoploprosopis with examples of most of the subgenera
of the Australian area, as these are defined by Michener (1965). There is
nothing in any of its characters to indicate an affinity with this fauna. The
hylaeines of southeastern Asia are so poorly known that comparison with them
seems impossible at the present time.

Specimens Examined’. 2 $ $ , Biliran Island, Philippine Islands, no further
data (L ACM).

Literature Cited

Bridwell, J. C. 1919. Miscellaneous notes on Hymenoptera. Haw. Ent. Soc., Proc.,
4:109-165.

Hedicke, H. 1926. Beitrage zur Apidenfauna der Philippinen (Hym.) (2. Beitrag zur
Kenntnis orientalischer Apiden.) Deutsche Entomologische Zeitschrift, Jahrg.
1926:413-423.

Michener, C. D. 1965. A classification of the bees of the Australian and South Pa-
cific Regions. Amer. Mus. Nat. Hist., Bull., 130:1-362.

Popov, V. 1939. Subgeneric groupings of genus Prosopis F. (Hymenoptera). Acade-
mic des Sciences de l’URSS, Comptes Rendus (Doklady), 25: 167-170.

Accepted for publication April 21, 1969.

LOS

ANGELES

COUNTY

MUSEUM

CONTRIBUTIONS
IN SCIENCE

umber 172 June 30, 1969

)

AVIAN FOSSILS FROM THREE PLEISTOCENE
SITES IN CENTRAL MEXICO

By Hildegarde Howard

Los Angeles County Museum of Natural History • Exposition Park
Los Angeles, California 90007

CONTRIBUTIONS IN SCIENCE is a series of miscellaneous technical papers
in the fields of Biology, Geology and Anthropology, published at irregular intervals
by the Los Angeles County Museum of Natural History. 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 Robert J. Lavenberg, Los Angeles County Museum of Natural
History, 900 Exposition Boulevard, Los Angeles, California 90007.

Robert J. Lavenberg
Managing Editor

INSTRUCTIONS FOR AUTHORS

Manuscripts for the LOS ANGELES COUNTY MUSEUM, CONTRIBU-
TIONS 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. Al-
though priority will be given to manuscripts by staff members, or to papers dealing
largely with specimens in the Museum’s collections, 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 Edi-
torial Committee. They may also be subject to review by competent specialists out-
side 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.

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. Alterations or changes in the manuscript after galley proof will
be billed to the author. Unless specifically requested, page proof will not 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.

Dorothy M. Halmos
Editor

AVIAN FOSSILS FROM THREE PLEISTOCENE
SITES IN CENTRAL MEXICO

By Hildegarde Howard1

Abstract: A small collection of avian bones from Chapala,
Tequixquiac and San Marcos represents more than twenty spe-
cies, 17 of which had no previous fossil record in Mexico. Of
nine extinct forms, five relate to the Pleistocene of Rancho La
Brea, California or San Josecito Cave, Nuevo Leon; one (from
Chapala) is referred to a Late Pliocene species; and three are
undescribed.

Since Loye Miller’s report (1943) on the large avifauna from San Jose-
cito Cave, in Nuevo Leon, there have been but few records of birds from the
Pleistocene of Mexico. A Pied-billed Grebe was reported by Wetmore ( 1949)
from Tepexpan, State of Mexico. An Ocellated Thrasher was tentatively iden-
tified by Robert Storer (1954) from Tequixquiac, State of Mexico. A sum-
mary by Downs (1958) of vertebrates from the Chapala Formation in Jalisco
included “probable Mexican Cormorant and flamingo.” Brodkorb (1963)
noted the Band-tailed Pigeon and Mourning Dove from Barranca Seca, in
Veracruz.

The present report concerns additional avian remains from Tequixquiac
and Chapala, as well as specimens from San Marcos, about 25 miles west of
Chapala in the state of Jalisco (see Fig. 1). The material from the three locali-
ties, totalling 44 unassociated bones, is in the collections of the Los Angeles
County Museum of Natural History (LACM). A preliminary report on this
material, plus a collection from Lago de Atotonilco, was presented at the
annual meeting of the Geological Society of America in Mexico City (Howard,
1968). The larger collection from Atotonilco is still under study.

With the exception of coracoids of the Pliocene Pliolymbus baryosteus
loaned by the University of Michigan Museum of Paleontology (UMMP),
and three modern skeletons of Phalacrocorax olivaceus from the University
of California Museum of Vertebrate Zoology, Berkeley, (MVZ), all compara-
tive material is from the LACM collections.

Chapala

Reports on the geology and vertebrate paleontology of the Lake Chapala
area, in eastern Jalisco, have been published by Palmer (1926), Downs (1958)
and Clements (1963). Briefly, the combined results of their studies indicate
that the fossils in the tilted sediments of the Chapala Formation, which occur
along the north shore of the present Lake Chapala, were deposited earlier,

1Research Associate in Avian Paleontology, Los Angeles County Museum of Natural
History

1

2

Contributions in Science

No. 172

and in a larger lake bed, than those recovered from the bottom of the existing
lake. An Early Pleistocene or Plio-Pleistocene age is suggested for the
Chapala Formation (Clements, 1963: 47).

Collections of vertebrate fossils in the Lake Chapala area were made by
the California Institute of Technology and the Los Angeles County Museum
of Natural History in cooperation with the Seminario para Estudios Cenozoi-
cos sponsored by the Instituto de Geologia, Universidad Nacional de Mexico.
No avian fossils were recovered from the bottom sediments of the present
lake. But 14 bird bones were obtained from the Chapala Formation at LACM
localities 1129 (type section, 1.3 miles north of the town of Chapala) and 1176
(near the west edge of the type beds). Other vertebrates from the Chapala
Formation include fish, crocodilian, turtle, mastodont (Cuvieronius) , peccary,
and horse {Equus and Nannipus) (Downs, 1958: 77).

Ten of the bird bones are specifically assigned to three species, as follows:

Pliolymbus baryosteus Murray, Extinct Grebe, 1 bone

Phalacrocorax olivaceus (Humboldt), Mexican Cormorant, 7 bones

Phoenicopterus ruber Linnaeus ?, American Flamingo, 2 bones

The remaining four fragments are too poorly preserved for even generic
identification. They represent a large goose, a duck, an egret or small heron,
and a coot or rail.

The grebe bone is an upper end of a right coracoid (LACM 2891) which
falls between Podiceps dominicus and P. caspicus in the dimension of distance
from scapular facet to head. However, in distance across the triosseal canal,
below the furcular facet, it is relatively broader than in either of the living
species. In both size and proportions it agrees with coracoids of Pliolymbus
baryosteus Murray, from the Late Pliocene of Kansas and Idaho (UMMP
27173 and 49577). A further, qualitative character, not noted in Murray’s
(1967) description of Pliolymbus, is also observed, i.e, all of the fossil speci-
mens, including the Chapala coracoid, have a more blunt brachial tuberosity
than does P. dominicus or P. caspicus. The Mexican coracoid, is therefore,
referred to Pliolymbus baryosteus. This occurrence considerably extends the
geographic range of that species, and tends to corroborate the age determina-
tion of the Chapala beds as Plio-Pleistocene.

The seven cormorant bones represent different elements (coracoid,
humerus, ulna, carpometacarpus, femur, tibiotarsus and tarsometatarsus), but
they were not associated and probably represent at least four individuals. The
coracoid (LACM 2778) was compared with the type of Phalacrocorax gole-
tensis Howard (LACM 4632) from the Pliocene of Michoacan. In each of the
characters that distinguish the Pliocene form from the living Mexican Cor-
morant, P. olivaceus (see Howard, 1965), the Chapala coracoid differs from
P. goletensis and resembles the living species. In fact, except for a slightly

1969

Fossil Birds from Central Mexico

3

Figure 1. Map of central Mexico, showing location of fossil sites, marked +

4

Contributions in Science

No. 172

longer tarsometatarsus (1 mm), and straighter, more medially placed anterior
intermuscular line on the femur, all of the elements can be matched in size
and qualitative characters in the four skeletons of P. olivaceus available for
comparison. As the position of the intermuscular line varies within the
Recent series, this character does not seem to be significant, and the Chapala
bones are referred to P. olivaceus.

The flamingo is represented by a slightly crushed coracoid (LACM 2770)
from locality 1176, and a right pedal phalanx (phalanx 1, digit 2, LACM
20867) from locality 1129. The coracoid is equal in size to that of large indi-
viduals of the existing American Flamingo. In qualitative characters, also, the
specimen agrees with coracoids from Recent skeletons of this species except
for possibly less clearly defined upper border of the furcular facet towards the
posterior side. However, there is considerable variability in shape and exten-
sion of the facet among Recent skeletons, and the significance of the slight
deviation found in the fossil bone is questionable.

Comparison was also made with the description of the coracoid of Phoe -
nicopterus copei Shufeldt, from the Late Pleistocene of Fossil Lake, Oregon
(Howard, 1946: 158). The two bones are very nearly the same length, but in
P. copei the furcular facet is much broader in anteroposterior dimension, and
the head is less thick. The coracoid is not represented in the other North
American fossil flamingos. However, from other elements represented, the
previously described Pliocene Phoenicopterus stocki Miller (1944) from the
Pliocene of Chihuahua, was obviously a much smaller bird than the one repre-
sented at Chapala, and P. minutus Howard, from the Pleistocene of California,
was even smaller than P. stocki. P. floridanus Brodkorb, from the Pliocene of
Florida, was probably the equal of the Chapala bird in size. The characters of
the leg bones, on which the description of the species is based (Brodkorb,
1953), seemingly present more evident distinctions from the Recent species
than are observable in the Mexican bone, but the relationship is indetermin-
able. The assignment of the Chapala coracoid to Phoenicopterus ruber is made
on a tentative basis in view of the possible significance of the character of the
furcular facet. No flamingos are found in central Mexico today.

Tequixquiac

In the 1930’s, the California Institute of Technology (CIT) collected
fossil vertebrates from the fluviatile deposits in the barrancas and limestone
fissures near Tequixquiac, in the northern part of the state of Mexico (Fig. 1).
The specimens are now included in the collections of the Los Angeles County
Museum of Natural History. The 15 bird bones were obtained from the fol-
lowing five localities, which bear LACM (CIT) locality numbers: 268, fissure
deposit in travertine quarry near El Tajo; 308, Barranca del Muerto; 309, Bar-
ranca del Rio Grande; 310, 61 kilometers del Gran Canal; and 311, La Can-
tera Vieja.

1969

Fossil Birds from Central Mexico

5

Hibbard (1955) recorded the mammalian fauna of the Tequixquiac area,
based largely on the collection in the Instituto de Geologia, Universidad Na-
tional Autonoma de Mexico, and reviewed previous studies of the deposits. He
states that the beds are referred to the Upper Becerra Formation, and may
represent two stages of Late Pleistocene deposition, the fissure deposits pos-
sibly being slightly older than the overlying horizontal beds. The mammalian
fauna that he records includes such typical Pleistocene forms as glyptodont,
Tremarctotherium, Mammuthus, Equus, Camelops, Bison, and Eucerather-
ium. He also notes remains of frogs or toads, turtles, and two gophers. A single
bird bone from one of Hibbard’s localities was tentatively referred to the
thrasher, Toxostoma ocellatum (Storer, 1954: 144).

From the fissure deposits, Furlong (1925) listed Capromeryx, Platygonus,
and Canis (Aenocyon) .

Eleven of the avian bones from the five LACM (CIT) localities are as-
signed as follows (extinct forms marked f ) :

f Ciconia ? sp., Extinct stork, 1
Anas acuta Linnaeus, Pintail Duck, 1
Anas cyanoptera Viellot ?, Cinnamon Teal, 2
Ay thy a americana (Eyton) ?, Redhead, 1
Ay thy a collaris (Donovan) ?, Ring-necked Duck, 1
t Breagyps clarki (Miller), La Brea Condor, 1
Aquila chrysaetos (Linnaeus), Golden Eagle, 2
t Spizaetus grinnelli (Miller) ?, Grinnell Eagle, 1
t Caracara prelutosa grinnelli (Howard), Extinct Caracara, 1

The other four fragments represent a swan, a hawk, and two passerines.
The stork bone is a poorly preserved distal end of a tibiotarsus (LACM
4645) from the fissure deposit (LACM (CIT) Loc. 268). The few diagnostic
features discernible indicate relationship to the extinct Ciconia maltha Miller
or the existing Maguari Stork, Euxenura maguari, rather than to either of the
members of the family recorded from Mexico in Recent time, i.e ., the Wood
Ibis and Jabiru. It is also dissimilar to the Pliocene Dissourodes milleri as de-
scribed by Short (1966). The size is less than the one available specimen of
the existing Euxenura maguari or a minimum specimen of Ciconia maltha
from Rancho La Brea. The bone is too fragmentary for accurate determina-
tion. In view of the probability of closer relationship to the widely distributed
Pleistocene stork of the United States than to the living South American
Maguari, the specimen is tentatively referred to Ciconia.

The most significant specimen in the collection is an axis vertebra (LACM
4638) of a condor. The specimen matches in all details an axis in the Rancho
La Brea collection assigned to the extinct Breagyps clarki (Miller) (LACM
G9346). The Breagyps axis is distinguished from that of Gymnogyps (both

6

Contributions in Science

No. 172

Recent and Pleistocene forms) by the position of the postzygapophyses. The
Mexican fossil resembles Breagyps in having the articular facets parallel to the
anteroposterior line of the vertebra. In Gymnogyps they are tipped in toward
the centrum anteriorly, and are also much more deeply cup-shaped. Gymno-
gyps, but not Breagyps, was present in San Josecito Cave.

A fragment of the distal end of a left tarsometatarsus (LACM 4646),
lacking the internal trochlea and all but the proximal tip of the middle troch-
lea, agrees in general size with some specimens of Aquila chrysaetos and the
extinct Spizaetus grinnelli. It resembles the latter as distinguished from Aquila
in the relatively shorter and more rounded external trochlea, deeper extensor
groove above the distal foramen, and more proximally extended middle troch-
lea. In view of the fragmentary condition of the specimen, the assignment to
Spizaetus grinnelli is tentative.

The identification of the caracara is based on an incomplete humerus
(LACM 4641), lacking the proximal end above the distal contour of the
bicipital crest. Distally the bone has a sharp ridge distal to the ectepicondylar
prominence, characteristic of the type of Caracara prelutosa (Howard, 1938:
226). The assignment to the Mexican race, C. prelutosa grinnelli, described
from San Josecito Cave (Howard, 1940), is based on the short, broad propor-
tions of the bone. A sight comparison with five humeri from San Josecito Cave
shows the Tequixquiac specimen to be shorter than the minimum cave bone.
The range in the cave series is, however, only 3 mm, and the minimum is only
1.5 per cent shorter than the mean. In a series of 67 humeri of the larger race,
C. prelutosa prelutosa from Rancho La Brea, the range is 15 mm and the
minimum is 9 per cent shorter than the mean. A similar range for the Mexican
race should include even shorter humeri than the one from Tequixquiac.

Inasmuch as swans are of rare occurrence in Mexico today, the presence
of the swan bone in the fossil collection is significant. The specimen is a frag-
ment of the distal end of a carpometacarpus that could belong either to Cyg-
nus or to Olor. Both genera are recorded in the Pleistocene of the United
States.

The Tequixquiac avifauna reflects the Late Pleistocene age of the de-
posit. All four of the identified raptorial species occur at Rancho La Brea and,
except for Breagyps clarki, were also found at San Josecito Cave.

San Marcos

Mr. Harold E. Smith, of the Sociedad de Ciencias Naturales de Lago de
Chapala, is responsible for the discovery of fossils in the lacustrine deposits
near San Marcos (Fig. 1). The occurrence was brought to the attention of the
Los Angeles County Museum of Natural History and collecting was carried on
by the museum in 1964-1965. Vertebrate fossils were obtained at two sites:
LACM locality 1952, on the east side of El Tecolote Island, in Laguna de
San Marcos; and LACM locality 1953, on the west side of the island, in

1969

Fossil Birds from Central Mexico

7

Laguna de Zacoalco. Dr. Reid Macdonald, who directed the field work, found
that deposits on both sides of the small island are part of the same lake bed,
and the fossils occurred in undisturbed material in the bottom of the lake,
which was then dry.

At this writing, the large mammalian fauna from San Marcos has not
yet been recorded, but, according to Dr. Macdonald, it is typical of the Late
Pleistocene. Fifteen bird bones represent nine species, two of which can be
assigned only generically. The taxa identified are as follows (extinct forms
marked t ) :

Pelecanus eryihrorhynchos Gmelin, White Pelican, 4
fPhalacrocorax sp., Cormorant, 1
t Mycteria wetmorei Howard ?, Wood Ibis, 1
Ardea herodias Linnaeus, Great Blue Heron, 1
fPhoenicopterus sp., Flamingo, 2
Chen hyperborea (Pallas), Snow Goose, 1
Anas acuta Linnaeus, Pintail Duck, 1
Aythya affinis (Eyton), Lesser Scaup Duck, 3
•fButeogallus fragilis (Miller) , Extinct Eagle, 1

The cormorant bone is a proximal fragment of a humerus (LACM 18501 )
lacking both deltoid and bicipital crests. The short, broad pectoral attachment
on the externa! tuberosity relates the specimen to Phalacrocorax auritus and
P. olivaceus in contrast to P. penicillatus and P, pelagicus, in which the attach-
ment is long and narrow. As far as size can be determined on the fragment, it
appears to fall between P. auritus and P. olivaceus, possibly closer to the latter.
Presumably an extinct species is represented.

A proximal end of a Wood Ibis carpometacarpus (LACM 9741) is com-
parable in size to Rancho La Brea specimen LACM K3529, tentatively as-
signed to Mycteria wetmorei (Howard, 1935: 253). Both fossils are equalled
by Recent specimens of M. americana in depth from the tip of the process of
metacarpal 1 to the internal crest, but exceed the existing species in other
measurements of the proximal end. This similarity in proportions suggests
specific identity of the La Brea and Mexican forms. However, the San Marcos
bone has a shorter proximal symphysis than have any of the other specimens
(Recent or fossil) compared. The assignment to M. wetmorei is, therefore,
tentative.

A proximal and a distal fragment of a left humerus (LACM 9738 and
LACM 18505, respectively), possibly from the same individual, represent a
large flamingo. Neither fragment provides qualitative details, but both- are
obviously larger than humeri of Recent flamingos. Measurements of breadth
and depth of the head and depth of the external distal condyle are 19 to 24
per cent greater than in the largest of eight available skeletons of the existing

8

Contributions in Science

No. 172

Phoenicopterus ruber (LACM 235). The largest of the North American fos-
sil flamingos, P. copei Shufeldt, is not represented by a humerus. The meas-
urements of other elements (see Howard, 1946: 158) compared with the
above-noted large skeleton of P. ruber (LACM 235), show a maximum of
only ten per cent greater size than in the Recent species. Of three P. copei
tarsometatarsi, two are slightly narrower than the P. ruber specimen, although
the third is seven per cent broader. The tibiotarsus is ten per cent broader proxi-
mally, the coracoid four per cent broader. These figures suggest that the San
Marcos flamingo was larger than P. copei, and is probably an undescribed
species. The humeral fragments are inadequate for a proper description.

A distal end of a left humerus (LACM 9740) represents a small eagle,
referred to Buteogallus fragilis (Miller). In qualitative characters it resembles
humeri of this species from the three California asphalt deposits. The bone is
slightly narrower than specimens from Rancho La Brea, but is closely matched
by humeri from McKittrick and Carpinteria.

The San Marcos avian representation agrees with the concept of a Late
Pleistocene age for these deposits. Two species, Mycteria wetmorei and Bute-
ogallus fragilis, are typical of Rancho La Brea.

Summary

Twenty-eight avian species are represented in the Los Angeles County
Museum of Natural History collections of fossil vertebrates from three central
Mexican Pleistocene localities— Chapala, Tequixquiac, and San Marcos. Eight
of the species, however, cannot be generically assigned. Eleven still occur in
Mexico today, although the most westerly record of the flamingo is from the
state of Veracruz (Lowery and Dalquist, 1951).

Combined Avifaunas of Chapala, Tequixquiac and San Marcos
t Pliolymbus baryosteus Murray
*Pelecanus erythrorhynchos Gmelin
*Phalacrocorax olivaceus (Humboldt)
t Phalacrocorax sp.

Indeterminate egret or small heron
*Ardea herodias Linnaeus
* Phoenicopterus ruber Linnaeus ?
t Phoenicopterus sp.
iMycteria wetmorei Howard ?
t Ciconia ? sp.

Indeterminate swan
Indeterminate large goose
*Chen hyperborea (Pallas)

*Anas acuta Linnaeus

1969

Fossil Birds from Central Mexico

9

*Anas cyanoptera Viellot ?

* Ay thy a americana (Eyton) ?

* Ay thy a collaris (Donovan) ?

* Ay thy a affinis (Eyton)

Indeterminate duck

t Breagyps clarki (Miller)

-\-*Aquila chrysaetos (Linnaeus)

iA-Spizaetus grinnelli (Miller) ?

t Buteogallus fragilis (Miller)

Indeterminate hawk

t -ACaracara prelutosa grinnelli (Howard)

Indeterminate coot or rail
Indeterminate small passerines (2)
t— Extinct

+— Also recorded from San Josecito Cave
*— Occurs in Mexico today

Nine extinct species are represented, five of which are assigned, at least
tentatively, to forms described from the Late Pleistocene of the United States;
three of these are recorded also from San Josecito Cave, Nuevo Leon, Mexico.
The grebe found at Chapala relates to the Late Pliocene of the United States.
Its occurrence is in keeping with the assumption that the Chapala Formation
was deposited earlier in the Pleistocene than the deposits at the other localities.
Three specimens, assigned only to genus, appear to represent undescribed spe-
cies, but are too fragmentary for description.

At least 17 of the birds from these localities have no previous fossil record
elsewhere in Mexico. San Josecito Cave, from which nearly all Mexican
Pleistocene avian records have come, represents a largely terrestrial environ-
ment, with cave-nesting and predatory birds predominating, and bones of
water birds comprising less than one per cent of the collection. The central
Mexican localities have added birds of the lakes and streams. At Tequixquiac
about 50 per cent of the bones found represent aquatic species, and at Chapala
and San Marcos, over 90 per cent are so assigned. Including all forms identi-
fied at least to genus, the resulting avifauna for the Pleistocene of Mexico now
totals 58 species.

Acknowledgments

For the fossil and Recent skeletons loaned for this study, I am grateful to
the University of Michigan, Museum of Paleontology, and the University of
California, Museum of Vertebrate Zoology, Berkeley, and to Dr. Claude W.
Hibbard, Bertram G. Murray, Jr., and Ned Johnson, who arranged the loans.
I want to express my appreciation also to Drs. Theodore Downs and Reid Mac-
donald of the Los Angeles County Museum of Natural History, for helpful
discussions concerning the field work in Chapala and San Marcos, and to them

10

Contributions in Science

No. 172

and to Dr. Herbert Friedmann, Director of the Museum, for continuing to
place the collections of the Museum at my disposal and provide most con-
genial surrounding in which to work. A special thank you goes to my husband,
Henry Anson Wylde, who prepared the map of the central Mexican localities.

SUMARIO

Una pequena coleccion de huesos de Aves procedentes de Chapala,
Tequixquiac y San Marcos, representan mas de veinte especies, de los cuales
17 no han sido encontradas en el record de fosiles en Mexico. De nueve
formas extintas cinco estan relacionadas con el Pleistoceno del Rancho La
Brea en California o Cueva de San Josecito en Nuevo Leon; una especie (de
Chapala) se refiere al Plioceno superior; y tres mas que no han sido descritas.

Literature Cited

Brodkorb, Pierce. 1953. A Pliocene flamingo from Florida. Chicago Acad. Sci.,
Nat. Hist. Misc., 124: 1-4.

. 1963. Two doves in the Pleistocene of Veracruz, Mexico. Condor, 65: 334.

Clements, Thomas. 1963. Pleistocene history of Lake Chapala, Jalisco, Mexico.
In Essays in Marine Geology in honor of K. O. Emery. Univ. So. Calif. Press,
Los Angeles, pp. 35-49.

Downs, T. 1958. Fossil vertebrates from Lago de Chapala, Jalisco, Mexico. In
Internat. Geol. Congress, 20th, Mexico City, 1956, Sect. VII: 75-77.

Furlong, E. L. 1925. Notes on the occurrence of mammalian remains in the
Pleistocene of Mexico, with a description of a new species, Capromeryx
mexicana. Univ. California, Publ. Geol. Sci., 15: 137-152.

Hibbard, Claude W. 1955. Pleistocene vertebrates from the Upper Becerra (Becer-
ra Superior) Formation, Valley of Tequixquiac, Mexico, with notes on other
Pleistocene forms. Univ. Michigan, Mus. Paleon., Contrib., 12: 47-96.

Howard, Hildegarde. 1935. The Rancho La Brea Wood Ibis. Condor, 37: 251-253.

. 1938. The Rancho La Brea Caracara: a new species. Carnegie Inst. Wash-
ington, Publ. 487: 217-240.

. 1940. A new race of caracara from the Pleistocene of Mexico. Condor, 42:

41-44.

. 1946. A review of the Pleistocene birds of Fossil Lake, Oregon. Carnegie

Inst. Washington, Publ. 551-: 141-195.

. 1965. A new species of cormorant from the Pliocene of Mexico. So. Cali-
fornia Acad. Sci., Bull., 64: 50-55.

. 1968. A preliminary report on Pleistocene birds of central Mexico. Geol.

Soc. Amer., Abstr. 1968 Annual meeting, Mexico City, Mexico, p. 142.

1969

Fossil Birds from Central Mexico

11

Lowery, George H., Jr., and Walter W. Dalquest. 1951. Birds from the State
of Veracruz, Mexico. Univ. Kansas, Mus. Nat. Hist., Publ., 3: 531-649.

Miller, Loye. 1943. The Pleistocene birds of San Josecito Cavern, Mexico. Univ.
California, Publ. ZooL, 47: 143-168.

. 1944. A Pliocene flamingo from Mexico. Wilson Bull., 56: 77-82.

Murray, Bertram G., Jr. 1967. Grebes from the Late Pliocene of North America.
Condor, 69: 277-288.

Palmer, R. H. 1926. Tectonic setting of Lago de Chapala. Pan-American Geol.,
45: 125-134.

Short, Lester L. 1966. A new Pliocene stork from Nebraska. Smithsonian Misc.
Coll., 149 (9): 1-11.

Storer. Robert. 1954. A fossil thrasher from the Pleistocene of Mexico. Wilson
Bull, 66: 144-145.

Wet more, Alexander. 1949. The Pied-billed Grebe in ancient deposits in Mexico.
Condor, 51: 150.

Accepted for publication June 20, 1969.

LOS

ANGELES

COUNTY

MUSEUM

CONTRIBUTIONS
(Z'fi IN SCIENCE

Number 173

June 30, 1969

FOSSIL LANTERNFISH OTOLITHS OF CALIFORNIA, WITH
NOTES ON FOSSIL MYCTOPHIDAE OF NORTH AMERICA

By

John E. Fitch

i

Los Angeles County Museum of Natural History • Exposition Park
Los Angeles, California 90007

CONTRIBUTIONS IN SCIENCE is a series of miscellaneous technical papers
in the fields of Biology, Geology and Anthropology, published at irregular intervals
by the Los Angeles County Museum of Natural History. 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 Robert J. Lavenberg, Los Angeles County Museum of Natural
History, 900 Exposition Boulevard, Los Angeles, California 90007.

Robert J. Lavenberg
Managing Editor

INSTRUCTIONS FOR AUTHORS

Manuscripts for the LOS ANGELES COUNTY MUSEUM, CONTRIBU-
TIONS 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. Al-
though priority will be given to manuscripts by staff members, or to papers dealing
largely with specimens in the Museum’s collections, 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 Edi-
torial Committee. They may also be subject to review by competent specialists out-
side 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.

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. Alterations or changes in the manuscript after galley proof will
be billed to the author. Unless specifically requested, page proof will not 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.

Dorothy M. Halmos
Editor

FOSSIL LANTERNFISH OTOLITHS OF CALIFORNIA, WITH
NOTES ON FOSSIL MYCTOPHIDAE OF NORTH AMERICA

By John E. Fitch1

Abstract: Otoliths (sagittae) from 30 species of lantern-
fishes (Myctophidae) are known from the North American
Cenozoic: 15 from Plio-Pleistocene (California), 13 from Mio-
cene (7 California; 5 Veracruz, Mexico; 1 Jamaica), and 2 from
Eocene (California). The high incidence of species from Cali-
fornia deposits appears to be a reflection of the proximity of the
shoreline to preferred myctophid habitat (deep water), rather
than a more abundant fauna, but both factors could be involved.

In addition to the 30 species known from their otoliths,
skeletal remains (imprints) of seven kinds of lanternfishes have
been found in Miocene diatomites of southern California. Oto-
liths in some of these imprints make it possible to compare them
with fossil myctophid sagittae that are unaccompanied by skeletal
remains.

Among the fish families inhabiting the ocean off California, only Cottidae
and Scorpaenidae contain more members than Myctophidae. A recent listing
by Fitch and Lavenberg (1968) shows 33 species of myctophids in 21 genera,
while Berry and Perkins (1966) note an additional eight species and two
genera from just outside the boundaries of the area covered by the California
list. In view of this great variety of species and the density of several nearshore
populations ( e.g ., Paxton [1967] reports 16,500 specimens captured in 48 tows
of a midwater trawl in the San Pedro Basin off southern California), it is not
surprising that lanternfishes are well represented in California’s fossil record.

Fitch (1966, 1967, 1968) and Fitch and Reimer (1967) have noted myc-
tophid otoliths in Pleistocene and Pliocene deposits of southern California, and
David (1943) described Lampanyctus bolini and L. petrolifer from skeletal
imprints in Miocene diatomites excavated in the Santa Monica Mountains of
southern California. Bagg (1912) illustrated an otolith (sagitta) from Tarle-
tonbeania crenularis (PI. 28, fig. a) in his report on Plio-Pleistocene Foramin-
ifera of San Pedro, but he failed to identify it except as an otolith.

Since 1963, I have investigated fossil exposures whenever possible in an
effort to shed light upon the fish faunas that lived here during various periods
of the earth’s history. To accomplish this, I have developed a routine for wash-
ing, screening, and sorting fossiliferous matrix that is very effective for finding
otoliths, teeth, vertebrae, and other fish remains (Fitch, 1969). To date, 12
Pleistocene and Pliocene deposits have yielded more than 60,000 otoliths
(sagittae) representing over 150 species, and more than 4,500 of these otoliths
have been from 15 kinds of myctophids (Table 1). In fact, lanternfish ear

Research Associate, Los Angeles County Museum of Natural History; Marine
Biologist, California Dept, of Fish and Game, Terminal Island 90731.

1

2

Contributions in Science

No. 173

stones have been present in every fossil deposit I have examined in southern
California that is Miocene or younger in age, and from which I have gleaned
more than 1,000 otoliths (Table 2 and unpublished data).

PLEISTOCENE AND PLIOCENE
Sites Investigated

Since marine fishes from these two epochs, representing possibly 1 1 mil-
lion years of time (Downs, 1968), are still living today (Fitch, 1969), I have
combined the two in this report for convenience of discussion.

California’s youngest marine Pleistocene, Palos Verdes Sand, was depos-
ited between 95,000 and 130,000 years ago (Fanale and Schaeffer, 1965),
during a period when ocean temperatures were higher than today at the lati-
tude of San Pedro (Fitch, 1964, 1966). Three southern California sites that
I have investigated rather thoroughly (“So. Calif, sites” in Table 1) have
yielded over 3,500 otoliths from at least 61 species, but only four otoliths
(three species) were from myctophids. The sandy matrix containing these
fossils appears to have been laid down at relatively shallow depths, possibly
not exceeding 60 feet. A deposit exposed during highway construction north
of Areata (listed as Crannell Road in this report) appears to be a northern
California equivalent of the Palos Verdes Sand. Two otoliths, of more than
800 found in material from this site, were from two species of lanternfish
(Table 1).

Two widely separated deposits of San Pedro Sand (Miraflores Street and
Ventura Freeway) have yielded fish remains indicating that cold oceanic
temperatures prevailed when they were being laid down (Fitch, 1967, 1969).
No age estimate is available for these early Pleistocene beds, but they too
appear to have been formed at depths shallower than 60 feet. Only nine of
nearly 4,000 otoliths from these two deposits were from myctophids (four
species).

I have investigated Timms Point Silt and its equivalent at Newport Beach
(Back Bay), and San Pedro, and near Carpinteria (Bates Road), and otoliths
are abundant at all three localities. No age estimate is available for Timms
Point Silt, but it is the oldest marine Pleistocene of southern California (Fitch,
1968, 1969), and appears to have been laid down at depths of 400 to 600 feet
when ocean temperatures were colder than at the same latitudes today. The
109 myctophid otoliths found in these three deposits were from nine species.

The youngest marine Pliocene of southern California, Lomita Marl, has
been estimated as being 3.04 million years old (Obradovich, 1965). Possibly a
ton of matrix from a single site yielded over 24 thousand otoliths when it was
washed, screened, and sorted. Among the more than 80 kinds of teleosts iden-
tified from these remains were many mesopelagics, a few bathypelagics, and a
half-dozen kinds of locally extinct northern fishes. The 14 species of myc-
tophids (over 3,100 otoliths) make this the richest assemblage of fossil lan-

Lanternfish Otoliths Found in Pleistocene and Pliocene of California

1969

Fossil Lanternfish Otoliths

3

q a

C/5 H

XX

11

Q-1

D- c/3
> <U

Is

=2

13 C

<3 £V

>3

dPQ
> •
£ ^
Z «

C/3

s .
£ £
H

c H

<u >
>H

qo +J

s

o*

3s

o'09

c/5

r- co

(N *-h

CO

(N

N N 't
wo

<N Tt
Tf

' — 'O n n

(N 'T

VO

-H <N

*-1 o\

<N (N

O

VO

\OHh

oc

^Hiooor^ooNovt^-

(N n ro

<N

r- ri

OO

On >o

^ £

(N

-

-

CO

00

VO

Tj- (N

HO00

|

oo

O

CO

V-)

|

<43

6»a

a

£

O

Q, 021

03 a a

* §

-a

S"a
a a

•r-

a,

) 7s 05
. r*»

05

C

05

a C
a

O

dee a

03

5 ©

^ ^ ja

cqUQ^

05

O 05

a ca
■a a

a, oo ^ . - -

? 2: ^ ~ ik
r*

05 ^ 05

a s s
a a

mi

B y 05^

|t 2 S

Q-. co ^—i

a 05

a £
a a

Ps

"a a
a a
a. a.

a a

kj Kj

& 8-5 § 8

s5°uh
S-g* g.s I
mi | g

— a o d-a b
H ^ S -S K 5
a s ^ q o e
o o*a ^-a

^5|,g

vo «/o
co

(N r-
(N

VO On
00

O o|-
'T

-H «0

co

m r-~
r-

00

£ C/3

^ <u
o’B
o g.

T3 1/3

£3

g-C
o a
o o

£o o

3s

a

o o

HZ

T0fal otoliths 3535 >800 2746 1200 2601 53 1243 24,299 5103 4285 1230 11,590

Total species >61 >28 >30 >40 >53 14 >37 >83 >55 >48 21 >50

*otoliths gleaned without aid of microscope

4

Contributions in Science

No. 173

ternfishes known in the world. The material appears to have been laid down
at depths exceeding 600 feet.

Two exposures of Pliocene that appear to be referable (or equivalent) to
the Pico Formation were sampled extensively during recent years. One of
these, on the mesa south of upper Newport Bay, yielded over 5,000 otoliths
(1,186 from lanternfishes) representing at least 55 species. Material from the
other deposit, in downtown Los Angeles, contained over 4,000 otoliths (122
from myctophids) from 48 species. Sagittae from other mesopelagics and from
several bathypelagics were present in both deposits, as were earstones from a
few locally extinct northern fishes. Based upon the fish remains, these beds
were laid down in depths of at least 600 feet at a time when ocean temperatures
were much colder than they are at the same latitudes today.

Myctophid otoliths have turned up in two other Pliocene deposits (Signal
Hill and San Diego Fm.), but these were gleaned by student volunteers who
were searching for mollusks and did not use a microscope. Thus, neither the
species nor the numbers offer reliable indices as to the lanternfish components
of these deposits. Five of the six species that were found at these two sites have
relatively large otoliths that are difficult to miss when searching through fossil-
iferous matrix, even without the aid of a microscope. The San Diego Forma-
tion yielded nearly 1 1,600 sagittae from about 50 species, all but one of which
(an extinct sciaenid) can be found living at the same latitude today. Most are
inhabitants of shallow water, but otoliths of mesopelagics and bathypelagics
are also present, although these do not show signs of abrasion by currents and
wave action as do the earstones of the shallow water forms. The sagittae of the
offshore species apparently arrived in this bed later than those of the inshore
fishes, but all appear to have been deposited in relatively shallow water, pos-
sibly no deeper than about 60 feet.

Species Accounts
Benthosema sp. (Fig. lb).

Benthosema sagittae were found in all three deep water Pliocene deposits,
and one of the three Pleistocene beds that were laid down in depths greater
than 400 feet. Until I have comparative material from B. glaciale and B. sub-
orbitale, the two most suspect species, they will need to remain unidentified.
B. glaciale is common in the western north Atlantic and has been reported
from the Bering Sea, whereas B. suborbitale is an inhabitant of oceanic waters
well offshore from California. The fossil otoliths are not from B. panamense,
a tropical species which is extremely abundant south of Magdalena Bay and
in much of the Gulf of California. In the Lomita Marl, Benthosema sagittae
comprised nearly four per cent of the 24,300 otoliths I found, so the species
was certainly common enough off California some 3 million years ago. Most
of these otoliths were from large adults and ranged in length to 2.5 mm, al-
though earstones of juveniles were also plentiful.

1969

Fossil Lanternfish Otoliths

5

Ceratoscopelus townsendi (Eigenmann and Eigenmann, 1889) (Fig. 1 o).

Otoliths of C. townsendi were found in all but one of the six Plio-Pleisto-
cene deposits that were laid down in depths greater than 400 feet, but they
were present in only one of the six shallow water deposits (Table 1). They
comprised about six per cent of the 5,100 otoliths from the Newport Mesa
locality, and among the nine species of myctophids in this deposits, only
Stenobrachius was more numerous. Paxton (1967) reported that C. town-
sendi has a center of diurnal distribution above 650 m and that it migrates into
the upper 10 m of water at night. It has been considered cosmopolitan in dis-
tribution, but recent investigations show that C. townsendi inhabits only the
eastern north Pacific (Basil Nafpaktitis, pers. comm.). The sagittae of a large
adult C. townsendi will be about 2.5 mm long; those I have found in fossil
deposits are mostly from large adults, and range in length to 2.7 mm.

Diaphus theta Eigenmann and Eigenmann, 1890 (Fig. 1 n).

Otoliths of D. theta were present in eight of my 12 Pleistocene and Plio-
cene deposits, being absent in the southern California Palos Verdes Sand (shal-
low deposition and warm water), in the Miraflores Street deposit (shallow
deposition and cold water), in the Bates Road exposure (moderate depth and
cold water), and at Signal Hill (material sorted without microscope). They
were most abundant at the three Pliocene sites representing deep deposition
and cold temperatures, comprising more than two per cent of the total otolith
yield from the Lomita Marl, and over five per cent of the total from the New-
port Mesa deposit. D. theta is restricted to the Pacific Subarctic water mass and
the transition area off western North America (Paxton, 1967), so its north-
south distribution extends from the Gulf of Alaska to about Cedros Island. It
appears to have a vertical distribution similar to that of C. townsendi. The
largest individuals inhabit more northerly waters and sagittae from these large
adults will average about 3.0 mm. All sizes of otoliths were found in the vari-
ous fossil deposits but most were from adult fish.

Electrona rissoi (Cocco, 1829) (Fig. 1/).

Otoliths of E. rissoi were most abundant in the Timms Point Silt, but even
there they were not common. They were present in only four deposits, two
Pleistocene and two Pliocene, but there is no pattern in their occurrence. One
otolith was found in Palos Verdes Sand representing shallow deposition and
warm water, and two were in the Signal Hill material which also represented
shallow deposition but at normal temperature for the latitude. The remaining
E. rissoi otoliths (15) were from deep water deposits that were laid down when
the ocean was colder than today.

Its present day distribution in the eastern north Pacific does not indicate
that it has a preference for cold water. Almost all records of this rarely caught
fish have been made offshore from the area between Point Conception and

6

Contributions in Science

No. 173

Figure 1. Myctophid sagittae found in Pleistocene and Pliocene of California.
Lengths (in mm) are given for each otolith; notations are made regarding its posi-
tion in the skull (left or right); and ages and localities of fossil deposits are indi-
cated. a. Symbolophorus califomiensis, 4.3, l, Lomita Marl, late Pliocene, San Pedro;
b. Benthosema sp., 2.3, /, Lomita Marl; c. Lampadena urophaos, 6.5, r, Lomita Marl;
d. Myctophum nitidulum, 3.1, l, Lomita Marl; e. Tarletonbeania crenularis, 2.0, l,
Lomita Marl; f. Stenobrachius leucopsarus, 2.1, /, Lomita Marl; g. T riphoturus mexi-
canus, 1.2, r, Lomita Marl; h. Lampanyctus regalis, 2.0, l, Lomita Marl; i. Parvilux
ingens, 0.8, /, Lomita Marl; j. Lampanyctus ritteri, 1.3, r, Lomita Marl; k. Protomyc-
tophum crocked, 1.7, r, Timms Point Silt, early Pleistocene, San Pedro; 1. Electrona
rissoi, 3.4, r, Timms Point Silt; m. Notoscopelus resplendens, 3.9, /, San Diego Fm.,
Pliocene, San Diego; n. Diaphus theta, 2.5, l, Lomita Marl; o. Ceratoscopelus town-
sendi 2.5, r, Pico Fm., Pliocene, Newport Beach. Photographs by Jack W. Schott.

1969

Fossil Lanternfish Otoliths

7

Magdalena Bay. Sagittae from a large adult will average about 3.8 mm in
length. The 18 otoliths noted here ranged from about 2.1 to 4.2 mm long.

Lampadena urophaos Paxton, 1963 (Fig. 1c).

Otoliths of L. urophaos have been found in only two deposits, both repre-
senting deposition at depths of at least 600 feet and during periods when ocean
temperatures were colder than usual. According to Paxton (1967), L. uro-
phaos has a diurnal center of distribution below 650 m, but migrates at night
to within 50 m of the surface. Nafpaktitis and Paxton (1968) list its geo-
graphical range as between 25° N and 42° N in the central and eastern Pa-
cific, and between 21° N and 38° N in the western Atlantic. They illustrate
the sagittae of the known species of Lampadena, and present characters for
differentiating them. L. urophaos has the largest otoliths of any lanternfish in
the eastern Pacific; those from a full grown fish will exceed 7.5 mm in length.
Only one of the six sagittae reported upon in this paper was entire, but the
broken fragments were unmistakable because of their size and the 90° notch
at the posterodorsal angle.

Lampanyctus regalis (Gilbert, 1891) (Fig. 1 h) .

Otoliths of L. regalis were found only in the Lomita Marl and Newport
Mesa (Pico Fm.) deposits, both representing cold water and depths of at least
600 feet. Paxton (1967) states that the diurnal center of distribution for
L. regalis is below 650 m, but at night it migrates to within 50 m of the surface.
He lists it as an inhabitant of the “Subarctic-Central Pacific” water mass, rang-
ing from the Gulf of Alaska to off Cape San Lucas on our coast.

The sagittae of S. regalis most closely resemble those of S. ritteri, but are
somewhat thicker at comparable sizes, and attain a larger maximum size.
There is a fair amount of variation in the ratio of length into height, but otoliths
from a typical large adult were 1.7 mm long by 2.3 mm high. The fossil otoliths
were mostly from adults, and range to about 1 .9 mm long.

Lampanyctus ritteri Gilbert, 1915 (Fig. 1/).

Otoliths of L. ritteri were present in the same two deposits that contained
L. regalis, but there were fewer than half as many. Paxton (1967) states that
during daylight hours L. ritteri has a center of distribution above 650 m, and it
migrates to within 50 m ot the surface at night, but no shallower. Thus, it in-
habits shallower depths during the day than L. regalis, but both move into
identical depths at night. Paxton classifies L. regalis as a “Subarctic-Transi-
tional” species based upon the water masses it inhabits. It is restricted to the
eastern north Pacific, having been recorded from the vicinity of Vancouver
Island south to about Magdalena Bay. Otoliths from large adults will average
about 1.5 mm long by 1.8 mm high, but the ratio of length into height is not
constant. The fossil sagittae were primarily from young adults.

8

Contributions in Science

No. 173

Myctophum nitidulum Garman, 1899 (Fig. 1 d).

Only a single otolith of M. nitidulum has been identified from the Plio-
Pleistocene of California. This otolith, from the Lomita Marl, is unmistaka-
ble. A Myctophum in the western Atlantic is reported to be conspecific with
M. nitidulum (described from off the west coast of Mexico at lat. 27° 50' N),
but differences in the sagittae of the two are so distinctive that they are readily
separable on the basis of otolith shape alone. Paxton (1967) did not note
M. nitidulum from the San Pedro Basin, and Berry and Perkins (1966) cap-
tured very few during offshore fishing with a midwater trawl. At times, how-
ever, large numbers of M. nitidulum can be attracted to a light suspended
above the ocean’s surface at night, especially in areas well offshore. Their range
in the eastern Pacific is from north of San Francisco to south of Cedros Island.
Sagittae of large adults will average about 2.5 mm long.

Notoscopelus resplendens Richardson, 1844 (Fig. Ira).

Five otoliths of N. resplendens were found in the Lomita Marl, and two
were among the sagittae gleaned from the San Diego Formation by student
volunteers. Paxton (1967) did not take N. resplendens in the San Pedro Basin,
but Berry and Perkins (1966) captured them at about 20 offshore stations
between San Francisco and Cape San Lucas. Only three of these stations
yielded more than three individuals, however, so they appear to be rare even
today. The elongate sagittae are impossible to confuse with any other species
that occurs off our coast. Those from a large adult will be 3.7 mm long or
longer. The fossil material was mostly from somewhat smaller individuals.

Parvilux ingens Hubbs and Wisner, 1964 (Fig. 1/).

Sagittae of P. ingens were present only in the Lomita Marl deposit, where
sagittae of 13 other myctophid species were found. Paxton (1967) states that
P. ingens has a diurnal center of distribution below 650 m, and that it migrates
to within 50 m of the surface at night, but was never taken shallower. He in-
cludes it with Symbolophorus californiensis and Lampadena urophaos as a
“Transitional species,” with a coastwise range extending from about the lati-
tude of Cedros Island to just north of San Francscio. The sagittae of P. ingens
could be confused with those from Lampanyctus ritteri and L. regalis, but they
differ in having a rounded dorsal profile and a deeply arched (concave) inner
face. Otoliths from a large adult measured 1.4 mm long by 2.0 mm high. Seven
of the eight from Miraleste Canyon were from large adults.

Protomyctophum crocked (Bolin, 1939) (Fig. 1/:).

Sagittae of P. crockeri were found at two Pleistocene and two Pliocene
sites, all representing deposition during cold periods, and three of the four
having been laid down in depths of about 600 feet. They ranked fifth numeric-

1969

Fossil Lanternfish Otoliths

9

ally among the 14 kinds of myctophid otoliths found in the Lomita Marl. All
of the deposits containing Protomyctophum otoliths also contained earstones
of Diaphus, Stenobrachius, and Tarletonbeania, and all but one contained
Ceratoscopeius. According to Paxton (1967), P. crocked inhabits depths of
400 to 600 m during daytime hours, and has been captured within 50 to 150 m
of the surface at night, but never shallower than 50. He classifies it as a “Sub-
arctic-Central Pacific species” according to the water masses it inhabits. It has
been captured from south of Cape San Lucas to off Vancouver Island. In more
northerly waters its range overlaps that of P. thompsoni, and since the otoliths
of these two species cannot be separated, some of those from the cold water
deposits ( i.e ., Miraleste Canyon, Timms Point, and downtown L.A.) could be
from P. thompsoni. Sagittae from large adults will average about 2.4 mm in
length. Those in the four fossil deposits were mostly from adult fish, and range
to 2.6 mm in length.

Stenobrachius leucopsarus (Eigenmann and Eigenmann, 1890) (Fig. If).

S. leucopsarus sagittae turned up in every Plio-Pleistocene deposit I have
investigated except Crannell Road (northern California equivalent of Palos
Verdes Sand representing warm ocean temperatures and shallow deposition),
and Signal Hill, where they could have been overlooked because of their small
size. In the Miraflores Street deposit, a single S. leucopsarus otolith was the
only myctophid represented, while at three localities they were the most
abundant of the lanternfish otoliths {i.e., Timms Point, Miraleste Canyon, and
Newport Mesa). The 606 otoliths in the Newport Mesa deposit comprised
more than ten per cent of the total otolith yield at that site, while the 1,120
found in Miraleste Canyon amounted to only about five per cent of the sagittae
from there. Paxton (1967) found their center of distribution during daylight
hours at 650 m; at night they migrate from there to within 50 m of the surface,
but seldom shallower. He categorizes them as a “Subarctic-Transitional spe-
cies” ranging from about Cedros Island to the Aleutian Islands. The range of
a close relative, S. nannochir, overlaps that of S. leucopsarus in northern Cali-
fornia during some years but is north of there during most years. Sagittae of
these two species are indistinguishable, so some of those from the cold water
deposits, particularly Timms Point, Miraleste Canyon, and Newport Mesa,
could be from S. nannochir. The otoliths of a large adult will average about
1.8 mm long. Those in the fossil deposits were mostly from adults, but sagittae
from juveniles were also plentiful.

Symbolophorus calif orniensis (Eigenmann and Eigenmann, 1899) (Fig. 1 a).

Otoliths of S. calif orniensis were present in five deposits, but only in the
Lomita Marl and San Diego Formation were more than two sagittae found.
S. calif orniensis ranges from about off Cedros Island to the vicinity of the
Oregon-California border, except during years when the ocean is quite warm,

10

Contributions in Science

No. 173

when it may reach waters off British Columbia. Its presence in the Crannell
Road deposit north of Areata was useful in associating that exposure with the
Palos Verdes Sand. During daylight hours its center of distribution is above
650 m, and from there it migrates into the upper 10 m of water at night (Pax-
ton, 1967). It is readily attracted to a light hung above the surface at night.
Paxton categorizes S. calif or niensis as an inhabitant of “Transitional” water,
along with Lampadena urophaos and Parvilux ingens. Other species of Sym-
bolophorus occur in the central Pacific and south of Cape San Lucas, but their
ranges never overlap that of S. calif or niensis . The only larger sagittae among
Californian lanternfishes are those of Lampadena. Otoliths of large adult
S. calif orniensis will average about 4.0 mm long; most of those in the five fossil
deposits were from adult fish.

Tarletonbeania crenularis (Jordan and Gilbert, 1880) (Fig. \e).

Otoliths of T. crenularis were present in seven of the 12 Plio-Pleistocene
deposits investigated, but they were abundant only in the Lomita Marl, where
they comprised more than one per cent of the total otoliths. Only one of the
six shallow water exposures contained an earstone from this species, and it
(Ventura Freeway) was one of the three deposits laid down when ocean
temperatures were colder than usual. Paxton (1967) categorizes T. crenularis
as a “Subarctic-Transitional species” along with Diaphus theta, Stenobrachius
leucopsarus, and Lampanyctus ritteri. Its coastwise range is from about off San
Diego to the Aleutian Islands. Its center of distribution during daylight hours
is above 650 m, and it migrates from there into the upper 50 m at night. As
with Myctophum nitidulum and Symbolophorus calif orniensis, Tarletonbeania
is readily attracted to a light suspended above the water’s surface at night. The
sagittae of large adult T. crenularis will average about 1.8 mm in length. Most
of those from the various Plio-Pleistocene deposits were from adult fish, and
some were longer than 2.0 mm. Bagg (1912) illustrated an otolith of T. crenu-
laris from the Timms Point Silt at San Pedro, but he identified it only as an
otolith.

Triphoturus mexicanus (Gilbert, 1890) (Fig. lg).

Otoliths of T. mexicanus were found at only three sites, all representing
deposition at depth and during periods of cold ocean temperatures. T . mexi-
canus has a diurnal center of distribution above 650 m, and migrates to 50 m
at night, but seldom shallower, according to Paxton (1967). It ranges along
the outer coast of Baja California to about the latitude of San Francisco, and
is categorized as having an “Eastern Equatorial” distribution by Paxton
(1967). The extent of its southern distribution is clouded by the contention
of some myctophid specialists (as yet unpublished) that the Triphoturus in the
Gulf of California and south of Cape San Lucas (to Panama and Peru) repre-
sent species that are distinct from T. mexicanus. Sagittae of T. mexicanus are

1969

Fossil Lanternfish Otoliths

1

never higher than long; those from large individuals will average about 1 .5 mm
long by 1.4 mm high, but this ratio of length into height may vary somewhat.
Most of the 1 1 fossil otoliths were from young adults.

MIOCENE MYCTOPHIDS

Since teleost otoliths and shells of marine mollusks usually are comprised
of the same type of calcium carbonate (i.e., aragonite), one can assume that
a fossil deposit containing an assortment of sea shells likely will contain fish
earstones also. By using this as a working hypothesis, I have routinely sampled
every shelly fossil exposure I have encountered if it is friable or can be broken
down without destroying or dissolving the aragonitic constituents. Not all such
beds produce otoliths (e.g., oyster reefs and sites where only calcitic sea shells
have not leached out generally yield very few sagittae), but deposits occasion-
ally are found that are rich with them. Such is the case with several Miocene
outcrops east of Bakersfield where I have conducted extensive investigations.
Although these Bakersfield deposits contain an abundance of otoliths, lantern-
fish sagittae generally are scarce.

No attempt will be made in this report to describe new species based upon

Table 2

Lanternfish Otoliths from Miocene of North America

California

SPECIES

Hwy.

178

Round

Mt.

Silt

Barker’s

Rnch.

Pomona
Fwy. '

Jamaica

Bowden

Fm.

Mexico

Veracruz

Diaphus sp.

261

Diaphus sp.

1

Diaphus sp.

56

Diogenichthys sp.?

1

Hygophum intermedius

1

Hygophum sp.

1

Lampanyctus sp.

353

15

Lampanyctus sp.

>100

Myctophum acutus

4

Myctophum pabloensis

2

Myctophum sp.

1

Notoscopelus sp.

57

Symbolophorus biatlanticus

2

Total myctophid otoliths

318

356

15

>100

56

10

No. myctophid species

2

4

1

1

1

5

Total otoliths

>3,000

>4,000 >25,000

>150

993

21

Total species

>20

>40

>70

>15

>50

13

12

Contributions in Science

No. 173

otoliths alone, because in 1968 some diatomite deposits were found that are
rich with skeletal imprints of myctophids which still retain their otoliths
(Fig. 3). Since the diatomite is the same age (middle to late Miocene) as the
friable shelly deposits east of Bakersfield, there is an excellent chance of lo-
cating lanternfish imprints to match the otoliths. If “imprints with otoliths”
can be matched to “imprints without otoliths” and to “otoliths without im-
prints,” a great deal of cluttering up of the literature through multiplication of
specific names can be avoided, and there is hope of obtaining an accurate
synoptic picture of the Miocene fish fauna.

A locality beside State Highway 178 (east of Bakersfield) contains oto-
liths that represent deeper water species than do the faunas in such well known
members of the Temblor Formation as Sharktooth Hill, Round Mountain Silt,
and “Barker’s Ranch.” Although many of the mollusk shells in these beds
near Highway 178 have leached out so that only their casts remain, sufficient
shelly matrix can be found to make sampling worthwhile. Otoliths from two
species of myctophids are abundant, and are in a better state of preservation
than many Pleistocene sagittae which may be 10 million years younger. The
commonest lanternfish is a species of Diaphus (Fig. 2b, Table 2). The 261
otoliths from this fish range in length to 2.8 mm, and comprise about 12 per-
cent of the total otolith yield from these beds. The other lanternfish from the
Highway 178 deposit (Fig. If) is from a species of Notoscopelus. Only 57
otoliths from this species have been found, and most of these are in relatively
poor condition. They range in length to 3.1 mm. Morid (family Moridae)
sagittae that pass through 20 mesh screen are the most abundant teleost of
more than 30 species found at this site.

Narrow lenses of shelly matrix in the Round Mountain Silt south of
Stanford University locality 2121 (Keen, 1943) have yielded nearly 4,400
otoliths representing at least 50 species. Four lanternfishes are represented
among the 356 myctophid sagittae, but for three of these ( Hygophum sp.,
Fig. 2g\ Diaphus sp., Fig. 2 a\ and Myctophum sp., Fig. 2 h) there is but one
otolith each, while 353 of the earstones are from a species of Lampanyctus
(Fig. 2e ). Although the Lampanyctus sagittae amounted to slightly more than
eight per cent of the total otolith yield from the Round Mountain Silt, the
number probably would be more than double what it is if I had searched all of
the fine residue from my samples. About 99 per cent of the Lampanyctus oto-
liths from Round Mountain Silt have been less than 1 mm long, so they pass
through 20 mesh screens with ease. Although they are retained in 30 mesh
screen, I have purposely avoided using mesh that fine for processing most of
the material I have dug from this site. To sort through residue that passes
through 20 mesh, but is retained by 30 mesh, is extremely time consuming;
so when initial sub-sampling shows that very few species are escaping the 20
mesh screen, I examine only enough fine residue to assure myself of an ade-
quate quantity of the tiny otoliths.

The oldest beds of the Temblor Formation that I have examined for oto-

1969

Fossil Lanternfish Otoliths

13

Figure 2. Myctophid sagittae found in Miocene of California and Jamaica. Lengths
(in mm) are given for each otolith; notations are made regarding its position in the
skull (left or right); and localities of fossil deposits are indicated, a. badly worn
Diaphus sp., 1.2. I, Round Mountain Silt, Bakersfield, Calif.; b. Diaphus sp., 2.7, /,
Hwy. 178, Bakersfield; c. Diaphus sp., 4.2, i, Bowden Fm., Jamaica; d. Lampanyc-
tus, sp., 1.2, r, Barker’s Ranch, Bakersfield; e. Lampanyctus sp., 1.0, r, Round
Mountain Silt; f. Notoscopelus sp., 2.6, l, Hwy. 178; g. Hygophum sp., 1.3, I,
Round Mountain Silt; h. Myctophum sp., 2.8, r, Round Mountain Silt. Photographs
by Jack W. Schott.

liths are the sandy “Barker’s Ranch” exposures. These beds outcrop in dozens
of canyons, gullies, and road cuts between the Kern River and Pyramid Hill,
and each layer contains a slightly different admixture of otoliths. The fish
fauna is predominantly from shallow water (judged by today’s standards), but
there is a sprinkling of remains from mesopelagics and bathypelagics at most
sites. Sagittae from croakers (family Sciaenidae) are dominant in all the beds,
from the standpoint of size, numerical abundance, and kinds. Myctophid oto -
liths, while not quite at the other end of the scale, have generally been small,
scarce, and from only one species — a Lampanyctus (Fig. 2d) which is iden-
tical to the Lampanyctus in Round Mountain Silt.

Skeletal imprints of myctophids are abundant in Miocene diatomites and

14

Contributions in Science

No. 173

Figure 3. Left (upper) and right sagittae in situ in skeletal imprint of Lampanyctus
sp. on diatomite slab from west of Pomona, California. Photograph by Jack W.
Schott.

shales, and based upon such imprints, David (1943) described Lampanyctus
petrolifer and L. bolini from specimens collected in the Santa Monica Moun-
tains. Examination of her type material (now in the Los Angeles County Mu-
seum of Natural History) shows that neither species belongs in the genus
Lampanyctus. Her Lampanyctus petrolifer is a Lepidophanes, while her Lam-
panyctus bolini is a Diaphus (R. J. Lavenberg, pers. comm.). None of the
specimens she selected as “cotypes” (paratypes) of L. bolini, including the

1969

Fossil Lanternfish Otoliths

15

imprint on the same slab as her “type,” belongs to the same genus ( Diaphus )
as the holotype. To date, seven genera of lanternfishes are represented in the
Miocene of southern California as skeletal imprints (R. J. Lavenberg, pers.
comm.), but until last year (1968) no imprint found had retained its otoliths.
At that time, during freeway construction and other excavating activities west
of Pomona, several beds of diatomite were exposed that were rich with fish
imprints, especially myctophids, and most have retained identifiable otoliths
(Fig. 3). In addition, loose earstones are scattered throughout these deposits
and impacted in some of the abundant coprolites.

Although I have examined material from Miocene deposits in Florida.
North Carolina, Virginia, Maryland, and Jamaica, none of these except the
Bowden Formation (Jamaica) has contained lanternfish otoliths. A 25-pound
field sample from the Bowden Fm. contained 993 sagittae from more than 50
species of teleosts, including several kinds of mesopelagics. Fifty-six of the 993
otoliths were from a species of Diaphus (Fig. 2c). Nafpaktitis (1968), in his
report on taxonomy and distribution of Diaphus in the north Atlantic, shows
at least 15 species presently inhabiting waters in the vicinity of Jamaica. My
comparative collection contains otoliths from 13 of these extant species, but
I am lacking in sagittae from D. garmani and D. subtilis. The fossil otoliths
are closest to D. problematicus among the species 1 have compared, but they
differ from D. problematicus in several significant features, particularly in the
texture and amount of spination on the ventral margin. Until sagittae from the
other critical species can be examined, it seems both inappropriate and inad-
visable to describe these fossil otoliths as new.

The only other fossil myctophids known from the Miocene of North
America are five species described (as four) by Weiler ( 1959) from otoliths
found in a core from oil well drilling near Veracruz, Mexico. His holotype of
Hygophus [=Hygophum] intermedins appears to be correctly assigned generic-
ally, but the otolith he illustrates as figure 19 and calls “ Hygophus interme-
dius?” is not Hygophum. From its outline and size, it could be Diogenichthys.
but Lampanyctus is also a possibility. In a later publication. Weiler (1968)
placed his “Otol. (Nyctophidarum) acutus"' [s/c.] in the genus Myctophum.
which, from his figured specimens, undoubtedly is correct. His other two spe-
cies, however, belong to different genera than those he assigned them to. His
“Ceratoscopelus? pabloensis” very likely is a Myctophum, which has a sagitta
characterized by a deeply rounded (convex) ventral margin, a slight concavity
at the posterodorsal angle, and an antirostrum that is “high up” as in Hygo-
phum. Ceratoscopelus, on the other hand, has an otolith with a gently rounded
ventral margin, an elongate rounded rostrum, and an entirely different shape
and placement of the antirostrum (see Fig. 1 o, which, although C. townsendi,
is very similar to C. maderensis, the type species of Ceratoscopelus) . Finally,
Weiler’s “Nyctophus [= Myctophum J biatlanticus ” probably is Symbolopho-
rus, but the possibility of it being Diaphus should not be overlooked. The
evenly curved dorsal and ventral contours, ratio of height into length, and

16

Contributions in Science

No. 173

Figure 4. Otoliths from two species of Diaphus found in Rose Canyon Fm., Eocene.
San Diego County, Calif, a. left sagitta, 1.3 mm long; b. right sagitta, 3.0 mm long.
Photographs by Jack W. Schott.

numerous rounded crenations on the dorsal and ventral margins are typical
of Symbolophorus. The ventral margin of Diaphus would be spinose instead of
scalloped as in Weiler’s illustration of “N. biatlanticus,” but spines and other
frills often are worn down prior to or during fossilization.

EOCENE MYCTOPHIDS

Numerous Eocene beds are exposed in canyons, cliffs, gullies, road cuts,
and excavations in San Diego County between Mission Bay and Oceanside

1969

Fossil Lanternfish Otoliths

17

and inland as far as Poway, and some of these contain well preserved otoliths,
teeth, and other fish fragmentia. Although I have barely explored two or three
of these outcrops, I have found sagittae of several deep sea teleosts, including
two species of myctophids (genus Diaphus). Although solitary fish scales
found in Cretaceous chalks in Alabama are thought to be from lanternfishes
(S. P. Applegate, pers. comm.), there is no record of a myctophid skeletal
imprint or otolith prior to the Eocene. Danil’chenko (1960) reports lantern-
fish skeletal imprints from Middle Eocene beds in Russia, and Weiler (1968)
lists four species identified from otoliths found in European Eocene. Thus, the
Diaphus sagittae from the Rose Canyon Fm. north of San Diego match the
oldest positive records for the family elsewhere in the world.

Two of the three otoliths I have found to date are from a Diaphus with a
deeper ostium (the anterior portion of the sulcus or groove that shows on the
inner face) than cauda (the posterior portion of the sulcus). The ventral mar-
gin of this otolith (Fig. Ah) appears to be roughened, but it is spinose in better
preserved material. The other species, represented by a single small otolith with
an indentation in the anterodorsal margin (Fig. 4a), has a series of spines
along the ventral margin, and there is no visible difference in the depth of the
ostium and cauda. Since I have barely begun investigating these beds, and have
only three sagittae on hand, it would be premature to give them specific names
at this time. In view of the small quantity of matrix I have processed to date,
there should be no problem obtaining sufficient otoliths of both species to ren-
der satisfactory descriptions, and to place “type” or “referred” material where
it will be available to other researchers.

Discussion

Lanternfishes extend as far back into the fossil record of California as
they do anywhere in the world, and the presence of Diaphus otoliths in Eocene
deposits north of San Diego shows that a generic character which evolved in
this family upwards of 60 million years ago (at least) has remained stable
ever since. Diaphus is considered by many ichthyologists to be the most highly
specialized of the lanternfish genera, and its lengthy fossil record would cer-
tainly bear this out.

Although myctophid remains have not been found in the Oligocene of
North America, this probably reflects a shortage of accessible fossiliferous
strata that were laid down in deep water, rather than a lack of lanternfishes in
the seas bordering our continent during that epoch. Fish remains are abundant
at Pyramid Hill (Oligocene) east of Bakersfield, but these represent a shallow
water fauna, as do the teleost sagittae in the otolith rich marls and sands of the
Gulf Coast Oligocene.

The lanternfishes of our Miocene seas left a better record than their
predecessors, and as a result, otoliths from 13 species in 7 genera have been
found in deposits in California (7 species in 5 genera), Mexico (5 species in

18

Contributions in Science

No. 173

4 genera), and Jamaica (1 species). In addition, skeletal imprints are abun-
dant in Miocene diatomites and shales of southern California, and although
these are known to comprise at least seven kinds of lanternfishes, only two
species (in two genera) have been described so far. Thus, myctophid genera
which appear to have been well established throughout the Miocene are:
Diaphus (present since Eocene), Diogenichthys, Hygophum, Lampanyctus,
Lepidophanes, Myctophum, Notoscopelus, and Symbolophorus.

During the Pliocene and Pleistocene, otoliths from extant species were left
in many deposits throughout California, but not elsewhere on the continent.
Sagittae from all but Diogenichthys and Lepidophanes, among the eight genera
present during Miocene, were recovered from Pliocene and Pleistocene de-
posits and nine additional genera were noted: Benthosema, Ceratoscopelus,
Electrona, Lampadena, Parvilux, Protomyctophum, Stenobrachius, Tarleton-
beania, and Triphoturus. Six of the 12 Pliocene and Pleistocene deposits that
yielded myctophid otoliths were laid down at relatively shallow depths, pos-
sibly 60 feet or less, yet lanternfishes are not known to inhabit such shallow
water. In spite of their mesopelagic existence, their otoliths can arrive in shal-
low, nearshore deposits in any of several ways: (i) fishes can be eaten by a
predator in deeper waters and their otoliths transported to a shallow area in
the predator’s digestive tract, which they can pass through in a relatively un-
altered condition; (ii) mass mortalities can occur and the dead fishes can be
transported into shallow areas by currents or wind action, or in the stomachs
of scavengers; (iii) in undertaking diurnal migrations near the heads of sub-
marine canyons or where the bottom gradient rises abruptly near shore, stray-
ing, predation, and natural mortality can result in otolith deposition at shallow
depths; (iv) spawning migrations take some species (and their otoliths) into
shallower areas than they inhabit the rest of the year; and (v) deep areas which
contain otoliths of offshore species can be, and are, folded and faulted upward
so that subsequent deposition will be from shallow species, and vice versa.
Each of these factors was responsible at one time or another for sagittae of deep
water species being in shallow water deposits.

Acknowledgments

This study was made possible by research grants (GB-1244 and GB-6490)
from the National Science Foundation under the sponsorship of the Los An-
geles County Museum of Natural History Foundation.

Many individuals have contributed to the success of my research by fur-
nishing freshly caught fish or their otoliths, informing me of fossil exposures
or helping me excavate various sites or both, or by contributing ideas and
sharing their special talents. I especially appreciate the help given me by:
Warren Addicott, Shelton P. Applegate, Frederick H. Berry, Richard Bishop,
Rolf Bolin, Henry E. Childs, Jr., Leonard Coleman, James Craddock, William
L. Craig, Jules Crane, Spano Giacalone, Daniel J. Gotshall, Jack Hopkins,

1969

Fossil Lanternfish Otoliths

19

Richard Huddleston, I. C. Johnson, George Kanakoff, Roy Kohl, Robert Lea,
Basil Nafpaktitis, John Paxton, Roger Reimer, Mark Roeder, Robert Wisner,
and William Zinsmeister.

The bulk of the myctophid otoliths in my comparative collection are the
result of many hours of work on pitching ships and in the laboratory by Robert
J. Lavenberg and Richard McGinnis. My son, Richard A. Fitch, washed,
screened, and sorted hundreds of pounds of fossiliferous matrix for me. Mrs.
Loretta Proctor typed the manuscript in record time, and Jack W. Schott took
the excellent otolith photos for me. If I have failed to acknowledge assistance,
and I am sure to have overlooked someone, it has not been intentional.

Literature Cited

Bagg, R. M. 1912. Pliocene and Pleistocene Foraminifera from southern Califor-
nia. U.S. Geol. Surv., Bull., 513: 1-153.

Berry, Frederick H., and Herbert C. Perkins. 1966. Survey of pelagic fishes of
the California Current area. U.S. Fish Wildl. Serv., Fish. Bull., 65: 625-682.

Danil’chenko, P. G. 1960. Bony fishes of the Maikop deposits of the Caucasus.
Transl. from Russian 1967 by A. Mercado, Israel Prog. Sci. Transl., Jerusalem.
vii+247p.

David, Lore Rose. 1943. Miocene fishes of southern California. Geol. Soc. Amer.,
Spec. Pap., 43: 1-193.

Downs, Theodore. 1968. Fossil vertebrates of southern California. Berkeley, Univ.
Calif. Press. 61 p.

Fanale, F. P., and O. A. Schaeffer. 1965. Helium-uranium ratios for Pleistocene
and Tertiary fossil aragonites. Science, 149: 312-317.

Fitch, John E. 1964. The fish fauna of the Playa del Rey locality, a southern Cali-
fornia marine Pleistocene deposit. Los Angeles County Mus. Nat. Hist., Con-
tribs. in Sci., 82: 1-35.

. 1966. Additional fish remains, mostly otoliths, from a Pleistocene deposit

at Playa del Rey, California. Los Angeles County Mus. Nat. Hist., Contribs. in
Sci., 119: 1-16.

. 1967. The marine fish fauna, based primarily on otoliths, of a lower

Pleistocene deposit at San Pedro, California (LACMIP 332, San Pedro sand).
Los Angeles County Mus. Nat. Hist., Contribs. in Sci., 128: 1-23.

. 1968. Otoliths and other fish remains from the Timms Point silt (early

Pleistocene) at San Pedro, California. Los Angeles County Mus. Nat. Hist.,
Contribs. in Sci., 146: 1-29.

. 1969. Fossil records of certain schooling fishes of the California Current

system. Calif. Mar. Res. Comm., CalCOFI Rept., 13: 71-80.

Fitch, John E., and Robert J. Lavenberg. 1968. Deep-water teleostean fishes of
California. Berkeley, Univ. Calif. Press. 155 p.

Fitch, John E., and Roger D. Reimer. 1967. Otoliths and other fish remains from
a Long Beach, California, Pliocene deposit. So. Calif. Acad. Sci., Bull., 66:
77-91.

20

Contributions in Science

No. 173

Keen, A. Myra. 1943. New mollusks from the Round Mountain silt (Temblor)
Miocene of California. San Diego Soc. Nat. Hist., Trans., 10: 25-60.

Nafpaktitis, Basil G. 1968. Taxonomy and distribution of the lanternfishes, gen-
era Lobianchia and Diaphus, in the north Atlantic. In Carlsberg Fdn. Oceanogr.
Exped., 1928-30, Dana-Rept. 73, 131 p.

Nafpaktitis, Basil G., and John R. Paxton. 1968. Review of the lanternfish genus
Lampadena with a description of a new species. Los Angeles County Mus. Nat.
Hist., Contrib. in Sci., 138: 1-29.

Obradovich, J. D. 1965. Isotopic ages related to Pleistocene events. INQUA VII
Int. Congr. Abstr., p. 364.

Paxton, John R. 1967. A distributional analysis for the lanternfishes (Family
Myctophidae) of the San Pedro Basin, California. Copeia, 1967: 422-440.

Weiler, Wilhelm. 1959. Miozane Fisch-Otolithen aus der Bohrung S. Pablo 2 im
Becken von Veracruz in Mexiko. N. Jb. Geol. Palaont., Abh., 109: 147-172.

. 1968. Otolithi Piscium (Neubearbeitung) . In: Fossilium Catalogus I. Ani-

malia, F. Westphal, edit., 117: 1-196.

Accepted for publication June 17, 1969.

LOS

ANGELES

COUNTY

MUSEUM

CONTRIBUTIONS
fll'Zt IN SCIENCE

Number 174

June 30, 1969

RESULTS OF THE 1968 AVIL EXPEDITION TO

MT. NYIRU, SAMBURU DISTRICT, KENYA.

ORNITHOLOGY

By

Herbert Friedmann and Kenneth E. Stager

Los Angeles County Museum of Natural History • Exposition Park
Los Angeles, California 90007

CONTRIBUTIONS IN SCIENCE is a series of miscellaneous technical papers
in the fields of Biology, Geology and Anthropology, published at irregular intervals
by the Los Angeles County Museum of Natural History. 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 Robert J. Lavenberg, Los Angeles County Museum of Natural
History, 900 Exposition Boulevard, Los Angeles, California 90007.

Robert J. Lavenberg
Managing Editor

INSTRUCTIONS FOR AUTHORS

Manuscripts for the LOS ANGELES COUNTY MUSEUM, CONTRIBU-
TIONS 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. Al-
though priority will be given to manuscripts by staff members, or to papers dealing
largely with specimens in the Museum’s collections, 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 Edi-
torial Committee. They may also be subject to review by competent specialists out-
side 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.

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. Alterations or changes in the manuscript after galley proof will
be billed to the author. Unless specifically requested, page proof will not 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.

Dorothy M. Halmos
Editor

RESULTS OF THE 1968 AVIL EXPEDITION TO
MT. NYIRU, SAMBURU DISTRICT, KENYA.
ORNITHOLOGY

By

Herbert Friedmann1 and Kenneth E. Stager2

Abstract: The 1968 Avil Expedition was organized, spon-
sored, and led by Mr. and Mrs. Gordon Avil, to make the first
comprehensive survey of the bird life of the montane forest on
the upper stretches of Mt. Nyiru, northern Kenya. Some 300
specimens of 80 species were obtained, of which 29 species were
taken in, or at the edge of, the montane forest, the rest being
from the lower slopes of the mountain. Previously less than a
dozen kinds of birds had been recorded from Mt. Nyiru, and
most of these were not forest species from the top of that moun-
tain. The present data on the Mt. Nyiru avifauna are compared
with those of Kulal and Marsabit. All specimens collected by the
Avil Expedition are in the Los Angeles County Museum of
Natural History.

The 1968 Avil Expedition of the Los Angeles County Museum of Natural
History had as its goal the ornithological exploration of Mt. Nyiru, Samburu
District, Northern Frontier Division, Kenya. The extensive montane forests of
this isolated mountain range had not been investigated systematically prior to
our work there in 1968. The 1966 Cheney Expedition (Friedmann and Stager,
1967) worked at the west base of the mountain but did not have the time or
opportunity to attempt a faunal survey of the extensive forest on the upper
portions of Mt. Nyiru. The desirability of such a survey was obvious, but its
realization had to be postponed for another trip.

Mr. and Mrs. Gordon Avil learned of this desire and in 1968 they gen-
erously sponsored and accompanied the expedition that resulted in the collec-
tions reported on in this paper.

The expedition departed from Nanyuki, Kenya, on June 1, 1968, and
traveled northward into the Samburu District, Northern Frontier Division, by
way of Rumuruti, Maralal and Baragoi. Base camp was established June 2 at
Turn (5240 feet [1588 m.] elevation), at the west base of Mt. Nyiru (Fig. 1).
The camp site was the same one occupied by the 1966 Cheney Expedition
(Friedmann and Stager, 1967). Collecting in the Turn area began on June 3
and continued until June 8. During this period an unused trail up the west face
of Mt. Nyiru from Turn was cleared by the Samburu, to enable our donkey
pack train of 30 animals to move our camp and supplies to the mountain for-
ests atop Mt. Nyiru. The move to the top was accomplished on June 8 and 9,

director, Los Angeles County Museum of Natural History

2Senior Curator of Ornithology, Los Angeles County Muceum of Natural History

1

2

Contributions in Science

No. 174

with collecting in the mountain forests atop Nyiru beginning on June 10. The
mountain top camp was situated in the forest at the edge of a grassy glade
approximately one mile east of the west rim of Mt. Nyiru and at an elevation
of 8400 feet (2545 m.). Collecting continued through June 21. On June 22,
the expedition returned to the base camp area at Turn where collecting opera-
tions continued until June 27. The return to Nanyuki began on June 28 and
Nairobi was reached on July 1, 1968.

Expedition personnel consisted of Mr. and Mrs. Gordon Avil, sponsors;
Kenneth E. Stager, ornithologist; Peter Saw, professional hunter (Monty
Brown Safaris), and his safari staff of thirteen men; Julius Ithia and Julius
Kyongo, professional bird skinners trained by John G. Williams.

The Museum is glad to make grateful acknowledgment to the generous

Figure 1. Map of northern Kenya area, showing area visited by the Avil expedition.

1969

Birds of Avil Expedition to Kenya

3

sponsors of the expedition, Mr. and Mrs. Gordon Avil of Beverly Hills, Cali-
fornia. Not only did Mr. and Mrs. Avil make the trip possible through finan-
cial support and actual participation in the field work of the expedition, but
also generously provided the funds necessary for the publication of this report
on the work accomplished.

Thanks are also due Chief Game Warden of Kenya, Mr. Nicholas Nganga
and his assistant, Mr. Stan Bleazard, for the granting of official permits neces-
sary for the accomplishment of our study. Special thanks are due Major Rod-
ney T. Elliott, Divisional Game Warden, Northern Frontier Division, Maralal,
for his hospitality and assistance to the expedition while it was working in his
administrative area. For informing us of some unpublished observations on
Kulal and Marsabit birds we express our thanks to G. Stuart Keith and John
G. Williams. During the course of the study of the specimens, we received kind
cooperation and assistance from M. A. Traylor of the Field Museum of Nat-
ural History, and R. Zusi of the National Museum of Natural History.

The senior author is responsible for the general discussion of Nyiru and
other north Kenyan montane forest avifaunas, and for the identifications and
systematic notes. The junior author made the collection, supplied the field notes
and the photographs, and the description of the area. Museum artist Mary V.
Butler made the illustration (Fig. 8) of Nectarinia reichenowi.

In the semi-arid stretches of northern Kenya, east of the Rift Valley, from
about the second parallel of north latitude, north to the Abyssinian border,
there are three isolated mountains that are crowned with extensive montane
forests — Nyiru, Kulal, and Marsabit. Mt. Nyiru is the southernmost of the
three, 2° ll'N, 36° 49' E, and rises from a base plain of about 5000 feet (1515
m.) to an elevation of a little over 9000 feet (2727 m.) . Kulal is about 40 miles
(667 km.) to the north and very slightly to the east, 2° 44' N, 36° 56' E, and
is not nearly as high, its summit being about 7730 feet (2342 m.). Marsabit,
the most isolated of the three, lies at 2° 16' N, 37° 57' E, approximately 80
miles (134 km.) to the east of both. It is the lowest of the three; its greatest
elevation is a little under 4700 feet (1424 m.) and its base plain is lower than
that of Nyiru, about 2200 feet (667 m.).

Geographically the closest montane forest to the south of Nyiru is the
Matthews Range, about 50 or 60 miles (84-100 km.) away. A much broader
gap of at least 200 miles (334 km.) separates these north Kenyan mountains
from the southernmost of the forested highlands of Ethiopia, to the north. Mt.
Nyiru is an old and isolated mountain quite as separated from the Matthews
Range to the south as it is from Kulal to the north and from Marsabit to the
east. So far as is known, its past history is probably similar to that of these
other two mountains. Moreau ( 1966: 191-192) has stated that a small number
of montane forests on isolated mountain tops today were probably isolated as
well during past glacial periods, but we do not know if he had any data on
Nyiru to cause him to ignore it when he concluded that of these areas, only two
— Kulal and Marsabit — were isolated bv as much as 30 miles (50 km.) of

4

Contributions in Science

No. 174

arid, low veldt, in which the forest dwelling montane species could not live.
The isolation of Nyiru is just as great, and the altitude to which it rises from
the base plain is even higher. It is clearly to be considered along with Kulal
and Marsabit. The only explanation we can suggest for Moreau’s apparent ig-
noring of Nyiru is that he may have considered it not as the isolated mountain
that it is, but as the northern extremity of the Matthews Range.

The bird life inhabiting these montane forests is mostly quite distinct from
that of the lowland forests on the lower slopes of these mountains, or of the
intervening, generally non-forested, semi-arid country, although there are some
birds that inhabit both. However, it does seem proper to look upon these high-
land sylvan avifaunas as isolated ecological entities. At the same time, it would
seem that their component species must be colonizers that reached each moun-
tain by individual flights, and not as a result of any mass alterations of the
topography or ecology of the entire area. They are not relict patches of a once
more extensive continuum of montane forest fauna. The birds may still recol-
onize these areas by such flights from time to time. The avifauna of each of
these three mountain forests is depauperate, and with little racial endemism,
as might be expected if recolonizing were taking place with any frequency.

Until now the avifauna of Marsabit was the best documented of the three,
with numerous records of individual species listed by van Someren (1922,
1932), Sharpe (1930), Jackson (1938), and Moreau (1966: 194-195).
Moreau very properly differentiated the montane forest records from the
large number of species listed by the others as from “Marsabit” and this re-
duced the number he considered pertinent to only seven species, all but one of
which ( Onychognathus walleri ) are represented in our present Mt. Nyiru col-
lection. Moreau’s list is not quite complete, as he states that the dove Aplopelia
larvata does not occur on Marsabit; but it is so recorded by Sharpe (1930) and
Jackson (1938). While Moreau does not include Tauraco hartlaubi in his
Marsabit list, he does mention it as having been alluded to rather than pre-
cisely recorded from there by van Someren. In his unpublished data on the
birds of Marsabit, Stuart Keith did not include this bird from the Marsabit
uplands, although he did note it on Kulal. This tends to corroborate Moreau’s
skepticism of the occurrence of this conspicuous bird in the montane forest on
Marsabit.

The avifauna of the Kulal montane forest, said to cover some 40 square
miles, is very imperfectly known, the only data in print being the information
assembled by Moreau (1966: 195) from unpublished specimen and sight rec-
ords of a small number of species supplied him by J. G. Williams and G. S.
Keith. Moreau does not state how many species have been found there, but
lists only seven that, to his current knowledge, do not occur on Marsabit. Of
these seven, all but two ( Aplopelia larvata and Andropadus tephrolaema ) are
represented in our present collection from Mt. Nyiru.

Prior to the collection reported on in the present paper, the bird life of
Mt. Nyiru was known from published records of some 11 species listed by

1969

Birds of Avil Expedition to Kenya

5

Jackson (1938), and some of these, such as Mirafra cantillans marginata,
Phylloscopus trochilus trochilus, Cisticola juncidis perennia, and Spinus cit-
rinelloides kikuyuensis are not forest birds. Of the remaining seven, Jackson
gives no indication as to whether they were taken in the true montane forest
at the top of the mountain or in the wooded areas around its base. However,
all seven are also included in our present collection and it may be assumed
that Jackson’s records came from similar habitats as our 1968 examples.

As is so often the case, there are in the literature references to specimens
pertinent to our present study, hidden in the discussion of relatively irrelevant
populations. As an example, van Someren (1939: 76), in discussing the birds
of the Chyulu Mountains in southern Kenya, mentions that Turdus olivaceus
polius ( = T. abyssinicus abyssinicus ) occurs at high elevations on Marsabit,
Kulal, and Nyiru, but fails to list either his specimens or his sources. We have
tried to compile such casual references, but cannot be certain that we found
them all.

The present collection contains 300 specimens of 80 species. Of these 80
kinds of birds only 29 were taken on the top of the mountain, in, or at the
edge of, the montane forest, and these are the birds of particular interest, as
they give us the information with which meaningful comparisons may be made
with the corresponding montane avifauna on Kulal and on Marsabit. Even of
these 29, five ( Indicator variegatus, Cossypha caffra iolaema, Corvus rhipidu-
rus, Corvus albicollis, and Laniarius ferrugineus ambiguus ) are not peculiar
to montane forests, but have a wide ecological range.

The other 5 1 species were taken chiefly on or near the lower slopes of Mt.
Nyiru, and all but a very few of them had previously been collected by Stager
in the same area in 1966 while on the Cheney Expedition to the Samburu Dis-
tirct. These species call for little comment, beyond listing the specimens with
localities and dates, plus Stager’s field observations. Users of this paper might
therefore be advised to consider this part of it as a supplement to the one by
the present authors (1967) on the Cheney Expedition. To expedite this usage,
the nomenclature of the present report has been kept uniform with that of the
earlier one, and, in the few instances where change has seemed called for, this
is explained in the accounts of the species involved.

If we combine the Kulal and Marsabit montane forest records available
in print with those kindly supplied by John Williams and Stuart Keith, and
correlate with them the Mt. Nyiru data from the present collection, we may
indicate the state of present knowledge of the similarities of the avifaunas of
these three highland forests in the following table. It should be kept in mind
that the apparent relative richness of that of Mt. Nyiru may be due to the fact
that we now have a good collection from that mountain, while equally com-
prehensive samplings are still to be made on the other two. If anything, the
higher altitude of Mt. Nyiru would lead one to suspect that it had a more re-
stricted, rather than a richer, fauna than Kulal or Marsabit. This thought can-

6

Contributions in Science

No. 174

not help but qualify the appropriateness of extracting significance from the
present data.

A few points may, however, be mentioned. The absence of records of
brightly colored, easily recognized species such as Nectarinia reichenowi,
Nectarinia famosa, and Cryptospiza Salvador ii from Kulal and Marsabit and
their abundance on Nyiru does suggest distributional divergence in each case.
As indicated in the text of this report, a few instances of racial endemism do
occur: Muscicapa adusta is represented by murina on Nyiru and by marsabit
on Marsabit; Zoster ops senegalensis by jacksoni on Nyiru and Marsabit, by
kulalensis on Kulal. If, as seems likely, Nectarinia reichenowi does not occur
on Kulal and Marsabit, the race lathburyi is endemic to Nyiru, and south to
the Matthews Range.

Table 1

Records of Occurrence of Montane Birds

Montane Forest Species

Nyiru Kulal Marsabit

Francolinus squamatus schuetti x

A polopelia larvata larvata

Tauraco hartlaubi x

Pogoniulus leucomystax x

Indicator variegatus x

Alcippe abyssinica abyssinica x

Pycnonotus latirostris saturatus x

Phyllastrephus fischeri placidus x

Muscicapa adusta murina x

Muscicapa adusta marsabit

Melaenornis chocolatina fischeri x

Cossypha caffra iolaema x

Pogonocichla stellata guttifer x

Turdus abyssinicus abyssinicus x

Zoothera piaggiae piaggiae x

Bradypterus cinnamomeus cinnamomeus x

Phylloscopus umbrovirens mackenzianus x

Chloropeta similis x

Coracina caesia pura x

Laniarius ferrugineus ambiguus x

Corvus rhipidurus x

Corvus albicollis x

Onychognathus walleri

Zosterops senegalensis jacksoni x

Zoster ops senegalensis kulalensis
Nectarinia mediocris mediocris x

Nectarinia tacazze jacksoni x

Nectarinia famosa aeneigularis x

Nectarinia reichenowi lathburyi x

Cryptospiza salvadorii salvadorii x

Fstrilda melanotis kilimensis x

Serinus dorsostriatus maculicollis x

Serinus striolatus striolatus x

x x

x

x

x

x

x x

x

x x

x

x

X

X

X

1969

Birds of Avil Expedition to Kenya

7

Figure 2. View of the west face of Mt. Nyiru from the vicinity of the expedition
base camp at Turn, June 1968.

Figure 3. Forest cover on top of Mt. Nyiru at 8600 ft. elevation, June 1968.

8

Contributions in Science

No. 174

Figure 4. Interior of forest on top of Mt. Nyiru at 8600 ft. elevation, June 1968.

The following description of Mt. Nyiru was written by Stager from his
personal experience and observations on the Avil Expedition. From the acacia
bush and grass covered plains at its west base (4000 feet [1212 m.] elevation),
the Mt. Nyiru massif rises abruptly to an elevation of slightly over 9000 feet
(2727 m.). The steep slopes of the mountain are clothed in heavy vegetation
and are well watered (Fig. 2). From the base camp at Turn, numerous water-

1969

Birds of Avil Expedition to Kenya

9

Figure 5. Interior of forest on top of Mt. Nyiru at 8400 ft. elevation, June 1968.

falls were visible cascading down deep, fissurelike canyons, filled with palms
{Phoenix sp.). The vegetation at Turn consists largely of flat-topped acacias.
Dombeya, Commiphora, Protea, Terminalia, Kigelia and Ficus. The slopes
above Turn have large stands of cedar ( Juniperus procera ), while the more
open hillsides support numerous large Euphorbias.

The montane forest atop Mt. Nyiru is extensive (approximately 60 square
miles) and magnificent. The summit of Nyiru consists of undulating hills and
valleys (Fig. 3), heavily forested on the slopes of the hills, with luxuriant
grassy glades or loriani in the valleys. Mt. Nyiru is covered with a primary
forest untouched by axe and fired only by occasional lightning strikes. The
dominant tree forms are podo ( Podocarpus milijianus ); musharagi ( Olea
hochstetteri) ; muzura (O cotea kenyensis)\ mueri ( Pygeum africanum) ; mun-
yenye {Suregada procera) ; rapanea ( Rapanea rhrododenoides) ; and cedar
( Juniperus procera ). The forest floor is open (Figs. 4 and 5) except at the
periphery of the grassy glades where undergrowth is encountered (Fig. 6).
Epiphytes abound in the form of ferns and mosses of many species, and one
dominant orchid, an Angraecum (sp.) with beautiful waxy white blossoms.
The tops of some of the hills are open grassy meadows (Fig. 7) with scattered
islands of bracken fern ( Pteridium aquilinium) . Mountain bamboo {Arundi-

10

Contributions in Science

No. 174

Figure 6. Heavy plant growth at periphery of glade, cleared for placement of mist
nets. Mt. Nyiru, 8600 ft. elevation, June 1968.

naria alpina ) occurs in many areas and large patches of it were observable on
the higher hillsides on the eastern portion of the range. Giant heath ( Erica
arborea ) occurs on Mowongo Sowan, the highest point of Mt. Nyiru, and
possibly elsewhere on the mountain. Cloud or mist forest condition prevails

1969

Birds of Avil Expedition to Kenya

11

Figure 7. Open park land showing islands of bracken fern and Leonotis, top of Mt.
Nyiru, 8600 ft. elevation, June 1968.

atop Nyiru, as attested by the extensive drapery of the trees with the lichen
(Usnea sp.) known as “old man moss” (see Fig. 3).

Mt. Nyiru is revered by the native Samburu as the home and birthplace
of the tribe. It is grazed by the cattle herds of the Samburu at certain times of
the year, but grazing rights are wholly restricted to the El Masula, the senior
clan of the Samburu. The Samburu using the forests and glades of Nyiru
practice an excellent form of forest conservation that is difficult to improve
upon. Each grassy glade is the property of a sub-section of the El Masula clan,
bears a particular name, and is never overgrazed. By custom, fires are taboo
within the forest, as is the cutting of trees. Firewood can only be obtained from
dead trees and fires are restricted to cleared areas in open ground. Manyattas
(family villages) are small, moved frequently and never burned for tribal
rights, as they are in the low country, at births and deaths.

The dominant mammals of the montane forest are elephant, buffalo,
leopard, bush buck ( Tragelaphus scriptus ), giant forest hog ( Hylochoerus
meinertzhageni) , olive baboon ( Papio anubis) , and hunting dog ( Lycaon
pictus ).

12

Contributions in Science

No. 174

Annotated List of Specimens
Family Accipitridae

Melierax poliopterus Cabanis

The pale chanting goshawk was met with at the west base of Mt. Nyiru, at
5240 feet, as it was in 1966 by the Cheney Expedition. Two males were col-
lected on June 3 and 26, the earlier one an immature bird in molting plumage
and in non-breeding condition, the later one an adult with enlarged testes. The
second specimen weighed 600 grams; their stomachs contained lizards, a snake,
small rodent remains, and insects.

Although no specimens were collected, the augur buzzard ( Buteo rufofus-
cus) was also observed atop Mt. Nyiru on several occasions between June 9
and June 22, 1968.

Milvus migrans parasitus (Daudin)

One specimen of the yellow-billed kite was taken at the west base of Mt.
Nyiru, 5240 feet, on June 26; a male with slightly enlarged gonads; weight 675
grams. Kites were common about the base camp at the west base of Mt. Nyiru
and became a pest to the safari staff attempting to hang out strips of fresh meat
to dry.

Several species of vultures were common on the plains at the west base
of Nyiru, but the only species recorded, though not collected, on the top of the
mountain was the hooded vulture ( Necrosyrtes monachus) . As many as six
individuals of this species were observed at one time about our camp at 8400
feet elevation.

Family Phasianidae
Francolinus sephaena grantii Hartlaub

This common francolin is represented by two specimens, one of each sex,
taken June 3 and 4, at the west base of Mt. Nyiru; both had enlarged gonads;
the male weighed 290 grams, the female 265; stomach contents seeds and
insects. This species was also taken by the Cheney Expedition in 1966, but
some distance to the south, in the Karissia Hills, and south of Maralal.

Francolinus squamatus schuetti Cabanis

Three examples of this francolin were taken near the top of Mt. Nyiru,
8400 to 8600 feet, June 12 to 16. Two of them were males with enlarged go-
nads, the other a female with no ovarian swelling. The female weighed 300
grams, the males 450 and 475, respectively; stomach contents seeds and in-
sects. Francolins of this species were commonly heard calling in the forest atop
Mt. Nyiru and were frequently flushed in small flocks of three to four birds.

1969

Birds of Avil Expedition to Kenya

13

Family Columbidae

Streptopelia senegalensis senegalensis (Linnaeus)

One male, testes enlarged, weight 60 grams, was taken on the west base of
Mt. Nyiru, 5240 feet, on June 8. A similar one, also with large gonads, was
collected in the same area in June 1966 by the Cheney Expedition. Pigeons
and doves were noticeably absent in the mountain forests of Mt. Nyiru. On
two occasions a dove was heard and seen in the deep forest. It resembled Aplo-
pelia larvata in sound and appearance, but no specimens were secured.

Turtur chalcospilos (Wagler)

As was the experience of the Cheney Expedition, the emerald spotted
dove was met with again at the base of Mt. Nyiru, where two specimens, one
of each sex, were taken June 6 and 26, both with slight gonadal enlargement;
weight, female 60 grams, male 55; stomach contents seeds and insects.

Family Musophagidae
Tauraco hartlaubi (Fischer and Reichenow)

Hartlaub’s turaco has a considerable elevational range, as six specimens
were taken at 8400 feet on Mt. Nyiru, June 11 to 13, while two years earlier
the Cheney Expedition collected a similar series in the Karissia Hills, 6500
feet, and heard, but did not obtain, others on the lower slopes of Mt. Nyiru.
All the birds from the top of Nyiru had enlarged gonads, and ranged in weight
from 210 to 260 grams; stomach contents berries, small fruits, and seeds.

Turacos of this species are extremely common throughout the forested
top of Mt. Nyiru. During daylight hours there seemed to be no time that the
calls of this bird were not ringing loudly through the forest.

Corythaixoides leucogaster (Riippell)

One female, with slightly enlarged ovary, was taken at the west base of
Mt. Nyiru, 5240 feet, June 4; weight 225 grams; stomach contents large red
berries. The species had also been taken there by the Cheney Expedition in
1966. A common species in the Acacia, Dombeya, and Commiphora thickets
around the base camp area.

Family Cuculidae
Cuculus solitarius solitarius Stephens
The red-chested cuckoo was not met with on the higher stretches of Mt.
Nyiru but two males were obtained at the west base of the mountain, 5240
feet, June 4 and 27; both with slightly enlarged testes; both weighing 60 grams;
stomach contents ants and other insects.

14

Contributions in Science

No. 174

Centropus superciliosus superciliosus Hemprich and Ehrenberg
The white-browed coucal was met with at the west base of Mt. Nyiru,
5240 feet, on June 3, when one male, with enlarged testes, was collected;
weight 130 grams.

Family Strigidae
Otus scops senegalensis (Swainson)

Three specimens of this little owl were obtained at the west base of Mt.
Nyiru, June 7, two males with slightly enlarged testes, and one female with a
small ovary; weights 60 to 65 grams.

White (1965: 193) considers caecus and graueri as synonyms of sene-
galensis, and we follow him in calling our present birds by the last name. They
are paler, less finely, less closely, and less darkly vermiculated than is one from
Isiola, which, although coming from a locality far to the south of Nyiru, re-
sembles the dusky, closely vermiculated Abyssinian population formerly called
caecus. This suggests that White is probably correct in relegating all these birds
to nominate senegalensis.

Two of the present specimens, but not the third, are somewhat larger
(wings 109 mm. in the smaller male, 122 and 123 mm. in the other two birds),
and all are somewhat darker, with less unmarked white in their plumage, than
two others from Bura, in the lower Tana River valley, both of which have
wings 1 15 mm. in length. These Bura examples belong to a new race Twomey
and Keith have recently described as O. s. nivosus (Ibis, 1968: 538).

The two Mt. Nyiru males are slightly darker on the back and wings than
the female. The small male, which may be subadult, has the blackish shaft
streaks on the underparts less well marked than in the larger male or the
female and has the gray vermiculations on the ventral feathers less spaced with
white, and paler gray. Our three birds had remains of beetles, orthoptera, and
other insect larvae in their stomachs.

All three of the above specimens were heard calling at night in the imme-
diate vicinity of the base camp and were easily located by eyeshining with an
electric torch.

Family Caprimulgidae
Caprimulgus stellatus simplex Neumann

Two examples of this little-known nightjar were collected in the plains
area near the base of Mt. Nyiru, 5000 feet, June 26; one of each sex, both in
non-breeding state; weight 60 grams each; stomach contents insects. The male
is in the rufescent plumage phase, the female is gray.

We allocate these specimens to the race simplex on geographic grounds,
as we have had no comparative material available, and follow White (1965:
206), who considers the birds of central Ethiopia and southeastern Sudan,
south to northwestern and central Kenya as of this subspecies.

1969

Birds of Avil Expedition to Kenya

15

Family Coliidae

Colius macrourus pulcher Neumann

Two specimens, one female, one unsexed, were taken on the west base
of Mt. Nyiru, 5240 feet, June 3 and 4; weight 35 and 45 grams. The Cheney
Expedition had obtained this mousebird in the same area two years earlier.

Family Coraciidae
Coracias naevia naevia Daudin

One female rufous-crowned roller, with slight ovarian enlargement, was
taken on June 26, on the west base of Mt. Nyiru, 5100 feet; weight 140 grams.

Family Bucerotidae
Tockus flavirostris fiavirostris (Riippell)

The yellow-billed hornbill was found on the west base of Mt. Nyiru, 5240
feet, June 25, when one male with enlarged gonads was taken; weight 275
grams. Two years earlier the Cheney Expedition also collected this species
there.

A young fledgling red-billed hornbill, Tockus erythrorhynchus, was pur-
chased alive from a Samburu warrior on June 26 at the base camp and retained
as a pet. The bird was fully fledged including the tail feathers. It was left, alive,
in Kenya.

Hornbills were carefully watched for in the forests atop Mt. Nyiru. The
only bird sighted was a single large hornbill, possibly a Bycanistes in flight over
the forest top at 8600 feet elevation, June 16, 1968.

Family Capitonidae
Lybius leucomelas diadematus (Heuglin)

One male, testes slightly enlarged, was taken on the west base of Mt.
Nyiru, 5240 feet, June 4; weight 30 grams. This species had also been collected
there by the Cheney Expedition.

Lybius lacrymosus lacrymosus (Cabanis)

The Avil Expedition, like the Cheney one two years earlier, found the
spotted-flanked barbet to be common on the west slope of Mt. Nyiru, 5240
feet, where, on June 4 and 7, they collected one male and two females, all in
non-breeding state; weight 25 grams each.

Pogoniulus leucomystax (Sharpe)

One male, testes slightly enlarged, was taken in the montane forest on Mt.
Nyiru, 8400 feet, June 12; weight 5 grams. This locality is considerably farther
north than the Karissia Hills, where the Cheney Expedition collected this
tinker-bird and which at that time was considered to be near the northern

16

Contributions in Science

No. 174

limit of its range. The above specimen was the only barbet observed or heard
on top of Mt. Nyiru.

Pogoniulus pusillus affinis (Reichenow)

On the west base of Mt. Nyiru, 5240 feet, one adult of each sex of the
red-fronted tinker-bird were collected, June 4 and 26; the female weighed 5,
the male 10 grams. These specimens agree with others from the Karissia Hills
(Cheney Expedition).

Trachyphonus darnaudii bohmi Fischer and Reichenow
Two males, testes slightly enlarged, collected on the west base of Mt.
Nyiru, 5240 feet, June 6 and 26; weight 30 grams each; agree with others taken
there earlier by the Cheney Expedition.

Family Indicatoridae
Indicator variegatus Lesson

One male scaly-throated honey-guide, testes slightly enlarged, was taken
on the top of Mt. Nyiru, 8400 feet, on June 13. This bird is not a montane
forest bird exclusively, but ranges down to sea level and up to 1 1,000 feet (on
Mt. Elgon). The specimen weighed 50 grams.

The greater honey-guide ( Indicator indicator ) was frequently observed
near the montane forest atop Mt. Nyiru, but no specimens were collected.
This species is even less of a forest bird than /. variegatus.

Family Picidae

Dendropicos fuscescens hemprichii (Ehrenberg)

The cardinal woodpecker is represented by one adult male, testes slightly
enlarged, taken on the west base of Mt. Nyiru, 5240 feet, June 3; weight 25
grams. It agrees with the examples obtained in the same place by the Cheney
Expedition, which were found (Friedmann and Stager, 1967: 21) to be
hemprichii and not lepidus, although the two races probably meet near Mt.
Nyiru.

Woodpeckers appeared to be totally absent from the forested top of Mt.
Nyiru.

Family Timaliidae
Alcippe abyssinica abyssinica (Riippell)

This little hill babbler was found to be a common bird in the montane
forest on Mt. Nyiru, 8400 to 8600 feet. A series of 14 specimens, seven of each
sex, was collected between June 10 and June 21; all the males and two of the
females had enlarged gonads, the other five females had small or only slightly
swollen ovaries. Their total weights ranged as follows: 11 weighed 20 grams
apiece, 2 weighed 15, and 1 weighed 25 grams.

1969

Birds of Avil Expedition to Kenya

17

These birds have been compared with a series kindly loaned by the Field
Museum of Natural History. They average slightly less rufescent on the back,
and, to a lesser degree, on the sides and the flanks, than examples from the
Kenya highlands farther south (Mt. Kenya, Kericho, Mau), but agree very
closely with one from Gojam, Ethiopia. Van Someren (1922: 245), on the
other hand, found that his Kenya birds (Elgon, Maraquet, Elgeyu, Molo,
Londiani, and Mt. Kenya) were less rufous on the back than Abyssinian
examples! All east African birds, from Ethiopia to northern Tanzania, are
currently considered as nominate abyssinica. We concur in so allocating our
Nyiru birds, which are similar to the Ethiopian specimen, the most nearly
topotypical abyssinica we have been able to examine. The slight discrepancy
between Nyiru and more southern Kenyan birds is here recorded for the ben-
efit of those who may have to deal with examples from Mt. Kenya south to
the Usambaras.

Further study with extensive material may yet reveal evidence in support
of van Someren’s chyulu, as his measurements (1939: 60) for that race not
only are smaller than those of his more northern Kenya highland birds, but
are smaller than those examined by us. The Nyiru males average slightly
larger than birds from the Kenya highlands; wings 69 to 75 (70.85) mm. as
compared with 66 to 70 (68) mm.; tail 63 to 66 (62.85) mm. as against 59 to
62 (61) mm.; culmen from distal end of nostril 7.4 to 8.0 (7.6) mm. as against
6.9 to 7.6 (7.2) mm. Strangely enough, Nyiru females do not reveal this size
difference.

The map given in Moreau (1966: 193) for this species shows no locali-
ties of record between Kapenguria, near Mt. Elgon, and ca. 7° N latitude in
Ethiopia. Mt. Nyiru thus appears to be the first station known for the species
in this vast area. The species has not yet been recorded from Kulal, Marsabit,
or the Matthews Range, but this does not mean that it is definitely absent in
those localities.

All the present specimens had insect remains in the stomachs.

Family Pycnonotidae
Pycnonotus barbatus dodsoni Sharpe.

A series of seven specimens of this common bulbul from the west base
of Mt. Nyiru, 5240 feet, June 4 to 26, are definitely dodsoni, not fayi (or tri-
color), as are the birds taken in the Karissia Hills two years earlier by the
Cheney Expedition. Although these races meet in the Samburu district, there-
is no sign of intergradation in either of these two population samples. The
Nyiru birds show the distinct pectoral mottling, and the white tips to the tail
feathers characteristic of dodsoni. Of the seven examples, five are males with
testes ranging from slightly to much enlarged; the two females had slight ova-
rian enlargement in one, none in the other. The birds weighed 20 to 35 grams.

A common bulbul along the stream courses of the base camp area, where
they were busily feeding on the fruiting Ficus.

18

Contributions in Science

No. 174

Pycnonotus latirostris saturatus (Mearns)

This bulbul was found to be very numerous in the montane forest, 8400
feet, on Mt. Nyiru, and 12 males and three females were collected there, June
10 to 21. Most of the males had enlarged testes, three only slightly so, but the
females showed little or no gonadal swelling.

The subspecies saturatus is similar to eugenius of western Uganda but is
a little more greenish on the top of the head and on the underparts, and has
slightly longer tail measurements.

Phyllastrephus fischeri placidus (Shelley)

Two specimens of this bulbul were taken in the montane forest, 8400 feet,
on Mt. Nyiru, June 13 and 14. Both were males with enlarged testes, and each
weighed 30 grams. They agree with good series from Mt. Kenya, Nyeri, and
other highland areas of Kenya. We have seen no material from Marsabit, and
cannot pass judgment on the race marsabit. However, the fact that van Som-
eren described it (1932: 343) from as low as 2000 feet on Mt. Marsabit,
precludes its being considered a true montane form there, and this suggests
that it is probably not a local endemism, but is the same as placidus, as Rand
(1958: 207) also considered it to be. In most areas, however, placidus is a
bird of higher altitudes.

Phyllastrephus strepitans Reichenow

On the west base of Mt. Nyiru, 5240 feet, on June 4, one male and one
female of this species were obtained, the male with slight gonadal development,
the female with none. At first glance these examples seem quite distinct from
other Kenyan ones, but, as pointed out in an earlier study (Friedmann, 1937:
1 16) and later confirmed by Rand (1958: 199), individual variation, plus the
marked effects of wear and fading, make it impossible to recognize races
within its range.

Our two birds weighed 20 and 25 grams, respectively.

Family Muscicapidae

Muscicapa adusta murina (Fischer and Reichenow)

This flycatcher was found to be common in the montane forest on Mt.
Nyiru, where one male and four females, all in non-breeding condition, were
collected at altitudes of from 8400 to 8600 feet, June 14 to 24.

We are advised by M. A. Traylor that in the forthcoming account of the
Muscicapidae in the continuation of Peters’ Check-list, the race interposita is
to be included in murina, thus differing from the conclusions of White (1963:
8), who recognized interposita. This leaves us with two races to consider. The
subspecies marsabit, described from the mountain of that name, is said to occur
on the higher parts of the mountains of northern Kenya from Moyale and
Marsabit to Laikipia, Kapenguria, Mt. Elgon, and eastern Uganda, intergrad-

S7 —

/ * t

(9

"T

x>

1969

Birds of Avil Expedition to Kenya

19

ing extensively with murina in the central Kenya highlands. On geographic
grounds we would expect the Mt. Nyiru birds to be marsabit , but they agree
with examples of inter posita (now murina ) from the Karissia Hills, not with
marsabit, supposedly somewhat rusty colored. They are also less richly colored
than one bird from Molo (topotypical interposita) .

Van Someren ( 1932: 259) reported that marsabit is smaller than murina ,
his 13 examples from Mt. Marsabit having wing lengths of 56 to 63 mm.,
mostly 56 to 57 mm. Our Nyiru birds measure 60 to 64 mm., and thus average
larger than van Someren’s topotypes of marsabit.

While marsabit is said to intergrade extensively with murina in the central
Kenya highlands, it seems to us that the integration begins as far north as
Nyiru and this, together with the wing lengths of our series, causes us to prefer
to consider the “intergrades” north to Nyiru as more properly referable to
murina. Otherwise we would be inclined to question the distinctness of marsa-
bit. This conclusion leads to the implication that this little flycatcher is one
of the very few species of montane forest birds in which the Marsabit and
Nyiru populations are racially distinct. We have not seen any material from
Laikipia, Kapenguria, or Elgon.

Melaenornis chocolatina fischeri (Reichenow)

The white-eyed slaty flycatcher was found to be numerous in the montane
forest on Mt. Nyiru, 8400 feet, where three males and one female were col-
lected June 10 to 15. One of the males showed gonadal enlargement; the other
birds did not. They weighed 20 to 25 grams. This species is part of the montane
forest avifauna but is not restricted to it, and occurs also at lower elevations,
as in the Karissia Hills, where the Cheney Expedition collected several exam-
ples.

Melaenornis pammelaina (Stanley)

Two males, testes slightly enlarged, were taken on the west base of Mt.
Nyiru, 5240 feet, June 6. The species had been found there by the Cheney Ex-
pedition as well, and apparently is well established there.

Batis molitor (Hahn and Kuster)

This little chin-spot flycatcher was found at 5240 feet, on the west base
of Mt. Nyiru, where two males, testes slightly enlarged, were taken June 3
and 4.

Terpsiphone viridis ferreti (Guerin)

Two specimens of the paradise flycatcher were collected at 5240 feet on
the west base of Mt. Nyiru, June 4 and 24; neither (one of each sex) was in
breeding condition; the male weighed 10, the female 15 grams.

20

Contributions in Science

No. 174

Family Turdidae

Cossypha caffra iolaema Reichenow

The Kenya robin-chat was found both at the west base of Mt. Nyiru, 5240
feet, and near the montane forest at its top, 8400 to 8600 feet. Specimens from
the two areas are alike. All in all, some seven examples were obtained, five
males and two females, all with small or only slightly enlarged gonads, June 6
to 26; weights 20 to 25 grams. Robin-chats were not found within the forested
areas proper, but appeared to be common in the bracken fern and Leonotis
thickets of the larger meadows.

Cichladusa guttata rufipennis Sharpe

The spotted morning warbler was found to be common at the west base
of Mt. Nyiru, 5240 feet, and four specimens were collected there, one male
with slightly enlarged testes and one female with a large ovary on June 4, one
whose sex was unrecorded on June 6, and another female with a small ovary
on June 26; weights 20 to 30 grams; stomach contents small beetles and other
insect remains.

These four specimens are very slightly duller, less rufescent above, than
others examined from Bura, in the lower Tana Valley, and from Lake Magadi,
in southern Kenya, but are rufipennis and not guttata. The Samburu area is
close to where the ranges of these two subspecies meet.

Pogonocichla stellata guttifer (Reichenow and Neumann)

The Kenya white-starred bush robin was found to be a very common bird
in the montane forest, 8400 to 8600 feet, on Mt. Nyiru, and eight specimens,
three males and five females, all in non-breeding condition, were obtained
June 11 to 18; weights 15 to 20 grams.

Mt. Nyiru is a new locality for this small thrush. In Moreau’s map (1966:
191) no locality records are indicated for it north of the Cherangani Moun-
tains, although he does list it from Kulal a few pages later ( 195) . These bright
colored bush robins appeared to be unusually tame about our camp atop Mt.
Nyiru and frequently foraged openly about our tents.

T urdus abyssinicus abyssinicus Gmelin

This thrush was found to be very common in the montane forest on top
of Mt. Nyiru, 8400 to 8600 feet, where a series of nine specimens, four males,
three females, and two unsexed birds, were collected June 10 to 20. The birds
showed a range from non-breeding condition to much gonadal enlargement;
their weights ran from 50 to 65 grams.

T urdus tephronotus Cabanis

One female, with only slight ovarian enlargement, was taken at 5240 feet
on the west base of Mt. Nyiru, on June 25; weight 55 grams.

1969

Birds of Avil Expedition to Kenya

21

Zoothera piaggiae piaggiae (Bouvier)

This montane forest thrush is represented in the collection by one male,
with slightly enlarged testes, 8400 feet, Mt. Nyiru, June 17; weight 45 grams:
stomach contents small snails. The specimen agrees with others from Kapen-
guria and from Mt. Moroto.

Family Sylviidae

Brady pterus cinnamomeus cinnamomeus (RUppell)

The bracken warbler was common in the montane woodlands zone on
Mt. Nyiru, 8400 to 8600 feet; nine specimens, five males and four females,
were obtained June 10 to 18; three of the males and one of the females had
enlarged gonads, the others, including some immature birds, did not. The birds
weighed 15 to 25 grams (average 19 grams).

These examples are uniformly darker, less rufous above, than ten others
examined from Mt. Kenya, the west Kenya highlands, Aberdares, and Mt.
Moroto in northeastern Uganda. The same difference in dorsal coloration
holds for the immature as well as for the adult birds. The Mt. Nyiru birds do
not agree with the description of elgonensis. However, it may be noted that
Granvik (1923: 239-240) found his Elgon specimens inseparable from nomi-
nate cinnamomeus.

In an earlier study Friedmann (1937: 167-170) had a much longer se-
ries, totalling 79 specimens, ranging from Ethiopia to northeastern Tanzania
(Usambara Mountains) and found the variation in coloration and size to be
inconstant enough to cast doubt on the validity of any races in this large area.
Van Someren (1932: 373) attempted to explain the situation by stating that
this warbler shows variation, “. . . due to different ecological factors, at varying
altitudes, producing either a paleness or an intensity of coloration, not bound
by geographical distribution in its usually accepted sense.” At our request
M. A. Traylor kindly compared the present Nyiru examples with the extensive
material in the Field Museum of Natural History and concluded that there
were no recognizable local races in eastern Africa, that mildbreadi of Ruwen-
zori was a darker and more brownish, less reddish subspecies, but that the
Nyiru birds, which on geographic and on presumed genetic grounds could
not possibly be the same as the Ruwenzori population, actually agreed more
closely with that group than with nominate cinnamomeus ! It may be noted, in
this connection, that Chapin (1953: 437) recognized chyulensis as a very
dark race, so the question of its relation to the Nyiru population is thereby
raised. It seems best, at least in the present state of our understanding, to con-
sider all east African birds part of a variable nominate race.

For the benefit of students who may wish to study the bracken warbler
in further detail, it may be recorded that the Nyiru birds, besides being darker,
are also generally longer winged than are the others examined from Mt.
Kenya, Aberdares, and Moroto: males — wing 65.5 to 66 as against 60 to 63

22

Contributions in Science

No. 174

mm.; females — 60 to 62 as against 54 to 58 mm. The tails are long in Nyiru
males (71 to 76 as against 61 to 71 mm. in others), but not in the Nyiru
females (61 to 69 as against 61 to 68 mm.).

Phylloscopus umbrovirens mackenzianus (Sharpe)

The brown woodland-warbler was found to be a common species in the
montane forest on Mt. Nyiru, 8400 feet, June 10 to 15, during which time six
specimens, two males with enlarged testes and four females with small ovaries
were collected; weight 5 grams in each case.

Moreau’s map (1966:192) indicates no locality records for this species
between the vicinity of Kapenguria and about 7° N in Ethiopia, but the bird
does occur on Nyiru, Kulal, and Marsabit; Moreau lists it from the last named
locality (1966: 194) although the record is not shown on his map.

Sylvietta whytii jacksoni Sharpe

One female in non-breeding condition, west base of Mt. Nyiru, 5240 feet,
June 4; weight 5 grams. The red-faced crombec had been found there in 1966
by the Cheney Expedition as well.

Camaroptera brevicaudata griseigula Sharpe
As was the experience of the Cheney Expedition in 1966 in the Karissia
Hills, to the south, the Avil Expedition also found this gray-backed warbler
common at the west base of Mt. Nyiru, 5240 feet, where 1 1 specimens were
taken between June 3 and 25, six males and five females, with gonads varying
from small to enlarged; weights 10 to 15 grams. Van Someren (1922: 228)
had recorded specimens of this warbler from the N’ziu River, Mt. Nyiru.

Cisticola nana Fischer and Reichenow
This tiny grass warbler was met with at the west base of Mt. Nyiru, 5240
feet, on June 8, when one female, ovary slightly enlarged, was collected; weight
5 grams.

Cisticola cinereola Salvadori

Two males of this grass warbler were collected on the west base of Mt.
Nyiru, 5240 feet, June 6 and 7; both had only slight testicular enlargement;
weight 15 grams.

Chloropeta similis Richmond

This yellow flycatcher-like warbler proved to be common in the montane
forest on Mt. Nyiru, 8400 to 8600 feet, June 10 to 18, when six specimens, five
males and one female, all in non-breeding state, were collected. They all have
the top of the head greenish, like the rest of the underparts, and are thus similis
and not a race of C. natalensis.

1969

Birds of Avil Expedition to Kenya

23

Mt. Nyiru is the northernmost locality in Kenya from which the species
has yet been recorded; it is known farther to the west, as far north as the
Imatong Mts., southern Sudan.

Family Campephagidae
Coracina caesia pur a (Sharpe)

The gray cuckoo-shrike was found in the montane forest on Mt. Nyiru,
at 8400 feet, on June 25, when an adult female with an enlarged ovary was
taken; weight 50 grams.

Family Dicruridae

Dicrurus adsimilis divaricatus (Lichtenstein)

The common drongo is represented by two specimens, both males with
slight gonadal enlargement, west base of Mt. Nyiru, 5240 feet, June 3; weight
30 grams each.

Family Laniidae
Eurocephalus riippelli Bonaparte

One female, ovary slightly enlarged, west base of Mt. Nyiru, 5240 feet,
June 7; weight 55 grams.

Dryoscopus gambensis malzacii (Heuglin)

This puff-backed shrike was found at the west base of Mt. Nyiru, 5240
feet, June 3 and 4, one non-breeding female and one male with enlarged go-
nads; weight 30 grams each. This species was already known from the Samburu
district (Karissia Hills).

Tchagra cruentata hilgerti (Neumann)

One male, testes slightly enlarged, west base of Mt. Nyiru, 5240 feet, June
25; weight 50 grams; others taken in the same place in 1966 by the Cheney
Expedition.

Tchagra jamesi jamesi (Shelley)

The Avil Expedition found this bird at the west base of Mt. Nyiru, 5240
feet, June 4, one male with slight testicular swelling; weight 30 grams.

Laniarius ferrugineus ambiguus Madarasz
This race of the boubou shrike was numerous in the montane forest on
Mt. Nyiru, 8400 to 8600 feet, June 12 and 16, two males and three females,
with gonads ranging from small to large; weight 50 grams in each case.

24

Contributions in Science

No. 174

Laniarius funebris funebris (Hartlaub)

The slate-colored boubou shrike was very common at the west base of
Mt. Nyiru, 5240 feet; five males and two females, all in non-breeding state,
taken June 3 to 24; weight 30 to 49 grams. Also found in that place in 1966
by the Cheney Expedition.

Lanius collaris humeralis Stanley

One male, testes slightly enlarged, west base of Mt. Nyiru, 5240 feet,
June 3; weight 35 grams.

Nilaus afer minor Sharpe

One female in non-breeding state was taken on the west base of Mt.
Nyiru, 5240 feet, June 6; weight 20 grams.

Family Paridae
Parus afer barakae Jackson

The gray tit was found at the west base of Mt. Nyiru, 5240 feet; one fe-
male, ovary not enlarged, June 3; weight 10 grams.

Parus albiventris Shelley

One male white-breasted tit, with small testes, was collected on the west
base of Mt. Nyiru, 5240 feet, June 4; weight 20 grams.

Family Remizidae
Anthoscopus musculus (Hartlaub)

One male, testes small, west base of Mt. Nyiru, 5240 feet, June 6; weight
5 grams.

Family Corvidae
Corvus rhipidurus Hartert

Two male fan-tailed ravens, both with enlarged gonads, were obtained
on Mt. Nyiru, one at the west base, 5240 feet, and one on the top, 8400 feet,
June 4 and 13. The species was already known from the Samburu District.

Corvus aibicollis Latham

This is another raven that ranges from the lower plains around Mt. Nyiru
to the top of the mountain; one male, with large testes, was taken at 8400 feet,
June 20. In 1966 the Cheney Expedition collected one at the west base of the
mountain.

Corvus capensis Lichtenstein

A single specimen of the Cape rook, a male in non-breeding condition.

1969

Birds of Avil Expedition to Kenya

25

was taken at the west base of Mt. Nyiru, June 6; weight 450 grams. Rooks of
this species were extremely abundant in the plains area west of Mt. Nyiru,
where flocks of 200 birds were not an uncommon sight.

Family Sturnidae

Lamprotornis chalybeus cyaniventris (Blyth)

Common at the west base of Mt. Nyiru, 5240 feet; two specimens of each
sex, the males with large, the females with small, gonads, June 3 and 4; weight
75 and 95 grams.

Spreo superbus (Riippell)

This starling is a common bird in the Samburu District; one of each sex
was taken on the west base of Mt. Nyiru, 5240 feet, June 3; both in non-breed-
ing state; weight 55 and 60 grams.

Buphagus erythrorhynchus (Stanley)

One male, testes enlarged, west base of Mt. Nyiru, 5240 feet, June 24;
weight 45 grams.

Family Zosteropidae
Zoster ops senegalensis jacksoni Neumann
This white-eye was very common on top of Mt. Nyiru, 8400 to 8600 feet
(2545 to 2606 m); 13 specimens, six males with enlarged gonads, seven fe-
males, all in non-breeding state, were collected, June 10 to 21 ; weights 10 to 15
grams,

Moreau (1957: 361-362) had seven specimens from the top of Mt. Nyiru
(9000 feet) and found that they were a little larger and less dully colored than
jacksoni from Marsabit and Laikipia, and approached kulalensis in this respect.
However, the last named race, found only on Kulal, has the abdomen grayish,
not yellow-green as in the Nyiru and Marsabit specimens. The present birds
are certainly more properly treated as jacksoni, although they are duller below,
less yellowish green, more grayish green, than typical jacksoni. White (1963:
89-92) also concluded that the populations on Nyiru, Marsabit, and the Mat-
thews Range were best included under that name. Moreau found his seven
Nyiru birds to be variable; the present 13 are quite uniform.

Family Nectariniidae
Nectarinia mediocris mediocris (Shelley)

This sunbird was found to be very common in the montane forest on Mt.
Nyiru, 8400 to 8600 feet, and 13 specimens were collected June 10 to 19, eight
males and five females, with gonads ranging from small to much enlarged; all
the birds weighed 5 grams each. The present species occurs north to Mts. Nyiru
and Kulal; it has not yet been reported from Marsabit.

26

Contributions in Science

No. 174

Nectarinia tacazze jacksojii Neumann

A common bird in the montane forest on Mt. Nyiru, 8400 to 8600 feet;
eight specimens, five males with enlarged testes and three females in non-
breeding state, were collected June 10 to 19; weights 10 to 15 grams.

Tacazze sunbirds frequented the more open areas of the mountain top
rather than the glades and were most commonly encountered around the
bracken fern and Leonotis thickets.

Nectarinia mariquensis osiris Finsch

The Avil Expedition found this sunbird on the west base of Mt. Nyiru
and collected two males and one female there, June 3 and 24; the female in
breeding condition, one male with enlarged, the other with small, testes. Two
years earlier the Cheney Expedition found the mariqua sunbird very numerous
in that same area.

Nectarinia famosa aeneigularis Sharpe
Three males, all with enlarged testes, collected at 8600 feet, June 17 to
19, attest to the presence in numbers of this race of the malachite sunbird in
the montane forest zone of Mt. Nyiru. It has not yet been reported from Kulal
or Marsabit, and because it is a conspicuous bird in the field the lack of records
from those two mountains may be indicative of its absence there.

Malachite sunbirds were the least abundant of the four species of the
family present on the top of Mt. Nyiru and, when encountered, proved to be
exceedingly wary. When approached they would immediately take wing and
fly great distances rather than just move a few yards to a new perch as did the
other species.

Nectarinia reichenowi lathburyi Williams

One of the prizes of the present collection is a superb series of 22 speci-
mens of this race of the golden-winged sunbird. The subspecies was described
from the summit of Mt. Garguez, 7100 feet, in the Matthews Range, and to
date is known only from that area and Mt. Nyiru, about 60 miles (100 km.) to
the north. All but two of 15 males, and four of the seven females collected had
large gonads; the birds weighed from 10 to 15 grams.

Golden-winged sunbirds were first observed on the west slope of Mt.
Nyiru at 7800 feet elevation as our pack train neared the western crest of the
mountain. The species proved to be exceedingly abundant in all of the forest
glades and small openings in the forest. It was equalled in numbers only by
Nectarinia mediocris.

These large sunbirds obviously serve as an important pollinator of the
large orange-blossomed mint ( Leonotis nepetifolia) , a common herb of the
forest glades and meadows atop Mt. Nyiru. The short velvet-textured feathers
covering the forehead of this sunbird are separated by a median bare patch

1969

Birds of Avil Expedition to Kenya

27

Figure 8. Drawing showing forehead of the golden-winged sunbird ( Nectarinia
reichenowi ) in contact with the corolla of Leonotis nepettfolia; also view of head
showing bare medial patch.

28

Contributions in Science

No. 174

2 mm wide and extending posteriorly from the culmen, often as much as 8 mm.
It was noted that this bare groove in all specimens collected was filled with a
bright red-orange pollen, especially in the male birds. The upper lip of the
corolla of Leonotis is long (3.8 cm) and decurved, the four stamens and anther
extend under this protective lip. Observation of this species of sunbird feed-
ing on Leonotis showed that as the bird thrust its long decurved bill deep into
the flower, the upper lip of the corolla covered the median area of the bird’s
forehead, thus permitting the groove to fill with pollen (Fig. 8). Examination
of the foreheads of specimens of the other two large sympatric species, Nec-
tarinia jacazze and N. famosa, reveals no such bare area on the forehead. It is
doubtful that the bare patch is acquired through abrasion, as an adult male of
T. reichenowi collected on Mt. Nyiru on October 11, 1955, by J. Smart, has
the entire head in pin feather molt and the bare median patch is without any
trace of feathering.

Family Ploceidae

Dinemellia dinemelli dinemelli (Riippell)

The white-headed buffalo-weaver is a common and conspicuous bird in
the bushveld of northern and eastern Kenya. On the west base of Mt. Nyiru,
5240 feet, two males were collected on June 3; one with small, one with en-
larged, testes; weights 70 and 75 grams.

Plocepasser tnahali melanorhynchus Bonaparte
One male with enlarged testes, one female with slight ovarian enlarge-
ment, west base of Mt. Nyiru, 5240 feet, June 3 and 7; weight 40 and 49 grams.
The species was found to be as common there as it was two years earlier by the
Cheney Expedition.

Passer griseus gongonensis (Oustalet)

The gray-headed sparrow is represented in the collection by one imma-
ture male, taken on June 7, west base of Mt. Nyiru, 5240 feet; weight 30 grams.

Ploceus velatus uelensis (Neumann)

One male, testes slightly enlarged, west base of Mt. Nyiru, 5240 feet,
June 4; weight 25 grams. Also found there by the Cheney Expedition.

Malimbus rubriceps leuconotus (Muller)

Two specimens, one of each sex, with slight gonadal enlargement, west
base Mt. Nyiru, 5240 feet, June 3; weight 25 grams each. Recorded from there
by the Cheney Expedition as well.

1969

Birds of Avil Expedition to Kenya

29

Family Estrildidae

Cryptospiza salvadorii salvadorii Reichenow
The crimson-wing was very abundant in the montane forest, 8400 to 8600
feet, on Mt. Nyiru, where 12 specimens were collected June 10 to 19, five males
and seven females, with gonads ranging from not at all to much enlarged;
weights 10 to 15 grams. The absence of records from Kulal seems likely to be
corrected in the future, considering its abundance on Nyiru. Its absence from
Marsabit may be due to the lesser altitude attained by that mountain.

These birds are less different from a series of kilimensis, from central
Kenya highlands, than the description of that race would suggest. There is
some variation in the Nyiru birds, and since White (1963: 178) found salva-
dorii to intergrade with kilimensis as far north as Mt. Uraguess, the Nyiru birds
may also be intergrades in varying degrees. No topotypical Ethiopian salva-
dorii have been available for comparison.

Estrilda melanotis kilimensis (Sharpe)

A bird of the montane forest edges, but not wholly restricted to the true
montane zone. This yellow-bellied waxbill was found in the montane zone,
8500 to 8600 feet, and also on the west base, 5240 feet, of Mt. Nyiru, June 16
to 25. A total of seven birds, five males, two females, with gonads ranging
from small to enlarged, were taken, weight 5 grams in all cases.

Due to an unfortunate lapsus, specimens of this species collected by the
Cheney Expedition in the Karissia Hills (Friedmann and Stager, 1967: 32)
were reported under the name Estrilda paludicola. They and the present
Nyiru specimens help fill the geographic gap between the ranges of kilimensis ,
otherwise known north only to Mt. Elgon and the south-central Kenya high-
lands, and of quartinia of Ethiopia.

Estrilda ianthinogaster (Reichenow)

Two males, one adult, one immature, west base of Mt. Nyiru, 5240 feet,
June 7 and 24; weight 10 grams each; also found there by the Cheney Expedi-
tion.

Family Fringillidae
Serinus dorsostriatus maculicollis Sharpe
One female, ovary not enlarged, was taken in the montane zone, 8600
feet, on Mt. Nyiru, June 17; weight 10 grams.

Serinus striolatus striolatus (Riippell)

This streaky seed-eater was found to be abundant in the montane zone,
8400 feet, on Mt. Nyiru, and 11 specimens were taken, June 10 to 18, six
males and five females, mostly in breeding condition; weights 15 to 25 grams.
This species has not yet been recorded from Kulal or Marsabit. The Cheney
Expedition did find it in the Karissia Hills.

30

Contributions in Science

No. 174

Literature Cited

Chapin, J. P. 1953. The birds of the Belgian Congo, Part 3. Amer. Mus. Nat. Hist..
Bull., 75 A: 1-821.

Dale, I. R., and P. J. Greenway. 1961. Kenya Trees and Shrubs. Buchanan’s Kenya
Estates Ltd., Nairobi: 1-654.

Friedmann, H. 1937. Birds collected by the Childs Frick expedition to Ethiopia
and Kenya Colony, Pt. 2. U. S. Natl. Mus., Bull., 153: 1-506.

Friedmann, H., and K. E. Stager. 1967. Results of the 1966 Cheney expedition to
the Samburu district, Kenya. Ornithology. Los Angeles Co. Mus., Contrib. Sci.,
130: 1-34.

Granvik, H. 1923. Contributions to the knowledge of the east African ornithology.
Journ. fur. Ornith., 71, Sonderheft: 1-280.

. 1934. The ornithology of north western Kenya Colony with special regard

to the Suk and Turkana districts. Revue Zool. et Bot. Africaines, 25: 1-190.

Gyldenstolpe, N. 1924. Zoological results of the Swedish expedition to central
Africa, 1921. Birds. K. Svenska Vetenskapsakad. Handl., (3) 1, (3): 1-326.

Jackson, F. J. (completed by W. L. Sclater) 1938. The birds of Kenya Colony and
the Uganda Protectorate. London: Gurney and Jackson. 3 vols., 1592 pp., 24
pis.

Moreau, R. E. 1957. Variation in the western Zosteropidae (Aves). British Mus.
(Nat. Hist.), Bull., Zool., 4, 311-433.

. 1966. The bird faunas of Africa and its islands. Academic Press, New

York. 424 p.

Neumann, O. 1903. Ueber einige afrikanische Vogel. Ornith. Monatsberichte, 11:
87-91.

Rand, A. L. 1958. Notes on African bulbuls. Fieldiana, Zool., 35: 145-220.

Sclater, W. L. 1930. Systema Avium Aethiopicarum. London: Taylor and Francis,
pt. 2: 305-922.

Sharpe, H. B. 1930-31. The birds of Marsabit mountain, Kenya Colony. Bateleur,
2: 44-47, 102-107; 3: 9-14, 35-39, 97-103.

Twomey, A., and S. Keith. 1968. New distributional records of some east African
birds. Ibis, 110: 537-548.

van Someren, V. G. L. 1922. Notes on the birds of East Africa. Novitates Zool., 29:
1-246.

. 1931. New races of birds from eastern Africa. East Africa and Uganda

Nat. Hist. Soc., Journ., no. 37: 193-197.

. 1932. Birds of Kenya and Uganda, being addenda and corrigenda to my

previous paper in “Novitates Zoologicae.” xxix, 1922. Novitates Zool.. 37:
252-380.

. 1939. Birds of the Chyulu Hills. East Africa and Uganda Nat. Hist. Soc.,

Jour., 14: 15-129.

White, C. M. N. 1963. A revised check list of African flycatchers, tits, tree creepers,
sunbirds, white-eyes, honey eaters, buntings, finches, weavers and waxbills.
Lusaka, Govt. Printers. 218 p.

. 1965. A revised check list of African non-passerine birds. Lusaka. Govt.

Printers. 299 p.

Williams, J. G. 1948. A new race of white eye from Kenya Colony. British Ornith.
Club, Bull., 68: 101-102.

. 1956. A New Golden-winged Sunbird from Kenya. British Ornith. Club.

Bull., 76: 136-139.

Accepted for publication June 6. 1969.

LOS

ANGELES

COUNTY

MUSEUM

CONTRIBUTIONS
IN SCIENCE

fUMBER 175

June 30, 1969

SYSTEMATICS AND EVOLUTION OF NEOTROPICAL
SALAMANDERS OF THE BOLITOGLOSSA HELMR1CH1 GROUP

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

Los Angeles County Museum of Natural History • Exposition Park
Los Angeles, California 90007

CONTRIBUTIONS IN SCIENCE is a series of miscellaneous technical papers
in the fields of Biology, Geology and Anthropology, published at irregular intervals
by the Los Angeles County Museum of Natural History. 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 Robert J. Lavenberg, Los Angeles County Museum of Natural
History, 900 Exposition Boulevard, Los Angeles, California 90007.

Robert J. Lavenberg
Managing Editor

INSTRUCTIONS FOR AUTHORS

Manuscripts for the LOS ANGELES COUNTY MUSEUM, CONTRIBU-
TIONS 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. Al-
though priority will be given to manuscripts by staff members, or to papers dealing
largely with specimens in the Museum’s collections, 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 Edi-
torial Committee. They may also be subject to review by competent specialists out-
side 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.

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. Alterations or changes in the manuscript after galley proof will
be billed to the author. Unless specifically requested, page proof will not 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.

Dorothy M. Halmos
Editor

SYSTEMATICS AND EVOLUTION OF NEOTROPICAL
SALAMANDERS OF THE BOLITOGLOSSA HELMRICHI GROUP

By David B. Wake1 and Arden H. Brame, Jr.2

Abstract: The Bolitoglossa helmrichi group includes B.
helmrichi and B. cuchumatana of Guatemala, B. flavimembris of
Guatemala and Mexico, and two new species, B. hartwegi from
Chiapas, Mexico, and B. stuarti from near the Mexican-Guate-
malan border. The group bridges the gap between the specialized
lowland members of the rufescens group and the more general-
ized species of the uplands of Nuclear Central America, and pro-
vides support for an earlier decision to synonymize the genus
Magnadigita with Bolitoglossa. In particular, hand and foot
morphology of these species represents all stages of intermediacy
between the slightly webbed, large digited structures of the up-
land species and the diminutive, fully webbed, small digited ones
of the lowland species. As a result of the discovery of the new spe-
cies, a nearly continuous morphocline is apparent which provides
insights into the morphological events accompanying the evolu-
tion of a tropical lowland salamander fauna.

Lungless salamanders of the family Plethodontidae are notably successful
in the New World Tropics, where more than forty per cent of salamander
species occur. All are members of a single radiation, the supergenus Bolito-
glossa (Wake, 1966). Salamanders are primarily a North Temperate group,
and their entrance into the tropics came long after the major radiation of the
order Caudata. The relative abundance of neotropical salamanders and the
correlation of increased species diversity with tropical environments invite
further scientific investigation. In earlier papers we have presented the results
of our systematic studies on lower Central and South American salamanders
of the genus Bolitoglossa (Wake and Brame, 1962, 1963a, 1963b, 1966a,
1966b; Brame and Wake, 1962a, 1962b, 1963a, 1963b). This paper is an
analysis of five species of the Bolitoglossa helmrichi group from the highlands
of Chiapas, Mexico, and Guatemala. These species provide data which eluci-
date evolutionary patterns associated with invasion of tropical lowlands, and
enable formulation of hypotheses concerning morphological evolution during
phylogeny. In future papers we plan to analyze the remaining species of the
genus, with emphasis on patterns of evolution.

The helmrichi group can be defined as follows: salamanders of the genus
Bolitoglossa having small to moderate size (adults average from 38 to 53 mm

Research Associate in Herpetology, Los Angeles County Museum of Natural His-
tory, and Department of Anatomy, University of Chicago, 1025 East 57th Street, Chi-
cago, Illinois 60637.

2Section of Herpetology, Los Angeles County Museum of Natural History.

1

2

Contributions in Science

No. 175

standard length (abbreviated SL throughout remainder of text), 60 mm in the
single adult of stuarti ); relatively broad heads (SL averages 5.5 to 6.3 times
head width, 6.7 in the rather narrow-headed adult of stuarti ) ; eyes moderate
in size, not strongly protuberant; snout pronounced to moderate; moderate to
high numbers of marginal teeth; limbs relatively long (limb interval averages
0.5 to 1.0); hands and feet with moderate to extensive webbing, moderate to
extensive flattening, and digits of reduced size; tail shorter than body; light
ventral pigmentation, especially on tail.

Content: Bolitoglossa cuchumatana, B. flavimembris, B. hartwegi, B.
helmrichi, B. stuarti.

In the following sections brief redescriptions and diagnoses are presented
for the Guatemalan species of the helmrichi group, B. flavimembris, B. helm-
richi, and B. cuchumatana, and two new members of the group are described.

Bolitoglossa flavimembris (Schmidt, 1936)

Oedipus flavimembris Schmidt, 1936: 158.

Magnadigita flavimembris, Taylor, 1944: 218.

Bolitoglossa flavimembris, Wake and Brame, 1963a: 386.

Holotype : FMNH 20381, an adult female from Volcan Tajumulco at
2195 m (7200 feet) on the trail above El Porvenir, San Marcos, Guatemala,

103

Material Examined : FMNH 20322 (cleared and stained), 20333 (2),
100128, UMMZ 80935, all topoparatypes, 10; UIMNH (field numbers) 2538,
2545, Mexico, Chiapas, Volcan Tacana, El Chicuite, 9, 2195 m (7200 feet);
FMNH 142081, Guatemala, San Marcos, near Aldea la Fraternidad between
San Rafael Pie de la Cuesta and Palo Gordo, 1 1 , 2400 m (7900 feet) .

Diagnosis’. A moderate-sized species of Bolitoglossa (1 adult male: 50.1
mm SL; 6 adult females: 40.5 - 64.5, mean 53.2 mm SL) with moderate num-
bers of maxillary (mean 49.1) and relatively low numbers of vomerine (mean
23.8) teeth in adults, distinguished from other members of the helmrichi group
by its larger size, more robust habitus, broader head, relatively smaller and
less completely webbed hands and feet (Fig. 15) and distinctive coloration
(“. . . back in life uniform Hair Brown, lighter on sides and venter, lightest on
chin, limbs Colonial Buff, feet dark gray above and below, and an obscure dark
line down middle of belly,” Schmidt, 1936).

Description: Bolitoglossa flavimembris is a moderate to moderately large,
robust species with a relatively short, broadly rounded snout. The one adult
male is smaller than the mean size of the six adult females; females of Bolito-
glossa are typically larger than males. The nostrils are small. Labial pro-
tuberances are poorly developed in males and females. A mental hedonic gland

:Bold face numerals here and below refer to plotted localities on map (Fig. 3).

1969 Systematics and Evolution of Neotropical Salamanders

3

50

40

0)

<1>

H

<D

.E 30

w-

Cl>

E

o

>

20

10

+ = helmrichi
• =hartwegi
s =stuarti
□ =flavimembris
o = cuchumatana

□ □ o

• •

++

• + □

• +

O □

30 40 50 60

Snout-Vent Length (mm)

Figure 1. Variation in vomerine dentition in the Bolitoglossa helmrichi group.

70

is poorly developed in the male. The head is moderately broad, but rather
variable (SL 5.4 - 6.5, mean 5.8 times head width in six females; 6.3 in male).
A relatively deep groove extends below the eye, following its curvature, but
does not communicate with the lip. The moderate-sized eyes are rather pro-
tuberant and are visible beyond the margin of the jaw when viewed from below.
A shallow postorbital groove extends posteriorly from the eye as a shallow,
irregular depression. At the posterior end of the mandible the groove proceeds
sharply ventrally and extends across the throat, anterior to the gular fold, as a
moderately defined depression. Vomerine teeth increase in number slowly with
increasing size (Fig. 1 ) . The teeth are typically in long, single series and extend
beyond the medial border of the internal nares. Maxillary teeth extend beyond
the center of the eye and increase in number slowly with increasing size (Fig.
2). Few premaxillary teeth (3-7) are present and they pierce the lip in the
adult male. The trunk and tail are robust. The stout tail is nearly round in cross
section and is rather strongly constricted at its base. Lightly pigmented postiliac
glands are barely evident. Limbs are robust and of moderate length. Limb
interval (costal folds between appressed limbs) varies from 1 to 2 (mean
1.5) in females and is 0.5 in the male. Hands and feet are moderate in size and

4

Contributions in Science

No. 175

the digits, while moderately webbed, are discrete (Fig. 15). Webbing is less
extensive than in other members of the group. Subterminal pads are well
developed, but the joints of the digits are not well marked. The hands and the
feet are relatively flattened with a rather large surface area. The fingers are, in
order of decreasing length, 3, 2, 4, 1; toes, in order of decreasing length, 3, 4,
(2-5), 1. Pertinent counts and measurements of material are found in Table 1.

Measurements of Holotype (in mm): Head width 10.1; snout to gular
fold (head length) 14.2; head depth at posterior angle of jaw 5.8; eyelid length
3.8; eyelid width 2.2; horizontal length of postorbital groove 3.3; distance
between ventral extension of postorbital groove and gular fold 5.8; anterior
rim of orbit to snout 3.7; horizontal orbital diameter 2.6; interorbital distance
3.3; distance between vomerine teeth and parasphenoid tooth patch 0.8; snout
to fore limb 18.9; distance separating internal nares 2.8; distance separating
external nares 3.3; snout projection beyond mandible 0.8; snout to posterior
angle of vent (SL) 59.4; snout to anterior angle of vent 55.6; axilla to groin
31.4; tail length 39.3; tail width at base 5.6; tail depth at base 5.9; fore limb
length 13.7; hind limb length 14.9; width of right hand 5.2; width of right
foot 5.8.

Coloration : This species is unusually colored for members of the genus.
Coloration is uniform dark brown dorsally and light ventrally, with light yel-
lowish limbs and yellowish tan tail venters. The dark feet are a strong contrast
to the yellow limbs, as pointed out by Schmidt ( 1 936) , who also presented good
photographs of the species.

Range : Known from the slopes of Volcan Tajumulco and Volcan Tacana
and from the mountains along the southwestern escarpment of the Guatemalan
Plateau, at elevations between 7000 and 8000 feet, in San Marcos, Guatemala,
and the border area of Chiapas, Mexico (Fig. 3). This report constitutes the
first record of the species in Mexico.

Bolitoglossa cuchumatana (Stuart, 1943)

Oedipus cuchumatanus Stuart, 1943: 14.

Magnadigita cuchumatana, Taylor, 1944: 218.

Bolitoglossa cuchumatana. Wake and Brame, 1963a: 386.

Holotype : UMMZ 89110, an adult male from an oak forest about 2
kilometers north of Nebaj, approx. 1900 m (6250 feet), El Quiche, Guatemala,

12.

Material Examined: UMMZ 89111-13, topoparatypes, 12.

Diagnosis : A small species (3 adult males: 35.2 - 43.5, mean 38.2 mm
SL; 1 female 33.4 mm SL) with low numbers of teeth (maxillary: 26 - 38,
mean 34.7; vomerine: 15 - 22, mean 18); distinguished from B. flavimembris
by smaller size, more extensively webbed feet, and banded (or mottled) body

1969 Systematics and Evolution of Neotropical Salamanders

5

90

80

+ = helmrichi
• =hartwegi
s =stuarti
□ =flavimembris
o =cuchumatana

s

70

o 50

□

40

o o

• o

□

30

20

30

40 50 60

Snout-Vent Length (mm)

70

Figure 2. Variation in maxillary dentition in the Bolitoglossa helmrichi group.

6

Contributions in Science

No. 175

coloration; from B. helmrichi by fewer maxillary and vomerine teeth; from
B. hartwegi and stuarti by less extensively webbed hands and feet, larger termi-
nal phalanges, and subterminal digital pads.

Description : This small, robust species has a relatively short, rounded
snout. Sexual dimorphism is indicated by the fewer maxillary and vomerine
teeth of the female and by the forward placement of the premaxillary teeth in
males. The nostrils are small. Labial protuberances are well developed in both
sexes, and are pronounced in males. Mental hedonic glands are moderately
developed in males. The head is rather broad (SL 5.1 - 6.0, mean 5.5 times
head width). A distinct groove extends below the eye, following its curvature,
but does not communicate with the lip. The moderate-sized, protuberant eyes
extend slightly beyond the lateral margins of the jaw. A shallow, postorbital
groove extends posteriorly from the eye as a shallow, irregular depression.
At the posterior end of the mandible the groove proceeds sharply ventrally
and extends across the throat anterior to the gular fold, as a poorly defined
depression. Vomerine and maxillary teeth increase in number with increasing
size (Figs. 1 and 2). Premaxillary teeth are few (1 - 4), and pierce the lip in
adult males. The trunk and tail are moderately robust. The swollen tails are
nearly round in cross section and are strongly constricted basally. The mod-
erately robust limbs are moderately long and limb interval varies from 0.5 to
2.0 (mean 1.0). Hands and feet are relatively large. Webbing is moderately
extensive (Fig. 15) and leaves one-half to one phalanx free of the web on the
inner digits. The digits bear small but discrete subterminal pads. The fingers
are, in order of decreasing length, 3, 4, 2, 1 ; toes, in order of decreasing length,
3, 4, 5, 2, 1. Pertinent counts and measurements of the type series are included
in Table 1.

Measurements of Holotype (in mm) : Head width 7.3; snout to gular fold
(head length) 12.2; head depth at posterior angle of jaw 3.2; eyelid length 2.9;
eyelid width 1.8; anterior rim of orbit to snout 3.3; horizontal orbital diameter
3.2; interorbital distance 3.9; distance between vomerine teeth and parasphe-
noid tooth patch 0.6; snout to fore limb 14.2; distance separating internal nares
2.0; distance separating external nares 2.4; snout projection beyond mandible
0.9; snout to posterior angle of vent (SL) 43.5; snout to anterior angle of vent
40.4; axilla to groin 22.1; tail length 37.7; tail width at base 3.6; tail depth
at base 3.6; fore limb length 11.7; hind limb length 11.8; width of right hand
4.5; width of right foot 5.5.

Coloration : Stuart (1943) has presented a description and an excellent
photograph of the species. He described the dorsal surfaces of the head, body
and tail as mottled with purplish brown and light reddish brown (pinkish in
life), forming a broad middorsal stripe. Narrower, light reddish brown stripes
extend from the eye to the base of the tail in three of the four specimens, with
coloration of the back gradually fading into the ventral coloration in the fourth.
Lateral surfaces are mottled with purple and reddish brown. Ventral surfaces

1969 Systematics and Evolution of Neotropical Salamanders

7

Table I

Measurements and Data for Bolitoglossa cuchumatana ,
B. flavimemkris, and B. helmrichi

H

su

to

Snout-Vent

Length

Axilla-Groin

Length

Head Width

Hind Limb
Length

Fore Limb
Length

Tail Length

Numbers of
Maxillary Teei

Numbers of
Vomerine Tee,

Limb Interval

Foot Width

B. cuchumatana

UMMZ 891 10**

8

43.5

22 A

7.3

11.8

11.7

37.7

38

18

0.5

5.5

UMMZ 89111

8

35.9

18.8

6.5

8.9

8.2

28.2

37

22

2

3.7

UMMZ 89113

8

35.2

—

6.6

8.8

—

26.4

38

17

1

—

UMMZ 89112

9

33.4

—

6.5

8.2

—

23.0

26

15

1

3.7

B. flavimemhris

FMNH 100128

8

50.1

26.2

8.0

13.8

12.0

38.2

46

21

0.5

5.2

FMNH 20333

9

64.5

33.4

9.9

15.6

15.2

45.7

50

25

2

6.8

FMNH 20381**

9

59.4

31.4

10.1

14.9

13.7

39.3

57

31

2

5.8

FMNH 20322

9

49.7

26.0

9.2

12.2

11.1

32.7

48

18

1.5

5.2

UIMNH 2538

9

44.9

23.0

7.9

11.2

10.7

31.1

45

23

1.5

4.2

FMNH 142081

9

40.4

19.2

7.2

10.3

8.7

36.2

50

25

2

4.3

B. helmrichi

UMMZ 89132

8

47.6

24.5

7.8

12.2

11.0

31.8

58

22

1

5.2

FMNH 20153**

8

38.4

19.9

6.8

11.2

9.9

35.8

55

25

0.5

4.3

UMMZ 89133

9

45.7

25.9

7.8

12.0

11.2

43.2

61

38

0.5

4.7

UMMZ 89137

9

42.6

21.2

7.2

10.5

9.3

38.6

51

27

1

4.1

** holotype

of the head and body are pale yellow, and the tail is light orange-yellow
mottled with gray.

Range : Known only from the type locality in the Sierra de los Cuchu-
matanes, El Quiche, Guatemala (Fig. 3).

Bolitoglossa helmrichi (Schmidt, 1936)

(?) Spelerpes morio (part) Brocchi, 1883: 113.

Oedipus helmrichi Schmidt, 1936: 152.

Bolitoglossa helmrichi, Taylor, 1944: 219.

Magnadigita helmrichi, Stuart, 1952: 5.

Bolitoglossa helmrichi, Wake and Brame, 1963a: 386.

Holotype : FMNH 21063, an adult male from mountains above Finca
Samac, west of Coban, Alta Verapaz, Guatemala, 13, at 1300 m (4264 feet).

8

Contributions in Science

No. 175

Material Examined : FMNH 20707 (8, 1 cleared and stained), topo-
paratypes; UMMZ S-1597 (cleared and stained); UMMZ 80929, 89136-7
(14), 89138, Guatemala, AltaVerapaz, FincaSamac, 13, 1350-1500 m (4440-
4920 feet); UMMZ 89132-34, 89135 (2), Guatemala, Alta Verapaz, Finca
Chichen, 14, 1700-1800 m (5600-5900 feet); UMMZ 89131, Guatemala,
Alta Verapaz, Finca Volcan, 15, 1410 m (4640 feet).

Diagnosis: A moderately small species of Bolitoglossa (5 adult males:

38.4 - 47.6, mean 43.1 mm SL; 5 adult females: 41.2 - 45.7, mean 42.8 mm
SL) with moderately high numbers of maxillary teeth (mean 56.4) and
moderate numbers of vomerine teeth (mean 30.9) in adults, distinguished from
B. flavimembris by smaller size, higher numbers of maxillary and vomerine
teeth, and coloration, from B. cuchumatana by higher numbers of maxillary
and vomerine teeth, from B. hartwegi and B. stuarti by less extensively webbed
feet, larger terminal phalanges, and discrete subterminal digital pads.

Description : Bolitoglossa helmrichi is a relatively small, rather stout
species with a moderately large, truncate snout. Males and females are about
the same size, but males apparently mature at a smaller size. The nostrils are
small. Labial protuberances are moderate in females and only slightly better
developed in males. Mental hedonic glands are moderately developed in males.
The head is broad (SL 5.2 - 6.1, mean 5.7 times head width in 4 adult males;

5.4 - 6.1, mean 5.9 in 5 adult females) . A relatively deep groove extends below
the eye, following its curvature, but it does not communicate with the lip. The
moderate-sized eyes are only slightly protuberant and are generally not visible
beyond the margin of the jaw when viewed from below. A shallow postorbital
groove extends posteriorly from the eye as a shallow, irregular depression. At
the posterior end of the mandible the groove proceeds sharply ventrally and
extends across the throat, anterior to the gular fold, as a moderately defined
depression. Vomerine teeth increase in number with increasing size (Fig. 1).
The teeth are typically in long, single series and extend beyond the medial
border of the internal nares. Maxillary teeth extend beyond the center of the
eye and increase in number with increasing size (Fig. 2) . The few premaxillary
teeth (2-3 in males, 3-7 in females) pierce the lip in adult males. The trunk
and tail are of moderate dimensions, neither robust nor slender. The tail is
nearly round in cross section and is strongly constricted at its base. The tail is
always considerably shorter than standard length. Light-colored postiliac glands
are barely evident. The slender limbs are moderately long. Limb interval varies
from 0 to 1.0 (mean 0.5) in males and from 0.5 to 1.5 (mean 1.0) in
females. Hands and feet are moderate in size and the digits are discrete (Fig.
15). Webbing is moderately extensive, but digital tips are free and bear well
developed subterminal pads. The fingers are, in order of decreasing length,
3, 2, 4, 1; toes, in order of decreasing length, 3, 4, 2, 5, 1. Pertinent counts
and measurements of four specimens are found in Table 1.

Measurements of Holotype (in mm) : Head width 6.8; snout to gular fold

1969 Systematics and Evolution of Neotropical Salamanders

9

Figure 3. Distribution of the species of the Bolitoglossa helmrichi group in Chiapas,
Mexico, and Guatemala. The map symbols indicate the collection localities for the
different species: solid circle^#. hartwegi\ open star in solid circles#. stuarti\
open circle=Z?. flavimembris ; open triangle=fi. cuchumatana\ solid triangle=#.
helmrichi. The numbers correspond to the locality data (indicated in bold face) in
the holotype, paratype, and material examined sections of the text.

(head length) 10.2; head depth at posterior angle of jaw 3.1; eyelid length 3.0;
eyelid width 1.7; anterior rim of orbit to snout 2.9; horizontal orbital diameter
2.2; interorbital distance 2.2; distance between vomerine teeth and parasphe-
noid tooth patch 0.8; snout to fore limb 12.1; distance separating internal

10

Contributions in Science

No. 175

nares 2.1; distance separating external nares 2.9; snout projection beyond
mandible 0.8; snout to posterior angle of vent (SL) 38.4; snout to anterior
angle of vent 36.3; axilla to groin 19.9; tail length 35.8; tail width at base
3.3; tail depth at base 3.3; fore limb length 9.9; hind limb length 11.2; width
of right hand 3.8; width of right foot 4.3.

Coloration: Schmidt (1936) presented information concerning coloration
in the type series, as well as photographs. We have examined 23 additional
specimens in the University of Michigan collections and coloration of this
sample is discussed below. In 15 animals (including five juveniles) a distinct
light-bordered dorsal stripe of tan blending to brown is present, with the stripe
most pronounced over the shoulders. The stripe is prominent and well devel-
oped in four individuals (including two juveniles) , where it is especially marked
dorsolaterally and anteriorly, darkening toward the midline. In one additional
individual the entire back is covered with a bright, broad stripe. This stripe
is bordered dorsolaterally by a broken line of dark pigmentation, and bears
a Y-shaped dark mark at the base of the head and scattered dark spots and
blotches down the back and onto the tail. In the remaining eight animals, no
stripe is present. In six of the eight, the dark ground color of the back is
marked with dark spots or blotches in rather irregular arrays, and in the
seventh light spots are present as well. The final individual has nearly uniform
dark dorsal coloration, with some light pigment dorsolaterally, forming an
indistinct border on the tail. The light border or stripe is present in all juveniles
that we have seen, and it appears that this stripe becomes obscured to varying
degrees in adults. Throats and body venters are nearly uniformly light gray in
the sample. Tails are present on 17 animals, and all have light yellowish pig-
mentation ventrally, although most have a few small dark spots and some
have larger dark blotches.

Range : Known from the mountainous regions of southwestern Alta
Verapaz, Guatemala (Fig. 3).

Over the past twenty years, specimens of an undescribed species of
Bolitoglossa have accumulated in several museums. This species, a highland
form, has been mistaken for the lowland species of the rufescens group, which
it resembles in its extensively webbed feet and general morphology. Despite
these similarities, its relatives apparently are members of the helmrichi group.
We name the species in memory of the late Professor Norman Hartweg of the
University of Michigan, an able scientist who spent many years studying
Chiapan amphibians and reptiles and who stimulated and guided many investi-
gations of the biology of tropical organisms.

Bolitoglossa hartwegi new species
Figs. 4-13

Holotype: UMMZ 121557, an adult female from 4.5 miles W San

1969 Systematics and Evolution of Neotropical Salamanders

1 1

Cristobal de Las Casas, 3, Chiapas, Mexico, collected by Floyd L. Downs,
June 30, 1960. The specimen was collected at about 2134 m (7000 feet)
elevation under a log at the top of a slope covered with a pine-oak forest.

Paratypes : All from Chiapas, Mexico. MVZ 57098-57115 (18), 6 mi.
SE San Cristobal de Las Cases, 4, 2226 meters (7300 feet); MVZ 57737-8,
32 mi. SE San Cristobal de Las Casas, 6, 2287 m (7500 feet); MVZ 66191,
35 mi. SE San Cristobal de Las Casas, 7, 2134 m (7000 feet) ; LACM 40603-7,
JFC-65-130 (cleared and stained, to be deposited LACM), Municipio de San
Cristobal de Las Casas, SE slope near summit of Cerro Zontehuitz, 2, 2865 m
(9300 feet); LACM 40601, Municipio de San Cristobal de Las Casas, SE

Figure 4. Holotype of Bolitoglossa hartwegi (UMMZ 121557). Dorsolateral (top)
and ventral (bottom) views.

slope near summit of Cerro Zontehuitz, 2, 2865 m (9400 feet); TCWC
21398-99, near summit of Cerro Zontehuitz, 2, 2805 m (9200 feet); LACM
40602, Municipio de Chamula, San Cristobal de Las Casas to Chenalho Rd.,
15.1 mi. N San Cristobal de Las Casas, 1 , 2835 m (9300 feet) ; LACM 44209,
12 mi. E Teopisca, 16, 2348 m (7700 feet); UAZ 20262, about 12 mi. SE
San Cristobal de Las Casas, 5, 2043 m (6700 feet).

Diagnosis : A moderately small species (7 adult males: 34.0 - 46.9, mean
40.6 mm SL; 11 adult females: 34.2 - 53.6, mean 45.1 mm SL) with moderate
to high numbers of maxillary teeth (female mean, 54.2; male mean, 50.0) and

12

Contributions in Science

No. 175

Figure 5. Head of holotype of Bolitoglossa hartwegi (UMMZ 121557).

moderate numbers of vomerine teeth (female mean, 27.7, male mean, 23.1)
in adults; distinguished from B. fiavimembris by smaller size, more extensively
webbed hands and feet with phalangeal elements, especially the terminal ones,
reduced in size and modified in shape (Figs. 9, 10, 15), no subterminal pads,
and lack of uniform body coloration; from B. cuchumatana by more extensively
webbed hands and feet, smaller terminal phalanges, no subterminal pads, and
more teeth; from B. helmrichi by more extensively webbed hands and feet,
smaller terminal phalanges, and no subterminal pads; from B. stuarti by smaller
size, broader head, less extensively webbed hands and feet, relatively longer
digits, and fewer maxillary teeth. Distinguished from members of the rufescens
group by larger size, broader head, larger limbs, hands and feet, and much
higher numbers of maxillary teeth.

Description : This moderately small, robust species has a relatively short,
truncate snout. Females are larger than males (see Diagnosis) . The nostrils
are small. Labial protuberances are well developed in both sexes, and are
pronounced in males. Mental hedonic glands are moderately developed in
males. The head is broad (SL 5.6 to 5.9, mean 5.7 times head width in males;
5.5 to 6.8, mean 6.0 in females). A distinct groove extends below the eye,
following its curvature, but does not communicate with the lip. The moderate-
sized, protuberant eyes extend slightly beyond the lateral margins of the jaw.
A shallow postorbital groove extends posteriorly from the eye as a shallow,
irregular depression. At the posterior end of the mandible the groove proceeds
sharply ventrally and extends across the throat, anterior to the gular fold, as a
poorly defined depression. Vomerine teeth increase in number with increasing
size (Fig. 1). The teeth are arranged in long or patchy series that extend well
beyond the medial border of the internal nares. The small maxillary teeth
extend beyond the center of the eye and increase in number with increasing

1969 Systematics and Evolution of Neotropical Salamanders

13

size (Fig. 2). The few premaxillary teeth (2-4 in males, 2-11 in females)
pierce the lip in adult males. The trunk and tail are robust. The swollen tails
are nearly round in cross section and are strongly constricted basally. Tail
length is always less than body length. Postiliac glands are only faintly indi-
cated. The slender limbs are moderately long; limb interval varies from 0 to
1.0 (mean 0.5) in males and from 0 to 2.0 (mean 1.0) in females. Hands and
feet are relatively large. Webbing is extensive (Figs. 9, 10, 13, 15). The digits
are discrete only distally and do not bear subterminal pads. The fingers are, in
order of decreasing length, 3, 4, 2, 1; toes, in order of decreasing length, 3, 4,
2, 5, 1. Pertinent counts and measurements of the material studied are included
in Table 2.

Measurements of Holotype (in mm): Head width 7.8; snout to gular
fold (head length) 1 1.8; head depth at posterior angle of jaw 3.9; eyelid length
3.4; eyelid width 1.8; anterior rim of orbit to snout 3.2; horizontal orbital
diameter 2.6; interorbital distance 2.9; distance between vomerine teeth and
parasphenoid tooth patch 0.7; snout to fore limb 14.0; distance separating
internal nares 2.1; distance separating external nares 2.3; snout projection
beyond mandible 0.9; snout to posterior angle of vent (SL) 43.2; snout to
anterior angle of vent 39.9; axilla to groin 22.3; tail length 31.0; tail width at
base 4.2; tail depth at base 4.2; fore limb length 10.3; hind limb length 1 1.2;
width of right hand 3.6; width of right foot 4.8.

Coloration (in alcohol): Coloration is highly variable in the species.
The holotype (Figs. 4, 6, 7) is dark dorsally with some mottling and irregu-
larity of pattern. Dorsal parts of the trunk and head are medium brown with
some darker spots on the dorsolateral surfaces of the costal folds. A slight
indication of a mid-dorsal band is present over the shoulders. The tail has a
lighter ground color with a broken, irregular line of dark pigment on its lateral
surfaces. Tips of the nasolabial protuberances are whitish. The trunk venter
is basically light, probably yellowish or whitish in life, with a coarse reticulum
of brown laterally. Midventrally the reticulum becomes dense and forms a
rather large patch. The throat is only lightly marked with dark pigment. The
tail venter is light with some coarse reticulations of dark pigment. The limbs
are dark dorsally and light, coarsely marked with dark brown, ventrally.

The general dorsal color pattern of the holotype is also encountered in six
paratypic adults. These differ from the holotype in having a very fine speckling
of black dorsally. The tail is lighter than the back in four of the six.

One large paratype is almost uniformly gray-brown with no light colora-
tion, even on the tail.

Dorsal bands are present in sixteen paratypes. In its highest state of
development the band is a solid tannish-orange with a small, inverted, dark
triangle behind the eyes and a few black spots in the midline (Fig. 8). Distinct
bands are evident in eight individuals, but varying numbers of black specks,
spots, and blotches have an obscuring effect in most. The band is obscure in

14

Contributions in Science

No. 175

Table II

Measurements and Data for Bolitoglossa hartwegi and B. stuarti

Sex

Snout-V ent
Length

Axilla-Groin

Length

Head Width

Hind Limb
Length

Fore Limb
Length

Tail Length

Numbers of
Maxillary Tee

Numbers of
Vomerine Tet

Limb Interval

Foot Width

B. hartwegi

MVZ 57115

$

46.9

25.3

7.9

12.0

11.8

34.2 48

19

0.5

5.1

MVZ 57737

$

45.4

22.8

8.1

11.7

11.7

* 67

22

1

5.1

MVZ 57098

$

41.8

21.1

7.2

12.0

11.3

* 51

20

0

5.7

LACM 40601

$

40.1

20.2

7.2

10.7

10.1

* 47

20

1

4.3

LACM 40606

$

39.8

21.8

6.9

11.2

10.5

20.9 58

37

0

5.2

LACM 40607

$

35.9

17.2

6.4

9.2

8.7

* 45

25

0

4.0

TCWC 21399

$

34.0

17.6

6.1

8.9

8.8

23.6 37

19

0.5

4.0

MVZ 57099

9

53.6

29.5

7.9

12.8

11.3

31.3 66

27

2

5.7

UAZ 20262

9

52.3

28.9

8.7

13.0

12.1

* 74

25

1

5.3

MVZ 57108

9

52.0

28.2

8.6

13.2

12.6

37.2 77

23

0.5

6.4

LACM 44209

9

49.7

28.2

8.2

11.8

11.3

35.5 66

20

1.5

5.9

LACM 40605

9

48.5

26.5

7.6

12.0

11.7

33.0 69

34

1

5.3

MVZ 57109

9

48.0

26.6

7.3

12.2

11.9

33.2 76

33

1

5.9

LACM 40602

9

46.0

25.0

8.2

11.2

10.2

— -* 52

23

—

4.9

UMMZ 121557**

9

43.2

22.3

7.8

11.4

10.3

31.0 53

19

1

4.8

JFC-65-130

9

41.3

21.9

7.2

11.4

10.2

30.0 47

41

1

4.8

MVZ 57112

9

39.0

20.0

6.7

10.2

9.4

26.9 47

30

1

4.1

MVZ 57738

9

38.3

19.6

6.8

9.3

8.3

* 24

24

1.5

4.0

MVZ 57111

9

34.2

16.6

5.7

8.3

8.1

22.6 21

16

1

4.0

LACM 40604

9

30.4

15.8

6.4

8.2

8.0

20.2 32

21

0

4.0

B. stuarti

UMMZ 123203**

9

60.0

32.1

8.9

15.5

15.6

37.2 85

40

0.5

6.2

LACM 44210

9

35.0

18.8

5.9

8.7

8.6

20.1 20

24

1.5

3.5

* regenerated tails
** holotype

the remainder, but is more evident than in the holotype. The tail is generally
a little lighter than the trunk, except in individuals with well developed dorsal
bands.

Most juveniles have indications of a light, dorsal band, and in these the
tails are distinctly lighter than the body. In some juveniles, however, the dor-
sum of the trunk and tail is solid dark brown.

The entire sample is basically light colored ventrally. Seven adults have

1969 Systematics and Evolution of Neotropical Salamanders

15

METRIC

Figure 6. Holotypes of Bolitoglossa hartwegi (UMMZ 121557, top) and Bolitoglossa
stuarti (UMMZ 123203, bottom), dorsal view.

Figure 7. Holotypes of Bolitoglossa stuarti (UMMZ 123203, top) and Bolitoglossa
hartwegi (UMMZ 121557, bottom), ventral view.

16

Contributions in Science

No. 175

Figure 8. Variation in dorsal color pattern of paratypes of Bolitoglossa hartwegi.
From left to right, MVZ 57109, MVZ 57108, MVZ 57737, MVZ 57099, MVZ 57738.

a coarse suffusion of darker pigment, generally in the form of punctate melan-
ophores, which is most prominent midventrally. In other individuals the
reticulum of dark pigment extends over most of the venter. In the most heavily
pigmented individuals, some guanophores may overlie the melanophore punc-
tations. Ventrolateral aggregations of guanophores are generally distributed
throughout the sample.

Juveniles are mostly whitish ventrally, with some reticulations of darker
pigment.

Osteology : Information concerning the osteology is derived from one
cleared and stained specimen and from stereo radiographs of the holotype and
several paratypes. For comparative purposes, we have examined cleared and
stained specimens of eight B. rufescens, one B. occidentalis, two B. helmrichi,
one B. flavimembris, four B. morio, four B. rostrata , two B. engelhardti, one
B. dunni and one B. franklini. Radiographs of these species and of B. stuarti
and B. cuchumatana have provided some information. Skeletons of many addi-
tional species of Bolitoglossa have been available and a comparative osteo-
logical analysis will be presented at a later date. For the purposes of this paper,
comparisons of B. hartwegi will be limited to members of the helmrichi group
and to other possibly related species, notably B. morio and members of the
rufescens group.

1969 Systematics and Evolution of Neotropical Salamanders

17

Figure 9. Left, dorsal view of right foot of holotype of Bolitoglossci hartwegi (UMMZ
121557) ; right. Figure 10, ventral view of right foot of holotype of Bolitoglossci hart-
wegi (UMMZ 121557).

Bolitoglossa hartwegi resembles the more generalized members of the
genus in many features of its skeleton, especially in its well developed skull,
but differs in other regards, especially hand and foot structure. The skull is
solid and the bones are, in general, well articulated. In general morphology
the skull is most similar to skulls of B. helmrichi, B. flavimembris and B. morio,
and initial comparisons will be mainly among these four species. In all, the
anterior cranial elements are well articulated. The nasals typically are mod-
erately large and articulate with the facial lobe of the maxillaries and the
facial expansion of the frontals. In B. hartwegi the nasal articulates with the
frontal processes of the premaxillary and with the prefrontals, and this occurs
in at least some of the other species as well. The nasals are generally protrusive,
but to a lesser extent than in B. rufescens and B. occidentalis. In B. morio,
nasals are sexually dimorphic, being large and protrusive in males, but quite
small in females. Members of the hartwegi group may prove to be similarly
dimorphic. Frontal processes of the premaxillary are usually somewhat dilated
distally in the four species, and the tips are solidly articulated with the facial
parts of the frontals. The processes are unique in B. hartwegi in being greatly

PMX

18

Contributions in Science

No. 175

£

E

r!

x>

* c»

e .2
£ —

<u

S-, -C
Oh .2
- J2

X £

1969 Systematics and Evolution of Neotropical Salamanders

19

expanded distally, with many erratic bony processes (Fig. 11). The processes
are larger in B. hartwegi than in B. helmrichi, B. flavimembris, and B. morio.
Prefrontals are small but discrete in B. hartwegi, B. helmrichi and B. flavi-
membris, but are larger in B. morio. The elements in B. morio are somewhat
smaller than in such generalized species as B. rostrata. The main articulation of
the prefrontals in all of these species, and the only articulation in some, is with
the frontals. The dorsal part of the maxillaries, the anteromedial part of the
prefrontals, and the lateral part of the nasals are notched in varying degrees
for passage of the nasolacrimal duct, with the maxillary notching usually
the greatest. The maxillaries are long and relatively stout with well developed
facial and palatal portions. At the posterior end of the palatal portion, a lobe
is formed which overlaps the lateral margin of the vomerine body, forming
the major articulation of the two bones. Vomers are well developed in the
species, but are usually well separated. In B. hartwegi the intervomerine
fontanelle is broad, but is slightly narrowed anteriorly by inward growth of the
vomerine bodies. Posteriorly the raised dental ridge of the preorbital process
extends medially, beyond the margins of the vomerine body, for a short distance
as a spinous process (Fig. 11). This process is a characteristic feature of many
of the derived species of Bolitoglossa and is especially well developed in B.
helmrichi, where it is long and arched. The processes are barely evident in
B. morio, B. flavimembris, and in such generalized species as B. dunni and
B. rostrata. Laterally the preorbital processes are long, and in B. helmrichi
they extend beyond the vomerine body. The facial portions of the frontals are
moderately well developed in these species, and the areas are especially large
in B. morio and B. flavimembris. The frontals are closely articulated with one
another at the anterior end of the interorbital region, but in most individuals
the bones are not typically in contact posterior to that point. In B. hartwegi,
however, the bones are articulated for their entire lengths. The frontals are
usually articulated with the parietals by means of a lateral lobe of each frontal
that overlaps each parietal. In B. hartwegi this articulation is particularly strong
and is augmented by a medial lobe of the parietal which overlaps the frontal.
All four species have a well developed, small, lateral parietal spur, which is
characteristically present in the tribe Bolitoglossini (Wake, 1966). Small,
laterally oriented crests are present on the otic capsule in the four species, but
they are quite variable in four specimens of B. morio and are probably variable
in the others also. In all four species the parasphenoid is a large bone with a
moderately broad anterior tip. The orbitosphenoids are well separated at this
point. Patches of posterior vomerine teeth borne on the parasphenoid approach
the midline but are not in contact. The right patch bears 66 teeth, the left, 72,
in the single B. hartwegi. In none of the species is a stilus present in even rudi-
mentary form on the operculum. Squamosals and quadrates are small. The
dentary is relatively weak and the prearticular is low and relatively small. The

20

Contributions in Science

No. 175

hyobranchial apparatus (Fig. 12) has a form that is rather typical of general-
ized members of the genus.

The skulls of B. occidentalis and B. rufescens differ from that of B. hart-
wegi in being much weaker (see also Hansen and Tanner, 1958). The bones
of the facial part of the skull are poorly articulated, if at all. The premaxillary
is small with very slender, slightly dilated frontal processes. The maxillaries are
very slender and are toothless in B. rufescens and bear a few small teeth in B.
occidentalis. Palatal parts of the premaxillary are nearly absent and those of
the maxillary are most prominantly represented by the lobe that forms the
articulation with the vomer. The nasals are strongly protrusive, extending far
anterior to the premaxillaries above the bulging nasal capsules. Prefrontals are
usually absent, and the facial parts of the frontals are greatly reduced. The
vomers are very greatly separated and are small, but the preorbital processes
are long and sinuous, extending beyond the body of the vomer both laterally
and medially. In some large B. rufescens, articulations between the nasals and
frontals are established, but this is not typical. The frontals and parietals are
separated medially and the only articulation between the two bones is laterally,
by means of the prominant frontal lobes. The parasphenoid is reduced anter-
iorly and the orbitosphenoids are nearly in contact at this point. Otic crests
and opercular stylar processes are absent. All skull elements are rather weakly
developed.

The vertebrae of B. hartwegi are of the generalized Bolitoglossa type, with
spool-shaped centra and unmineralized intervertebral cartilages. No basapo-
physes are present. The column is comprised of 1 cervical, 14 trunk or dorsal,
1 sacral, 2 caudosacral, and 26 caudal vertebrae. The holotype has only 12
trunk vertebrae; the sixth is abnormal and bears three pairs of ribs. Despite the
anomalous vertebral reduction, the number of costal grooves is 13, as in all
members of the genus. The holotype has 23 caudal vertebrae, MVZ 51099 has
24, MVZ 57109 has 27, and LACM 44029 has 23. Low numbers of caudal
vertebrae are found in the rufescens group and in other members of the helm-
richi group as well. Information for the various species follows: B. occidentalis,
23, 27, B. helmrichi, 26, 28, 32; B. flavimembris, 21, 22; B. cuchumatana, 24,
26; B. rufescens, 20 (1), 21 (4), 22 (2), 28 (1); B. stuarti, 20, 25. Related
members of the rostrata group have numbers of tail vertebrae in the high por-
tions of the ranges of these species, and often have more. Members of the pre-
dominantly lowland platydactyla group have as many as 40 caudal vertebrae.

Transverse processes are borne on the first 17 caudal vertebrae. The
processes diminish in size in a nearly progressive pattern from the base toward
the tip of the tail. The first and second caudal vertebrae lack hypophysial keels,
and the third has a moderate- sized keel that is not free dorsoventrally as in
some related species ( cf . B. morio, Wake and Dresner, 1967, Fig. 4). Hypo-
physial keels are prominent on all but the last three vertebrae.

The first caudal vertebra of B. hartwegi has large and complex transverse

1969 Systematics and Evolution of Neotropical Salamanders

21

Figure 12. Hyobranchial apparatus of Bolitoglossa hartwegi.

processes that are bifurcated distally. The anterior prong is the dominant one
and it extends sharply anterolaterally, well ahead of the anterior end of the
centrum. The smaller and slenderer posterior prong extends almost directly
lateral from the base. This pattern is broadly distributed in the Bolitoglossa
of northern Middle America except in the platydactyla group and a few addi-
tional species (see below for discussion). Species with this condition also tend
to have rudimentary to well developed tibial spurs, structures generally lacking
in other species of the genus. In B. hartwegi the tibial spur is a well developed
tibial crest, and the free distal portion is very short. Distinct tibial spurs are
also present in all members of the rufescens and helmrichi groups that we have
been able to examine directly, and they are visible in radiographs of B. cuchu-
matana and B. stuarti.

Hands and feet of B. hartwegi are large and are characterized by stout,
flattened metacarpals and metatarsals, fusions of the third and fourth distal
carpals and the third, fourth and fifth distal tarsals, and marked reduction in
the more distal phalangeal elements (Fig. 13). These features, found also in
members of the rufescens group, are discussed more fully below.

Habitat : Information concerning the habitat of B. hartwegi has been ob-
tained from the field notes of Floyd L. Downs deposited at the University
of Michigan Museum of Zoology, of William J. Riemer deposited at the
Museum of Vertebrate Zoology, University of California, and of Dennis E.
Breedlove, University of California, Berkeley, provided through the courtesy
of Joseph F. Copp. The holotype was taken under a log in a pine-oak woodland.
All specimens collected by Dr. Breedlove (LACM 40601-07) were taken on
the ground, either in crevices of moss covered rocks, under moss on rocks, or

22

Contributions in Science

No. 175

in fossiliferous shale. The animals were found in areas where many seepages
were present, and all but one were found in cloud forest. The vegetation of the
cloud forest on Cerro Zontehuitz includes the following as dominants, accord-
ing to Dr. Breedlove: Drimys granadensis var. mexicana, Oreopanax capitatus,
O. xalapensis, Per sea fioccosa, Ternstroemia tepezapote, Quercus acatenan-
gensis, Miconia glaberrima, Magnolia sharpii, and Podocarpus oleifolius.
LACM 40602, collected at a considerably lower elevation, was found in a
fossiliferous shale outcrop on a northeast slope in an area with a cover of shrubs
and bunch grasses. The dominant plants, identified by Dr. Breedlove, were:
Myrica cerifera, Muhlenbergia macroura, Garrya laurifolia, Coriaria thymi-
folia, Rondeletia sp., and Fuchsia arborescens.

Specimens collected by W. J. Riemer were taken under and in rotten logs
and under the bark of standing stumps up to 3 feet above the ground. One was
taken from a bromeliad that was attached to a tree about 10 feet above the
ground. Dr. Riemer noted that all of his specimens (MVZ numbers) were
taken from cover that was found in relatively open, sunny places in the rather
heavy forest.

Range : Bolitoglossa hartwegi occurs in the mountainous regions of east-
central Chiapas (Fig. 3).

We take pleasure in naming the second undescribed species for Professor
L. C. Stuart of the University of Michigan, the foremost student of Guatemalan
amphibians and reptiles.

Bolitoglossa stuarti new species
Figures 6,1, 14, 15

Holotype : UMMZ 123203, an adult female from 7.8 mi (by road) SE of
Ciudad Cuauhtemoc (El Ocotal), 8, Chiapas, Mexico, in Huehuetenango,
Guatemala, at about 950 m (3120 feet) elevation, collected by T. J. Cohn,
September, 1961.

Paratype : LACM 44210, a young female from 1.4 mi (by road) S of La
Trinitaria, 17, Chiapas, Mexico, at 1615 m (5300 feet) elevation, collected by
Dennis R. Paulson, July 14, 1965. The specimen was collected from an area
of dry forest.

Diagnosis : A moderately large species (to 60.0 mm SL) with high num-
bers of teeth (to 85 maxillary and 40 vomerine); distinguished from B. flavi-
membris, B. helmrichi and B. cuchumatana by extensively, nearly completely
webbed hands and feet, reduced phalangeal elements (especially terminals),
and absence of subterminal pads; distinguished from B. hartwegi by larger
adult size, narrower head, more extensively webbed hands and feet, shorter
digits, and higher numbers of maxillary teeth in adults. Distinguished from
species of the rufescens group by much larger size, proportionately longer
limbs, larger hands and feet, and many more maxillary teeth.

1969 Systematics and Evolution of Neotropical Salamanders

23

Figure 13. Hands and feet of Bolitoglossa. Cartilage stippled. Cutaneous outline indi-
cated. Drawn through use of microprojector. A. Right hand of B. hartwegr, B. Right
foot of B. hartwegi, C. Right foot of B. rostrata, D. Right foot of B. rufescens.

Description : This moderately large species is robust, relative to such
forms as B. helmrichi, and has a rather short, moderately truncate snout. The
nostrils are small. Labial protuberances of the nasolabial groove are small. The
canthus rostralis is long and rounded. The head is relatively narrow (SL 6.7
times head width in adult holotype). A distinct groove extends below the eye,
following its curvature, but does not communicate with the lip. The eyes are
of moderate size and only slightly protuberant; they extend only slightly beyond
the lateral margins of the head. A well defined postorbital groove extends

24

Contributions in Science

No. 175

Figure 14. Head of holotype of Bolitoglossa stuarti (UMMZ 123203).

posteriorly from the eye as a shallow depression for 3.1 mm (in holotype) , then
proceeds sharply ventrally at the level of the posterior end of the mandible and
extends across the gular area as an indefinite depression parallel to and 5.2 mm
(in holotype) anterior to the sharply defined gular fold. Vomerine teeth occur
in long, slightly patched series that extend beyond the lateral borders of the
internal nares in the adult holotype; large numbers are present (24 and 40 in
the paratype and holotype, respectively). Maxillary teeth are numerous in the
holotype (85), and they extend posteriorly to a point about four-fifths through
the eye. The premaxillary teeth do not pierce the lip. The tails are especially
short (0.58 and 0.62 times SL in the paratype and holotype, respectively).
Tails are rounded and strongly constricted at the base. Postiliac glands are
conspicuous. Limbs are slender and long; limb interval is 0.5 in the holotype
and 1.5 in the paratype. The hands and feet are characterized by great width
relative to length. The digits are short and pointed. Webbing of hands and feet
is so extensive as to be nearly complete, with only the tips of the longer digits
protruding from the webbing (Fig. 15). The digits are discrete only where
their tips protrude from the web, and there are no subterminal pads. The
fingers are, in order of decreasing length, 3, 4, 2, 1 ; toes, in order of decreasing
length, 3, 4, 2, 5, 1.

Measurements of Holotype (in mm) : Head width 8.9; snout to gular fold
(head length) 14.7; head depth at posterior angle of jaw 4.4; eyelid length 4.2;
eyelid width 2.2; anterior rim of orbit to snout 5.0; horizontal orbital diameter
3.3; interorbital distance 3.1; distance between vomerine teeth and parasphe-
noid tooth patch 0.8; snout to fore limb 18.9; distance separating internal nares
2.8; distance separating external nares 3.2; snout projection beyond mandible
1.1; snout to posterior angle of vent (SL) 60.0; snout to anterior angle of vent
54.8; axilla to groin 32.1; tail length at 37.2; tail width at base 4.3; tail depth at
base 4.4; fore limb length 15.6; hind limb length 15.5; width of right hand 4.4;
width of right foot 6.2.

Coloration (in alcohol) : The holotype is generally dark dorsally and light

1969 Systematics and Evolution of Neotropical Salamanders

25

ventrally. A broad dorsal band of light pigment extends from in front of the
eyes nearly to the tip of the tail. The ground color of the band varies from a dull
tan to a bright yellowish-tan. It is heavily mottled with dark brownish spots and
blotches in the trunk region, but is bright and relatively unmarked above the
tail base. Lines of dark pigment extend from behind the eyes and converge in
the mid-dorsal area of the neck. The lateral surfaces of the trunk and tail are
covered with an area of dark pigment that extends from the posterior margin of
the eye almost to the tail tip. This pigment is lightest on the sides of the trunk
and darkest along the tail. It is sharply demarcated from the dorsal band along
both trunk and tail. Ventrolaterally on both trunk and tail the pigment lightens.
All ventral surfaces are very light but have a more or less even suffusion of
punctate melanophores. Melanophores are least numerous on the very light
throat. All ventral surfaces are covered with large, scattered guanophores.

The tip of the snout is dark, but the tips of the nasolabial protuberances
are light. Dorsal surfaces of the limbs are mottled with dark brown and light
tannish pigment, but ventral surfaces are very light gray.

The juvenile paratype is lighter in general coloration than the holotype.
The dorsal surfaces of the head and trunk are light reddish brown, becoming
progressively lighter laterally and ventrally. Laterally the reddish element of
the dorsal color disappears and the melanic reticulum is converted to isolated,
punctate melanophores which are intermixed with small guanophores. This
light “salt and pepper” effect is characteristic of the entire ventral surface, but,
because the unpigmented flesh is exposed, the dominant coloration, especially
on the tail venter, is a light yellow. Bright yellowish pigment lies above the base
of the tail, and extends to near the tail tip, as in the holotype.

Osteology : Information has been derived from a series of stereo radio-
graphs of the two known specimens. The skull has an unpaired premaxillary
with broadly separated, slender frontal processes. The nasals are large and
protrusive, and bear medially directed bony processes that in part overlap the
frontal processes of the premaxillary. Moderate-sized, elongate-ovoid pre-
frontals are present. Vomers are widely separated medially, and small, medi-
ally directed processes carry the teeth medially, beyond the vomer body. The
facial parts of the frontals are moderately well developed. Frontals and
parietals are not sutured to their paired elements medially. The articulation be-
tween frontals and parietals is exclusively lateral. No stilus is present on the
operculum. The vertebral column includes 1 cervical, 14 trunk, 1 sacral, 2
caudosacral and 20 (paratype) to 25 (holotype) caudal vertebrae. The last
two trunk vertebrae lack ribs. Elsewhere in the helmrichi group, only the last
trunk vertebra lacks ribs, except in three of six B. hartwegi, which are like
B. stuarti. The arrangement of the vertebrae in the tail base region is unusual
in the two specimens of B. stuarti. The last caudosacral and the first caudal
vertebrae have extremely short centra, and the vertebrae in this region appear
to be “crowded” together. The combined length of the centra of these two

26

Contributions in Science

No. 175

vertebrae is just slightly longer than the length of the centrum of the second
caudal vertebra. The transverse processes of the first caudal vertebrae are
stocky but are distinctly bifurcated. The anterior prong of each process extends
far forward, below the transverse process of the last caudosacral vertebra, and
may extend as far as the posterior margin of the first caudosacral vertebra.
This general pattern is typical of the helmrichi group, especially B. hartwegi,
but the “crowding” is more extreme in B. stuarti than in the others. A distinct,
stout tibial spur is present in the holotype, but it appears to be attached to the
tibia for most of its length. Metacarpals and metatarsals are flat and broad
(Fig. 15). Phalanges are highly variable in shape, but are greatly reduced in
size distally. Proximal phalanges are usually stout, rectangular, flattened plates,
but terminal phalanges are reduced to tiny dots of bone. The phalangeal
formulas are 1-2-3-2 in the hand and 1-2-3-3-2 (1-2-3-2-2 on one side) in the
foot of the holotype. The counts cannot be made for the feet of the paratype,
but one hand has a formula of 1 -2-3-2.

Remarks : Bolitoglossa stuarti is apparently a larger species than B. hart-
wegi, its close relative and close geographic neighbor. Although we have only
two specimens, the holotype is larger than any B. hartwegi by over 6 mm,
and the paratype is a juvenile at a size at which B. hartwegi is sexually mature.

Range : Known from the type locality and from the nearby La Trinitaria
region along the Mexican-Guatemalan border at moderate elevations (950-
1600 m) (Fig. 3).

Evolution of the HELMRICHI Group

Relationships : The five species of the helmrichi group may be further
grouped into three assemblages: B. flavimembris ; B. helmrichi and B. cuchu-
matana', B. hartwegi and B. stuarti.

The most primitive may be B. flavimembris, which is the largest of the
species and has the least specialized hands and feet. Digits are well developed
and discrete, and subterminal pads are well developed. The species has a broad
head. These are features characteristic of other genera and we consider them
primitive within the group.

Somewhat intermediate between the primitive B. flavimembris and the
more derived B. hartwegi and B. stuarti are B. cuchumatana and B. helmrichi.
They are smaller than B. flavimembris and B. stuarti, and do not get as large
as the larger B. hartwegi in our sample. The two species have similarly shaped
hands and feet, with more webbing than B. flavimembris, but with more dis-
crete digits and less webbing than B. hartwegi and B. stuarti. Both species have
subterminal pads. Many maxillary and vomerine teeth are present in B. helm-
richi, fewer are present in B. flavimembris, and dentition is reduced even fur-
ther in B. cuchumatana, which has less teeth than any other member of the
group. Heads are broader in B. flavimembris than in B. cuchumatana, and
B. helmrichi is somewhat intermediate.

1969 Systematics and Evolution of Neotropical Salamanders

27

The most derived species of the helmrichi group are B. hartwegi and
B. stuarti. While B. stuarti has such possibly primitive features as large size and
high numbers of maxillary teeth, the hands and feet are highly specialized with
extensive webbing and the digits and phalanges are reduced in size. The smaller
B. hartwegi also has many teeth, and its hands and feet are highly specialized
in having extensive webbing, phalangeal reductions and mesopodial fusions.
Subterminal pads are absent in both B. stuarti and B. hartwegi, and the tips of
fingers and toes are characteristically pointed. Bolitoglossa hartwegi has longer
digits, more phalangeal bone, and less webbing than B. stuarti. Bolitoglossa
hartwegi, B. stuarti, B. helmrichi, and B. cuchumatana all have generally simi-
lar coloration, although there is much variation. All four commonly have a
dorsal light band which tends to become obscured somewhat medially and
posteriorly. As a result, a pair of dorsolateral light stripes, extending from
behind the eyes over the shoulders and onto the back, are frequently present.
All members of the helmrichi group have light ventral coloration, especially on
the tail.

Relationships to species outside the group are somewhat difficult to assess,
but it is likely that the helmrichi group is closely related to two other groups.
The pattern of relationships is such that one is tempted to suggest that a “slice
of time” has been preserved in the Guatemalan and Chiapan highlands.
Bolitoglossa flavimembris is a nearly perfect link between the helmrichi group
and the more primitive rostrata group, restricted to the highlands, and B. hart-
wegi is a nearly perfect link between the helmrichi group and the more derived
rufescens group of the lowlands. Together the three groups form what we think
is a natural assemblage with a common ancestor. All members of the three
groups share a derived character not widely distributed elsewhere in the genus,
a unique configuration of the transverse processes on the last caudosacral and
first caudal vertebrae related to tail autotomy (Wake and Dresner, 1967). The
processes on the last caudosacral vertebrae are usually nearly perpendicular
to the body axis, or slightly bent posteriorly, and the processes on the first
caudal vertebrae are sharply directed anteriorly and are distally bifurcated.
This pattern is related to the sharply constricted region at the base of the tail,
which is more marked in these groups than in other members of the genus.
Correlated with this feature are several other characters, such as the tendency
for members of the group to have tibial crests or free tibial spurs. We hope to
present analyses of these features in future publications.

The rostrata group is, in our view, a large one, the members of which are
large, with somewhat longer tails, less webbing of hands and feet, and longer
digits than members of the helmrichi group. We include the following species:
B. rostrata, B. engelhardti, B. nigroflavescens, B. dunni, B. franklini, B. morio,
B. macrinii, B. riletti, B. omniumsanctorum, B. lincolni, B. resplendens, and
B. brevipes. This group includes the genotype of Magnadigita Taylor {nigro-
flavescens). Stuart (1952) placed B. dunni, B. engelhardti, B. cuchumatana,

28

Contributions in Science

No. 175

B. helmrichi and questionably B. robusta, B. subpabnata and B. macrinii in a
dunni group, said to include bromeliad dwellers with large, protruding eyes, a
single phalanx of the longest toe free of a web, flattened rostra, and streaks or
stripes of red or brown coloration. His franklini group included B. franklini ,
B. lincolni, and B. nigrofiavescens, robust species of moderate size with rounded
snouts and moderate sized eyes. These species were said to have two phalanges
of the longest toe free from the web and to be brightly colored with red and
yellow, either in spots or stripes. They are bromeliad dwellers. Salamanders of
his morio group, including B. morio and B. omniumsanctorum, are robust
species of moderate size with rounded rostra. Eyes are smaller and less pro-
truding than in the dunni group, and two phalanges were said to be free from
the web. They are mottled brown and white, and flecked with white. Both
species are ground dwellers. Though similar to the dunni group in most re-
spects, B. rostrata was excluded because of its ground dwelling habits. The
terrestrial B. flavimembris was assigned provisionally to the franklini group,
but its different ecology and foot structure were noted. What we have done is
to remove B. flavimembris, B. helmrichi, and B. cuchumatana to our helmrichi
group, remove B. robusta and B. subpalmata to other groups, and place all
other species discussed by Stuart in our rostrata group. In addition, we have
added B. riletti, B. brevipes, and B. resplendens, species described since 1952,
to the rostrata group. Certainly the rostrata group could be further subdivided,
but we are unable to do so at the present time. It is apparent that B. lincolni,
B. resplendens and B. brevipes are close relatives, B. franklini and B. nigro-
flavescens are close relatives, and B. macrinii and B. riletti appear to be close
relatives; B. riletti is smaller than other members of the rostrata group. Boli-
toglossa engelhardti is close to the helmrichi group, and it is also rather similar
to B. rostrata in many features. With its very large head and long limbs, B.
dunni is in a relatively isolated position. We are unsure of the relationships of
the poorly known B. omniumsanctorum. We hope to discuss relationships in
and of the rostrata group more thoroughly in future publications.

The rufescens group includes B. rufescens, B. occidentalis, and B. bili-
neata (which we consider to be a synonym of B. occidentalis) . This is a group
that includes species that are smaller than members of the helmrichi group,
and have greatly reduced dentition, poorly developed skull bones, reduced
phalangeal elements, mesopodial fusions, extensively webbed and flattened
hands and feet with sharply pointed digital tips, and short tails.

Bolitoglossa flavimembris is the largest member of the helmrichi group
and has the least webbing of hands and feet. In these features it resembles
members of the rostrata group, yet its digits are more reduced and pointed and
it has more webbing than any member of that group.

Bolitoglossa hartwegi resembles the rufescens group in having extensive
hand and foot webbing, mesopodial fusions, and reduced phalangeal elements.

1969 Systematics and Evolution of Neotropical Salamanders

29

It differs sharply, however, in its larger size and its high numbers of marginal
teeth.

Taylor (1941, 1944) and Hansen and Tanner (1958) recognized that
B. rufescens and B. occidentalis differed considerably from the extensively
webbed species of the platydactyla group, and suggested that two natural
groups were represented. It is now apparent that the rufescens group is related
not to the platydactyla group but to the helmrichi group. Our new knowledge
of the helmrichi group demonstrates again the futility of separating species of
Bolitoglossa into two groups on the basis of foot structure. This tendency has
resulted in comparison of the rufescens group with the large platydactyla group
rather than with its true relative, B. helmrichi, which had been placed in a
different genus. Clear adaptive trends within groups of closely related species
span the extremes of foot types in Costa Rica (Wake and Brame, 1963b) and
South America (Brame and Wake 1963a) as well as in Nuclear Central
America (see also Wake and Brame, 1963a).

Several years ago we suggested that the genus Magnadigita Taylor could
no longer be recognized (Wake and Brame, 1963a). However, should it be
found that the assemblage that includes the rostrata, helmrichi, and rufescens
groups is a natural one that is sharply differentiated from other members of
Bolitoglossa (genotype mexicana) , the name Magnadigita is available for it.
The genus Bolitoglossa as currently recognized includes over 50 species, and
division of the genus into natural assemblages is desirable; we do not recom-
mend such action, however, until the group as a whole becomes better known.

Evolutionary Patterns and Trends: In the view of Taylor ( 1944), species
of Magnadigita had large, discrete digits and slight webbing, while species of
Bolitoglossa had indistinct digits and extensive webbing (see also Wake and
Brame, 1963a). A complete bridging of these two extremes is found in the
helmrichi group, and hand and foot structure in this group is considered in
detail in the following paragraphs.

In discussing degree of webbing we recognize five categories: a) un-
webbed; b) slightly webbed (well over 2 phalanges of the longest toe free of
web, and all toe tips free of web and bearing well defined subterminal pads) ;
c) moderately webbed (about 2 phalanges or less of longest toe free of web,
subterminal pads on the longest 2 to 3 toes); d) extensively webbed (no sub-
terminal pads, toe tips pointed, marked indentations between tips of toes);
e) completely webbed (little or no indentation between toe tips, only longest
toe tip pointed and obvious). No members of Bolitoglossa are unwebbed.
Bolitoglossa rostrata falls in category b, B. flavimembris, B. cuchumatana and
B. helmrichi in category c, and B. hartwegi in category d (Fig. 15) . Bolitoglossa
stuarti, B. rufescens, and B. occidentalis fall in category d and category e, with
some variation from individual to individual. The best examples of category e
are individuals of the Ecuadorian species B. sima (Brame and Wake, 1963a,

30

Contributions in Science

No. 175

Fig. 20 F) and the Panamanian species B. schizodactyla (Wake and Brame,
1966b, Fig. 2); the latter, however, lacks pointed digital tips.

Within the helmrichi group a clear trend in foot structure is evident, from
the moderately webbed foot with large, discrete digits and well developed
subterminal pads of B. flavimembris, through the increasingly flattened, more
fully webbed feet of B. cuchumatana and B. helmrichi, to the extensively
webbed feet with relatively indistinct digits found in B. hartwegi and in a more
extreme form in B. stuarti (Fig. 15). Not much ecological information is avail-
able for the group. The known altitudinal range is from 950 m (B. stuarti ) to
over 2800 m ( B . hartwegi ), and B. hartwegi is known to range over 800 m
altitudinally. We consider these elevations to be moderate to relatively high
for Bolitoglossa. Apparently all occur in areas where bromeliads are common,
and at least some individuals of all five species have been collected in brome-
liads. The B. flavimembris taken by Schmidt (1936) were collected under
stones and logs, but FMNH 142081 was collected from a bromelid in a wet,
densely forested ravine. Wake and John E. Woods collected over 50 specimens
of B. flavimembris in the vicinity of Finca La Lucha, above La Fraternidad,
Guatemala, in March and April, 1969. Some were taken in bromeliads, but the
majority were collected under ground cover and under the bark of logs. Stuart
(1943) found three of the four B. cuchumatana in bromeliads, and considered
B. helmrichi to be definitely a bromeliad inhabitant, stating (p. 16) “I have
never seen it on the ground.” Series of all five species with good ecological
information are badly needed, and only speculations may be made concerning
ecological correlations of morphological characters at this time. Members of
this group have a distinct tendency in the direction of increased webbing of
hands and feet, and this may be correlated to some degree with attainment of
arboreality by at least some of the species and movement into areas of some-
what lower elevation. Yet, the extensively webbed B. hartwegi has been taken
mainly in terrestrial situations at moderate to moderately high elevations, so
the correlation is not high. Members of the rostrata group with much less
webbing appear to be fully as arboreal as B. hartwegi ( e.g ., B. franklini, B.
nigroflavescens ) .

In Bolitoglossa, highland species tend to be terrestrial and to have the
least amount of webbing in the genus, while lowland species tend to be ar-
boreal and to have the greatest amount of webbing. This is a generalization and
there are some exceptions; a few species with extensive webbing inhabit brome-
liads in the highlands (e.g., B. hartwegi ). The slightly webbed B. franklini
occurs in bromeliads. In addition, some species with relatively little webbing,
such as B. subpalmata, are terrestrial in the highland portions of their range, but
are found in bromeliads at lower elevations. All lowland species have extensive
to complete webbing. Hands and feet are morphologically complex in Bolito-
glossa, and to discuss them strictly in terms of degree of webbing can be quite
misleading. They are functional complexes, modifications of which may involve

1969 Systematics and Evolution of Neotropical Salamanders

31

Figure 15. Outline of right feet of species of Bolitoglossa. Cutaneous outline, meta-
tarsals and phalanges indicated. Cartilage stippled. All drawn to same scale. Drawn
through the use of a microprojector from cleared and stained (C) or X-rayed (X)
specimens. A. B. rostrata (C), B. B. flavimembris (C), C. B. cuchumatana (X), D.
B. helmrichi (C), E. B. hartwegi, holotype (X), F. B. hartwegi (X), G. B. stuarti .
holotype (X). H. B. occidentalis (C). I. B. rufescens (C).

32

Contributions in Science

No. 175

the whole structure or only parts of it. In studying these features one must be
concerned with degree of webbing, but also with the shape of the foot, the
shape of the individual digits and their relative flattening, the characteristics of
the subterminal pads or their absence, the numbers and shapes of the pha-
langes, metacarpals and metatarsals, the numbers of tarsals and carpals and
their relationships, and proportional changes in the musculature and skeletal
morphology of the unit. We as yet have not had adequate samples to analyze
all of these features, but we can present some information for the hehnrichi
group and related species.

In foot structure as well as in habitat and elevational range, members of
the hehnrichi group generally fall between highland and lowland relatives. We
have compared the five species with B. rostrata, B. engelharclti, B. dunni, B.
morio, B. nigroflavescens and B. franklini, highland forms with discrete digits,
subterminal pads, and slight webbing, and B. occidentalis and B. rufescens,
lowland species with padlike feet, indistinct, flattened digits, no subterminal
pads, and extensive webbing. Within the hehnrichi group, the feet of B. flavi-
membris are nearest those of the highland relatives in most features, while
those of B. stuarti and B. hartwegi are closest to those of the lowland relatives.
Feet of B. hehnrichi and B. cuchumatana are somewhat intermediate in certain
features, but very distinctive in others (Figs. 15, 16). The hehnrichi group also
bridges the gap between the highland and lowland species in carpal and tarsal
structure. Highland species are like most species of Bolitoglossa in having eight
carpal and tarsal elements, while B. rufescens and B. occidentalis have seven
elements in both carpus and tarsus. B. flavimembris has eight, B. hehnrichi has
eight in six of the eight carpi and tarsi examined and seven in one carpus and
one tarsus, and B. hartwegi has seven elements in both carpus and tarsus. In all
species the reductions are the result of fusions rather than losses.

Toe tips are bluntly rounded in the highland and more terrestrial species,
and they bear well developed pads on the ventral surface of the terminal parts
of the longest two to four digits. In the helmrichi group the toes have rounded
or triangular tips. Subterminal pads are present in all members of the rostrata
group and in B. flavimembris, B. hehnrichi, and B. cuchumatana, but are ab-
sent in B. hartwegi and B. stuarti, and all members of the rufescens group.

In skeletal detail the feet of the members of the helmrichi group are also
intermediate between the highland and lowland assemblages. The fourth digit
of the hind foot has three phalanges in most Bolitoglossa, but in B. rufescens
and B. occidentalis only two are found, apparently as a result of the fusion of
two smaller elements (see also Hansen and Tanner, 1958). The members of
the hehnrichi group have three with the exception of B. hartwegi (three in
three animals, three on one side and two on the other in another animal, and
two on both sides in a fifth animal) and B. stuarti (three on one side and two on
the other in the holotype). Terminal phalanges are large and distally expanded
in generalized Bolitoglossa, including also B. flavimembris, B. cuchumatana,

1969 Systematics and Evolution of Neotropical Salamanders

33

and B. helmrichi, but the elements are small and pointed or erratically shaped
in B. hartwegi, B. stuarti, B. occidentalis and B. rufescens (Fig. 15). Lengths
and widths of all phalangeal elements in the osteological preparations and
radiographs available to us have been enlarged by projection and measured,
and some preliminary results are portrayed in Fig. 16. In terms of the lengths
of digital elements, the sum of the phalangeal length is greater than that of the
metatarsal length in B. franklini, B. rostrata, and B. morio, but less in the
helmrichi group and in B. rufescens and B. occidentalis. B. flavimembris may
be somewhat intermediate, but we do not have adequate samples to confirm
this observation. The sum of the lengths of the terminal phalanges is greater
than 50 per cent of the sum of the lengths of the remaining phalanges in
B. franklini, B. rostrata, B. morio, B. helmrichi, B. cuchumatana, and B. flavi-
membris, but less than 50 per cent in B. rufescens, B. occidentalis, B. hart-
wegi and B. stuarti.

It is evident from Figures 15 and 16 that B. helmrichi and B. cuchumatana
are quite distinct from the other species in terms of phalangeal proportions, and
that B. flavimembris is the closest to a link between them, the rostrata group,
and the other members of the helmrichi group. B. cuchumatana tends to have
slightly less webbing and slightly better developed phalangeal elements than
does B. helmrichi. The second phalanx of the fourth toe is extremely small in
B. helmrichi, but the corresponding bone is longer than either the terminal or
first phalanx in B. cuchumatana. In both species the terminal phalanges are
expanded distally, thus resembling more generalized members of the genus.

Bolitoglossa stuarti and B. hartwegi are close to B. rufescens and B. occi-
dentalis on Figure 16. In B. stuarti and B. hartwegi the phalanges, metacarpals
and metatarsals are flattened and bear bony laminae, similar to those found in
B. rufescens and B. occidentalis. Only slight indications of these laminae are
seen in other species (Figs. 13, 15). An indication of the flattening of these
elements and the growth of bone into the web is that the width of the fifth
metatarsal is more than half its length in B. hartwegi, B. stuarti , B. occidentalis
and B. rufescens, but less than half in the other species. Marked phalangeal re-
duction is found in these four species, with tendencies to lose a phalanx in the
fourth toe. In addition, the terminal phalanges are greatly reduced and are
pointed, in contrast to the situation in more generalized species.

As a result of the phalangeal reductions, carpal and tarsal fusions, and
webbing increase in these four species, hand and foot pads result that are
resilient structures, adapted for functioning as a unit. Independent digit action
is reduced in favor of the total hand or foot action associated with movement of
these organisms across wet, smooth surfaces in arboreal situations. The flat-
tened pad and unitary functioning make locomotion by surface tension phe-
nomena possible. Observations on the extensively webbed feet of living B.
yucatana, B. adspersa and B. rufescens reveal that during locomotion the hands
and feet are rolled flat onto the surface, proceeding from the ankle or wrist

34

Contributions in Science

No. 175

distally in a rather slow and deliberate action of the entire pad, assuring that
complete contact with the underlying surface is achieved.

Figure 15 illustrates a kind of morphocline in regard to foot structure in
the helmrichi group and its relatives. While it is not likely that this represents
a chronocline or a true phylogeny, it is quite possible that there are preserved
in this series of species, morphological stages representative of ancestral con-
ditions of some of the more derived types of foot structure. The discovery of
B. hartwegi and B. stuarti fills an important gap in our knowledge of this genus
and demonstrates both the pattern and manner in which extensively webbed
forms are derived from more generalized types. In addition, it provides in-
sights into factors associated with invasion and success in tropical lowland
environments. We suggest that lowland invasions became possible with the
evolution in the highlands of such species as B. hartwegi, forms adapted for
life in both terrestrial and arboreal habitats and having the fully webbed feet
associated with exclusively arboreal life in more specialized members of the
genus. With the evolution of arboreality, movements into the heavily populated
tropical lowlands became possible. All members of Bolitoglossa in the tropical
lowland fauna are extensively to completely webbed, arboreal species, and
it seems likely that terrestrial life in tropical lowlands for this genus is pre-
cluded. One reason may be competition with the numerous lowland species of
lizards, frogs and snakes which overlap niche requirements of more generalized
Bolitoglossa. In the case of Bolitoglossa, invasion of tropical lowlands has been
accomplished several times in different geographic areas, always, to our knowl-
edge, by specialized rather than generalized species. Specialization associated
with arboreality has provided, for these salamanders, evolutionary access to
the new ways of life associated with tropical lowlands.

Paedomorphosis also has been an important factor in the evolution of
tropical salamanders, and it is possible that it may have been of significance in
foot evolution. In B. rufescens and B. occidentalis, and to a lesser extent in
B. hartwegi and B. stuarti, the feet have a rather embryonic appearance. In
embryos of B. subpalmata, which has slightly webbed feet as an adult, pads are
present, with protrusions where the digits will form. The form of the adult foot
suggests that in such forms as B. rufescens, the digits have failed to grow out of
the embryonic pad. The digits are very short and there is a distinct gradient of
development from the proximal to the distal elements. The distal elements are
highly irregular in shape and degree of ossification. The apparent extensive
webbing in these forms is not so much the result of an increase in the integu-
mentary web as the failure of the digits to grow out of the embryonic pad. The
growth differentials between the digits and the carpal or tarsal regions of the
hand or foot are quite different in, for example, B. rostrata and B. rufescens,
and in the latter the growth of the digits relative to the growth of the tarsal
portions is much reduced. Correlated with this developmental trend, which
clearly is adaptive, is the trend to stiffen the entire pad and make it a more

1969 Systematics and Evolution of Neotropical Salamanders

35

+ = helmrichi

r =rostrata

• = hartwegi

f = franklin!

s = stuarti

m = morio

□ = fiavimembris

# =rufescens

o = cuchumafona

a = occidentalis

INDEX B

Figure 16. Proportional relations of phalangeal and metatarsal elements in Bolito-
glossa. Index A. Total length of terminal phalangeal bone divided by total length of
bone in phalanges other than the terminal ones; Index B. Total length of phalangeal
bone divided by total length of metatarsal bone. Each point results from measure-
ment of one foot of one individual.

effective climbing structure. For example, the flattening of the metacarpals and
metatarsals and the digits, and the growth of lateral weblike extensions of
bone into the integumentary pad add to the unification of the foot as a func-

36

Contributions in Science

No. 175

tional unit. The fusions of tarsal and carpal elements are also involved in this
process.

Evidence that paedomorphosis is operative in B. rufescens and related
species may be seen in such features as the absence or irregular development
of prefrontal bones, the general low level of ossification of skull elements, and
the large size of the sensory capsules relative to surrounding cranial elements.

Not all species with extensively webbed hands and feet are involved with
a paedomorphic trend. In B. schizodactyla of Panama the digits are very long
with highly specialized terminal phalanges, and the integumentary part of the
web is greatly increased (Wake and Brame, 1966b). Feet in other species of
Bolitoglossa are specialized in yet different manners. We think that careful
analysis of foot structure in the various species of Bolitoglossa is likely to pro-
vide much information concerning evolutionary relationships and adaptive
processes in the genus.

Acknowledgments

We are grateful to Joseph F. Copp, Scripps Institution of Oceanography,
for making available materials from his private collection (JFC), now de-
posited in the Los Angeles County Museum of Natural History (LACM) . This
material was collected by Dennis E. Breedlove, University of California,
Berkeley, and we express our appreciation to him for taking time from his
botanical field work to collect many important salamanders and provide
habitat information. We greatly appreciate the long continued support of
Charles F. Walker of the Museum of Zoology, University of Michigan
(UMMZ), and Robert F. Inger and Hymen Marx of the Field Museum of
Natural History (FMNH). They have permitted access to their collections
and preparation of cleared and stained specimens. Loans have also been
obtained through the courtesy of Robert C. Stebbins, Museum of Vertebrate
Zoology, University of California (MVZ), Charles H. Lowe and Robert L.
Bezy, Department of Zoology, University of Arizona (UAZ), Richard J.
Baldauf and John R. Meyer, Texas Cooperative Wildlife Collection at Texas
A and M University (TCWC), and Hobart M. Smith and Macreay Landy,
Museum of Natural History, University of Illinois (UIMNH). We have bene-
fited from discussions with L. C. Stuart concerning Guatemalan salamanders.
George B. Rabb provided important radiographs of B. cuchumatana. Photo-
graphs were made by Les Siemens and drawings of the holotypes were made
by Lewis Larsen, both of the University of Chicago. This work has been
supported in part by a grant from the National Science Foundation (GB
6423X) and by the Dr. Wallace C. and Clara A. Abbott Memorial Fund of the
University of Chicago.

Addendum

Since submitting this paper, we have been able to examine two skeletons of

1969 Systematics and Evolution of Neotropical Salamanders

37

Bolitoglossa veracrucis, a species known only from the type series (Taylor,
1951). We are grateful to W. E. Duellman for allowing us to prepare two
cleared and stained specimens.

In the original description, B. veracrucis was compared with B. odonelli,
B. mulleri and B. yucatana, but it is now apparent that its closest relationships
are with the helmrichi group. In its long, truncate, prominent snout, its large,
extensively webbed hands and feet, generally similar head dimensions, rather
short tail, and moderately high numbers of maxillary and vomerine teeth, it
resembles both B. hartwegi and B. stuarti. Skeletal structure provides the most
convincing evidence of relationship to the helmrichi and rufescens groups. The
elongate transverse processes of the first caudal vertebra bear small, accessory
processes that are similar to but smaller than those of B. hartwegi. Only seven
carpals and tarsals are present, the result of fusions, and the digits are reduced
in length and flattened as in B. hartwegi and its relatives. Phalangeal formulae
for two specimens are 1-2-3-2 (Kansas University 26942) and 1-2-3-1 (KU
26953) for the hands, and 1-2-3-2-2 (1-2-3-1-2 for one foot of KU 26953)
for the feet. The metacarpals and metatarsals are enlarged, flattened, platelike
bones that occupy much of the base area of the web. The flattening is more
extreme than in other species examined. Digits and phalanges are shortened
and resemble those of B. hartwegi. Phalangeal bone has a total length that is
less than 80 per cent of the total metatarsal length, and the total length of
terminal phalangeal bone is less than 50 per cent of the total length of the non-
terminal phalanges ( cf . Fig. 16). Terminal phalanges are pointed and very
small. Phalangeal formulae variation results from the failure of some terminals
to ossify. In all of these features, B. veracrucis resembles B. hartwegi, B. stuarti,
and the rufescens group.

Bolitoglossa veracrucis has a well-developed, solidly articulated skull that
is more like those of B. helmrichi and B. hartwegi than those of the rufescens
group or of other species that we have examined. The snout portion is par-
ticularly well developed, and the nasal bones are larger than in any member of
the helmrichi group. They extend ventrally at the anterior end of the snout to
partially envelop the medial border of the olfactory capsules. Premaxillary
bones are less well developed than in B. hartwegi, with shorter and less dilated
frontal processes. Prefrontals are small and round, with the greatest dimension
being less than one-half that of the elongate prefrontal of comparably-sized
B. hartwegi. Maxillary bones are longer than in either B. hartwegi or B. helm-
richi, and they extend beyond the posterior limit of the eye. The intervomerine
space is large; accordingly the spinous medial process, which bears the vomer-
ine teeth, is longer than in B. hartwegi or B. helmrichi. The tibial spur is
distinct and free in one tibia of one animal, and is distinct, but partly joined to
the main shaft of the tibia at the tip of the spur, in the other three tibias ex-
amined. Many trunk vertebrae of both specimens bear well developed basa-

38

Contributions in Science

No. 175

pophyses. All trunk vertebrae but the last bear well developed ribs in both
individuals.

This combination of characters suggests close relationship of B. vera-
crucis to B. hartxvegi and its relatives and indicates that the species should be
considered a member of the helmriclni group. Its spotted color pattern (Taylor,
1951), consisting of a light mottling on a dark background, its prominent snout,
and features of the skull and feet distinguish it from the members of the group
discussed in this paper. The type locality of B. veracrucis (35 km SE Jesus
Carranza, Veracruz, Mexico, 350 feet elevation) is considerably lower than
the known range of any other member of the group. In this paper we have
discussed the relationship of the upland helmrichi group to the lowland mem-
bers of the rufescens group, and have pointed out morphological features that,
in a sense, preadapt the group to life in lowland, arboreal habitats. Discovery
that the lowland B. veracrucis is a member of the helmrichi group supports our
theory. It is a close relative of montane species, yet it exemplifies an extreme
in those trends that seem to be associated with arboreality and the ability to
invade and succeed in tropical lowland habitats.

SUMARIO

El genero Bolitoglossa incluye cerca de 60 especies de salamandras neta-
mente neotropicales. En este trabajo se definen las especies del grupo helmrichi
de Bolitoglossa y sus relaciones evolucionarias son discutidas. Tres especies se
reconocen ahora en este grupo: B. flavimembris de la region de la costa mon-
tanosa del Pacifico en Guatemala y Chiapas, Mexico; B. cuchumatana de la
Sierra de los Cuchumatanes, Guatemala, y B. helmrichi de las montanas de
Alta Verapaz, Guatemala, mas dos especies que aqui se describen: B. hartwegi
de las tierras altas de Chiapas central, Mexico y B. stuarti de la region fronteri-
za de Chiapas, Mexico y Huehuetenango, Guatemala. Miembros de este grupo,
presentan evidencia de relacion general a especies de tierras altas que antes de
incluian en Magnadigita y tambien un especies de rufescens del grupo de
Bolitoglossa especializados a las tierras bajas. La posicion intermedia de estas
especies es una demostracion mas de la tendencia en Bolitoglossa de presentar
especies relacionadas que ocurren en una area geografica restringida y para
estos grupos de mostrar una diversidad morfologica considerable, especial-
mente de las manos y patas. Las especies de tierras altas mas generalizadas,
tienden a tener membranas interdigitales mas reducidas y a ser terrestres. Las
especies de tierras bajas son arboreas y presentan las membranas mas desar-
rolladas. Muchos estados intermediaries en estructura de las paras, se encuen-
tran en el grupo de helmrichi y las modificaciones son aparentemente parte de
la tendencia a la adaptabilidad arborea. A1 menos dos aspectos de adaptation
intervienen: una tendencia pedomorfica que produce un incremento inicial en
las membranas asociado con retention del pie embrionico y una tendencia

1969 Systematics and Evolution of Neotropical Salamanders

39

adaptiva funcional con seleccion de estructuras locomotoras especializadas que
operan en los apendices parcialmente modificados.

El descubrimiento de nuevos miembros del gmpo helmrichi permite la
demostracion de un morphocline continuo en la estructura de las patas, el cual
a su turno proved aspectos en la adaptacion morfologica de las salamandras
neotropicales a las tierras bajas.

Literature Cited

Brame, A. H., Jr., and D. B. Wake. 1962a. A new plethodontid salamander (genus
Magnadigita } from the Cordillera Occidental of Colombia. Biol. Soc. Washing-
ton, Proc., 75:71-76.

. 1962b. A new plethodontid salamander (genus Bolitoglossa) from Vene-
zuela with redescription of the Ecuadorian B. pal mat a (Werner). Copeia, 1962:
170-177.

— . 1963a. The salamanders of South America. Los Angeles Co. Mus. Nat. Hist.,

Contrib. Sci., 69:1-72.

— . 1963b. Redescription of the plethodontid salamander Bolitoglossa lignicolor

(Peters), with remarks on the status of B. pa lust r is Taylor. Biol. Soc. Washing-
ton, Proc., 76:289-296.

Brocchi, P. 1881-1883. Etude des Batraciens de PAmerique Centrale. In Mission Sci-
entifique au Mexique et dans PAmerique Centrale, Rech, Zool., 3(2): 1-122.

Hansen, A. M., and W. W. Tanner. 1958. A comparative osteological study of cer-
tain species belonging to the genus Bolitoglossa (Amphibia). Great Basin Nat.,
18:85-100.

Schmidt, K. P. 1936. Guatemalan salamanders of the genus Oedipus. Field Mus.
Nat. Hist., Chicago, Zool. Ser.. 20: 135-166.

Stuart, L. C. 1943. Taxonomic and geographic comments on Guatemalan salaman-
ders of the genus Oedipus. Michigan Univ., Mus. Zool., Misc. Publ., 56: 1-33.

. 1952. Some new amphibians from Guatemala. Biol. Soc. Washington, Proc.,

65:1-12.

Taylor, E. H. 1941. New amphibians from the Hobart M. Smith Mexican collec-
tions. Kansas Univ. Sci. Bull., 27 : 141-167.

. 1944. The genera of plethodont salamanders in Mexico, pt. I. Kansas Univ.

Sci. Bull., 30:189-232.

. 1951. A new Veracrucian salamander. Kansas Univ. Sci. Bull., 34:189-193.

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

Wake, D. B., and A. H. Brame, Jr. 1962. A new species of salamander from Colom-
bia and the status of Geotriton andicola Posada Arango. Los Angeles Co. Mus.
Nat. Hist., Contrib. Sci., 49: 1-8.

— . 1963a. The status of the plethodontid salamander genera Bolitoglossa and

Magnadigita. Copeia, 1963:382-387.

40

Contributions in Science

No. 175

. 1963b. A new species of Costa Rican salamander, genus Bolitoglossa. Rev.

Biol. Trop., 1 1:63-73.

. 1966a. Notes on South American salamanders of the genus Bolitoglossa.

Copeia, 1966, 360-363.

. 1966b. A new species of lungless salamander (genus Bolitoglossa) from Pan-
ama. Fieldiana, Zool., Field Mus. Nat. Hist., Chicago, 51:1-10.

Wake, D. B., and I. G. Dresner. 1967. Functional morphology and evolution of tail
autotomy in salamanders. Jour. Morph., 122:265-306.

Accepted for publication Dec. 1, 1968.

Z -J

RIES SMITHSONIAN INSTITUTION NQIlfUllSNI NVINOSHilWS S 3 I H Y 8 8 1 1 LIBR

— c n *- — ip

LI1SNI NVINOSHillNS S3IHV8 8n LIBRARIES SMITHSONIAN INSTITUTION NOI1

2: co

CO

hr? loL, x O

RIES^SMITHSONIAN INSTITUTION NOlinillSNI NVtNOSHilWS^Sa I aVU 8 n^LI B F

00 ~ c/5 ~ C/0 ~

> W'<

LU

CE

31 <
cn

rn «3«0r 5 N^pr 2 5 ^ = *

.IlSNl^NVINOSHllWS S3 I 8 YH 8 II LI 8 R AR I E S^SMITHSONIAN^INSTITUTION ZN0I1
2 r* 2 n z

RiES SMITHSONIAN INSTITUTION NOtlfUliSNl NVINOSHilWS SBI8V88I1 LI B F

RIES SMITHSONIAN INSTITUTION NQIlfUllSNI NVIN0SH1IWS S3 I 8 ¥8 8 il LIBF

U)

m

— CO Sr CO

LI1SNI NViNOSHimS S3I8Y8 8I1 LIBRARIES SMITHSONIAN INSTITUTION N0I1

~ r>> rr> — C/5

>

,R I ES^SMITHSONIAN^ INSTITUTION NO UHIIISN I ~N¥IN0SH1IW$'~ S3 I 8 ¥8 8 11 LI B f

* s-

e 5 't&gr “ • 5 2

SMITHSONIAN INSTITUTION NOIXflilXSNI NVINQSHilWS S3 I a VH 8 IT LIBRARIES
2 f“ 2 r- 2 *■" ^

O . X^iO^/X 2 XCy^VaTx V- O X\V^hT>v m

00

m ^ x^jiusr^ rn

oo n ^ co \ S to

NVINOSHXIWS S3iavaail LIBRARIES SMITHSONIAN INSTITUTION NOliniliSN

</5 Z ... W 2 w

5 /&rrnK\ q . ,/ • z ,*c?\. . -i /Jim®

CO

o
2

«» S 2 *\ >

“SMITHSONIAN^ INSTITUTION NOUriilXSNI NVINOSHilWS^SB I a V 8 S3 IT LIBRARIE

CO CO ZZ ..... C/5

V..

~

o xu^jxc.x — X^?iixs^X o v XgUixs^X o

NVIN0SHIIWSZS3 I HVB a II LIBRARIES SMITHSONIAN^INSTITUTION NOliniliSN
2 r- z £ ___ z

° ^iir^x ~ xf5o*ox 2 m 2

CO

» /£'

> p

^ lyP

2 >- ✓"V /^v’ ~

m <- ^ x^»va5^x rn X^osv^x m

SMlTHSONlAN"’lNSTITUTIONC/>NOliniliSNI~NVINOSHlliMS S3 1 BVB 311 UBRAR IE:

CO 2 ^ CO 2 CO . 2 >

5> > 's/'yosvAiX > *2

CO z co — z CO

nvinoshxiws S3 lava an libraries smithsonian institution NotxnxixsN

V CO — CO ~ CO

X x- 1 1 « ^nrr^~^ <*- < . i Z

CO

'-t.\

/•v

o

SMITHSONIAN INSTITUTION NOIXflilXSNi NVINOSHilWS S3 I a VH 8 IT __L I B R AR 1 E

° .-5 1 5 jEx 1 § /#— 75

m v^v. pcx r; n^^s^x m ^ ^ *x m

co ~ co VS co

NviNosHims saiavaan libraries smithsonian jnst tion noi-ioliisi'

2 CO 2 CO 2

CO

< 4 s

— i

1
CO

2 jKV X o

„ N ~ Vcfc&^v tz ^ e

2 ^ > V^ilxS^X ^ V >

CO Z CO * Z CO z

SMITHSONIAN INSTITUTION NOimillSNI NVINOSHUWS SBlUVaail LIBRARIE

co 5 co ~ . CO X

kiumnoii i ii*i(' o i \i w \i n in i « rs rA a « r- #-«, a iti irAMiAHi imptitiitiasii