Se ee
Sct cEs Shsapteesn pers, 5
Astras mecemenrarereren ey

tn ee Se ea eee 2 RAs RN

ee tant

Se

Soe Spe ees

Liat
oe
—*

SC etelot
peewee
Sepa

aes

Ry eat omnes =
Fe oh ZS"

PS

~ oe a
ND a a ee

= wale ~ —_ - —
SN ee
SE ————— eee

NSTITUTION NOILNLILSNI_ NVINOSHLIWS

D for Be ANY oO Gj) as PLAS ro) Vy fp Yo a2 \%) 3] O thy im
= WO = AMR OE WY FS 5 Yi Ea) 2, |
x W : Ne = |
2) x 2 5 s a uz
”
ud

Ni

= ;'
S32!4YVYsIT LIBRARIES SMITHSONIAN INSTITUTION

NOILNLILSNI
NOILNLILSN!

LIBRARIES
NOILNLILS

S31NYVYUSIT LIBRARIES SMITHSONIAN INSTITUTION NOILNLILSNI

S3iyvudit
INSTITUTION
INSTITUTION

INSTITUTION

NSTITUTION NOILALILSNI NVINOSHLINS SSIYVHYEIT LIBRARIES

NVINOSHLIWS S31uvudit_ LIBRAR!
NVINOSHLINS S3IYVYa!IT_ LIBRARIES

NVINOSHLIWS
SMITHSONIAN
SMITHSONIAN
SMITHSONIAN

Z te S a
LIBRARIES SMITHSONIAN_INSTITUTION NOILNLILSNI NVINOSHLINS S3Z1¥vugIT

Ss

NOILNLILSNI
LIBRARIES

INSTITUTION NOILALILSNI NVINOSHLINS S3IYvYYaIT LIBRARIES

INSTITUTION NOILNLIASNI
INSTITUTION NOILNLILSNI
Sa!YVYGIT LIBRARIES

SAIYVYURITILIBRARIE

INSTITUTION

Saiu¥vudit

LIBRARIES SMITHSONIAN INSTITUTION NOILOLILSNI_ NVINOSHLIWS saruvudli

os

z ” z 7) z n }

< = < = Wi < = W)

z =f Fa a typ = 4 UY
ane Ae re) 28 Ys fe) a= 1 x

“” $2) Y”) Tp) Jf: 2) wo os

ae ©) ac 0 [VL = fe) Gy:

= > > = >

7) = S 23 79) Zz

NSTITUTION NOILMLILSNI NVINOSHLIWS saruvuai LIBRARIES SMITHSONIAN INSTITUTION

? 2 2 oe
z = z z = WY
oo — Lae = loa ,
= fe) BN 2 re) eae |
S Zz — = ay
S3IUVUSIT LIBRARIES SMITHSONIAN INSTITUTION NOILNLILSNI
z Es z i z
= = 3 = o =
> =e > = > 4 >
a i: 2 = a -
O z i D z Oo Zz
INSTITUTION, NOILLILSNI_ NVINOSHLINS, SAIUVYSIT LIBRARIES, A
S\ = as = at CK = <
= Bh Share = z NS = z
se. KY ro) 35 9° MK ~ = 9
suff % Be 2 2 2
2 “yy = 2 E “© 2 =
. BS “EG = > = Se = i
be Zz ”n Zz Ww eS wo
Sa1uyvugi7 LIBRARIES SMITHSONIAN ANOILMLILSNI_NVINOSHLINS Sa1uvuaty
Zz & 2 0 WJ Sv. = WN uu V.
LR APR 2g fg 3 PR = KR 3 NO = RB

y - Lai Misi:
NX 2 GE 5) 2 5 GG 2
NY 2 Dy = (297 f& ©
sY \ ENS = ee f =
~~ 5 2 5
PILALILSNI_NVINOSHLIWS pale) BRARI ES SMITHSONIAN
a ul z tu Yyy
g - Z - Wy
4 WJ tS =i a OY 2, Yfyy,
c AW y < = <t UY, Ves
4 XN SS oo 4 load ‘ Gu
= S foo) = Ag
2 I lee 3 z
|BRARIES SMITHSONIAN INSTITUTION NOILNLILSNI NVINOSHLINS
eG r
oO se oO -
— ow -_ o
5 a = »
F 3 > res >
F z = Z
Z m Z o
PIN IILSNI oot lyYvUug roti BRARI ES SMITHSONIAN
as = Pan Ks =
oe. “ q Ss x : S XN Y x
4Jf ? gy EN 38
Viitjp ca = Sx 2
|BRARIES SMITHSONIAN _ INSTITUTION NOILNLILSNI_ NVINOSHLIWS
Sy S ae it
“pop 3 : 3
w Uy, S tc c
2 “Uj rs} = 5 2
z i = <
DILALILSNINVINOSHLIWS | $3 1yYVvug fae LIBRARI ES _ SMITHSONIAN
== fe) = fe) J
> We 5 z =
a WA, 2 aa >
Na WE E
a \ E a =
SMITHSONIAN INSTITUTION NOILOLILSNI
52) ma * wn z=
= g = gh, :
= S \ i 8G.2
w 2 LN @ Vid m7)
= SW “WS oO Vi ff,
Z e AWS 7 Zee fo OE
_ NVINOSHLIWS S3IYVYEIT LIBRARIES SMITHSONIAN
E us z
a S Yi
— =~ Wh
Se * : Ee Uff
a Oh
2 BS fe)
z

IBRARIES SMITHSONIAN INSTITUTION NOILNLILSNI NYINOSHLIWS

INSTITUTION NOILALILSNI

NVINOSH

INSTITUTION

INSTITUTION NOILALILSNI

S31YVYSIT LIBRARIES SMITHSONIAN

INSTITUTION

NVINOSHLIWS

S3IYvugit

Ssaiuvudia LIBRARIES

”
tu
e a
. <
a fo
o mo.
a a
rua {i
3 3 6 Shee
= = ea :
=e a 3 ra 8
ke : > kK Pah
=" zu = a
2 : 2 4
SSN pot tYVNdIT_ LIBRARIES SMITHSONIAN
< = Aes NE = <x
, a =4 uv ‘s \S = >
4 Z, 5 = y 5 AS = 5
“iy = z = NS Z E
“Gy = > ‘ = Cp =
\ (7) Zz ” ors 7)
IBRARI ES SMITHSONIAN _INSTITUTION NOILNLILSNI NVINOSHLINS S3IYVY
z = Ce ”
) Ser = Bee igo 2 Seqp) ad eae) D a ASS pa ASE
jE PG 36E 7 = (EPMA) SS = G
e a) <--277 J2the wg < ce Fe) = WS < EF

INSTITUTION

NVINOSHLINS SAIYVudlT

gi

NVINOSHLINS S31YVYSdIT LIBRARIES SMITHSO

Sa1uvuaiq_.

RY 2
Wwe
NY =
SW >
NOILALILSNI_N
me
wn
=I
Coe}
S
=
oO
za
LIBRARIES §
=z
°
ja
2.
aN
—
Yn
=
NOILMLILSNI A
a
LY §
Wy = |
Yi =
LIBRARIES S$
Zz |
O aa
2 =
<3
= a
3 ate
Zz |
NOILNLILSNI_N
= ce
ae
= -
= en
Zz yA
LIBRARIES S
ra ) w
< =
z 4
{e) AG
(7p) Ww
ac oO
= z
a 8
NOILALILSNI ON
” Zz
ud ”
= zy
< c
a 4
22 re)
= 2
S
c =z
S °
2 5.
= EP
~ 5
o z
NOILNLILSNI N
wn =
z &
2 1h tp 5
$4 ff 2
Oo |
ae
LIBRARIES §
= Ww
7) Soa a
at 3 4\ of
c a) SS 4

N
.

STITU

weliuy
SaVIg
2 )
€ a
Tew. wit

a
&
<<
~w

DINO
S >
& ce
2 FA
r=
a} <9

o
zs
=
=

Vile
Be

" DUNO
Oo <>
fS
2
=
& =

w
a
4
Cc

nett uy,
igrs
e)
a
ZF; ~az

Ld
bag
<

x
wo
<

DINO
ky <4 ON
'S a
=
= c

on
Qa
GA A

Uf
Z
G

we
e
<x

v

ey,
a
t
ww
mS

Ss & Wary - KY FY | (Na i Nay - ONS! Ray = KA GF GY Sl =e &®

UNIVERSITY OF KANSAS MISCELLANEOUS

PUBLICATION

MUSEUM OF NATURAL HISTORY NO. 71

Late Pleistocene Herpetofaunas
From Puerto Rico

By
Gregory Pregill
AOSONIAN
how”
fo }
f
' JS
LIBRA

UNIVERSITY OF KANSAS
LAWRENCE 1981

7 =
\ ~
‘A ~w

fy

SRG

UNIVERSITY OF KANSAS PUBLICATIONS
MUSEUM OF NATURAL HISTORY

The University of Kansas Publications, Museum of Natural History, beginning
with volume 1 in 1946, was discontinued with volume 20 in 1971. Shorter research
papers formerly published in the above series are now published as Occasional
Papers, Museum of Natural History. The Miscellaneous Publications, Museum of
Natural History, began with number 1 in 1946. Longer research papers are pub-
lished in that series. Monographs of the Museum of Natural History were initiated
in 1970. All manuscripts are subjected to critical review by intra- and extramural
specialists; final acceptance is at the discretion of the Director.

Institutional libraries interested in exchanging publications may obtain the Occa-
sional Papers and Miscellaneous Publications by addressing the Exchange Librarian,
University of Kansas Library, Lawrence, Kansas 66045. Individuals may purchase
separate numbers of all series. Prices for all publications of the Museum may be
obtained from the Publications Secretary, Museum of Natural History, University of
Kansas, Lawrence, Kansas 66045.

Tue UNIVERSITY OF KANSAS

MusEuM oF NATURAL HIsTORY

MISCELLANEOUS PUBLICATION No. 71

May 8, 1981

Late Pleistocene Herpetofaunas From

Puerto Rico ,
/

By

GrecorY PREGILL
Department of Vertebrate Zoology
National Museum of Natural History
Smithsonian Institution
Washington, D.C. 20560

THE UNIVERSITY OF KANSAS
LAWRENCE

1981

ee

S SP) a)

a &

No PO
Via ®

—
_
~~

CA

UNIVERSITY OF KANsAS PUBLICATIONS, MusEUM OF NATURAL HisToRY

Editor: R. M. Mengel
Managing Editor: Joseph T. Collins

Miscellaneous Publication No. 71
pp. 1-72, 26 figures; 5 tables
Published May 8, 1981

Museum oF Natural History
THE UNIVERSITY OF KANSAS
LawreENce, Kansas 66045

U.S.A.

PRINTED BY
UNIVERSITY OF KANSAS PRINTING SERVICE
LAWRENCE, KANSAS

CONTENTS

Mes PTR ISTIC che ae a ee i
PAO KEN ON VED GINGIGN (Ss ee ee ee 22
MURTRIOIDCIUOG’ 4 ee ee eee 3
PHYSICAL DESCRIPTION OF PUERTO RICO
(COO TRT LOS apa ce a 4
ICO RICA MInTKe eZ ON CS eemenes ees eS ae ee 4
CAVE LOCALITIES AND FOSSIL DEPOSITS _............--- 7
Gavessmtherbarahoma Avea 2 11
Oi CTBELOSSI IES ite seme eens eed eRe 15
Lies LE (Lae TD YES ONO eee Be a 15
SYSTEMATIC PALEONTOLOGY
Amphibia
Leptodactylidae
Meptodacilusvalbilabris Sexo sea 17
ICULRenOMGeHLUsESp ps tee me) ae 19
Bufonidae
CIELO PINUMCMIC TI geen ere eee em ne yee i ee SN) Te 20
Testudines
Emydidae
Ciiguscniy Sct mde CUSSALDy se ee 23
Sauria
Gekkonidae
Spracnodactylus. sp, inGety 23
Iguanidae
ING OUES ae a a Eo 24
ANOIIS) CUIDU GY, OE as we TN ree ens a 24
VAN OU SM CIA MCTASEOLCHIUS) eee =e we ee ee 26
KROES GOLAN LL ee Tl
NEOUS US PULTE aa NTN DI Ne Se Te es ne 27
INGUOUES GOO TIITTIS 2 cee ee ee ee eee 28
ISGUDUES: SD) Dh, a eo eee 28
CUGUUINED FOURGUES: ea a ok a a en ori sea Ee ee se 29
PEC LOCE PICU peerage Pret SNIP Re TS 34
liciocephalusnethemdeci new, species) 35
EClocepRalusm pants. muew SPECIES —2 ee ee 39
Scincidae
WGLOUGE TOGIONC UA ae ae ae ea 43
Teiidae
PACIVCE LS (EMC AS 10 eee mn aN k= 65 tT RE UR il ith eek Ws yn Py ale 43
Anguidae
Diploclosstismpleciqesee as ee ee ee 45
Amphisbaenia
Amphisbaenidae

PMN DIUUSOUCI gS) RUNG ey = te Ss 48

Serpentes
Typhlopidae
Typhlops sp. indet.. 22-205 ete 49
Boidae
Epicrates inornaius 222.2 =. 02 eee 50
Colubridae
ef. Alsophis portoricensis eee 52
cl. Arrhyton exiguum: 22552 2 eee 53
DISCUSSION
Comparison of the Cave Faunas EE eee 53
The Paleoenvironment _....0. eee 57
Zoogeographic! Considerations ee ee 60
SUMMARY AND CONGEUSONS 63
LITERATURE CITED 22.2020 22 eee 65
APPENDIX ps2. c:ccco eee ee ee ee 68
APPENDIX Ui vce $.ee ieee AS oo ee ee 70

INTRODUCTION

The Caribbean sea and the thousands
of islands and cays composing the
Greater and Lesser Antilles are among
the most biologically complex areas of
the world. Some understanding of this
complexity has been achieved in recent
years by the application of modern tech-
niques and principles of population ge-
netics, ecology, and island biogeography.
To this end, amphibians and reptiles are
particularly instructive subjects, but their
usefulness in addressing biological prob-
lems in the Antilles has not been fully
appreciated. Many of the _ urgent
questions remaining about their origins
and systematic relationships bear di-
rectly on the evolution of the West
Indian biota. Among the more valuable
data pertinent to such inquiries are the
fossil remains of these animals found as
owl pellets in limestone caves and sinks.
The available West Indian (terrestrial)
vertebrate fossil record is no older than
late Pleistocene, partly because lime-
stone caves are geologically ephemeral
_structures, and also because of a general
scarcity of nonmarine sediments with
the potential of containing older fossils.
A few vertebrate fossils of Tertiary age
have been recovered from marine sedi-
ments, however, including remains of
extinct sirenians known from the Eo-
cene(?) to early Miocene of Jamaica,
Cuba, Puerto Rico, and Sombrero (see
Verona, 1974 for a synopsis), and also
shell fragments of a pelomedusid turtle
described from the San Sabastian For-
mation of Puerto Rico (Wood, 1972).

In spite of their recency, the amphib-
ian and reptile fossils from the Antilles
often differ from living species taxonom-
ically, morphologically, and distribution-
ally. Herpetofaunal remains are not as
well known as are those of birds and
mammals, and most amphibian and rep-
tile fossils are from the Greater Antilles.
Fossil lizards have received the most
attention and have been described from
Cuba, Jamaica, Hispaniola, the Bahamas,

Cayman Islands, Barbados, and Barbuda
by Etheridge (1964a, 1965, 1966a,
1966c), Auffenberg (1959), Hecht (1951),
Ray (1964), and Morgan (1977). Ex-
tinct tortoises are known from widely
scattered localities in the Antilles (Auf-
fenberg, 1967, 1974). Fossils of snakes
have been reported from Cuba, the Cay-
man Islands, and Barbuda (Koopman
and Ruibal, 1955; Auffenberg, 1959; Mor-
gan, 1977), but thus far frog fossils have
been reported only from Barbuda (Auf-
fenberg, 1959, Lynch, 1966), and the
Cayman Islands (Morgan, 1977). Oc-
casionally, amphibian and reptile fossils
are also recovered from archaeological
sites in the islands.

There has been no thorough treat-
ment of a large fossil herpetofauna and,
in general, paleontological research in
the West Indies has languished in recent
years. Many systematic, zoogeographic,
and paleoecological puzzles remain un-
solved. For example, on Puerto Rico
the only amphibian or reptile fossils
known are the isolated remains of the
rock iguana, Cyclura, the specific iden-
tity of which has been questioned since
Barbour described it in 1919. Moreover,
the cave region of northwest Puerto Rico
has been known for years as a rich
source of vertebrate fossils, but most of
the studies were done over 50 years ago.
Wetmore (1920, 1922) described many
fossil birds collected by H. E. Anthony,
including the extinct barn owl, Tyto
cavatica, responsible for most of the ac-
cumulations of small fossils. A few years
later Anthony himself (1925-1926) re-
viewed the abundance of small mammal
fossils. Many of these were bats, but
also included were an extinct insectivore,
Nesophontes edithae, the rodent Elas-
modontomys, and two species of extinct
sloths of the genus Acratocnis. More
recently, another series of bats and ro-
dents was described by Choate and Bir-
ney (1968) from three Puerto Rican
caves worked by James W. Bee of the

2 MISCELLANEOUS PUBLICATION MUSEUM OF NATURAL HISTORY

University of Kansas in 1957. The col-
lection of James Bee also included a
number of fossil lizards, but these were
not studied. While casually examining
this lizard material a few years ago, I
found dentaries belonging to the lizard
Leiocephalus, although the genus pres-
ently is not known from the island. While
I was pursuing this problem in 1976 and
1977, field parties from the Smithsonian
Institution revisited Anthony’s localities
and excavated new caves. Subsequently,
I worked some of these sites in 1978.
Thousands of vertebrate bones were re-
covered from a number of caves. The
herpetofaunas from these efforts and
those of the James Bee collection are the
subject of this paper. The birds and
mammals are being studied by other
researchers.

The fossil material includes 12 liz-
ards, 4 snakes, 3 frogs, an amphisbaenid
and a turtle. There are two new species
of Leiocephalus and abundant remains
of Cyclura. The diversity of species ex-
ceeds any previously known fossil her-
petofauna from the West Indies and as
such, invites comparative discussion of
the animals past and present.

The first section of the paper de-
scribes the physical features of Puerto
Rico and the nature of the cave deposits.
The majority of the text is devoted to
description of the fossils (Systematic
Paleontology) with elaboration on the
osteology of poorly known species. Be-
cause descriptive accounts rely, at least
in part, on the availability of adequate
comparative material, some species are
treated more thoroughly than others.

The last section discusses the fossil
assemblages in terms of species composi-
tion, bias owing to owl predation, the
paleoenvironment, and finally the zoo-
geographic implications. Studies of this
sort lend themselves to wholesale specu-
lation, but it is not my intention to
answer, nor even raise all relevant ques-
tions. Rather, by exploiting the potential
of the Puerto Rican herpetofaunas I have
attempted to synthesize pertinent aspects

of the fossil record, and in part, focus the
paleontological perspective of West In-
dian amphibians and reptiles.

ACKNOWLEDGEMENTS

Studies of this scope encompass the
overlapping efforts of many people. Con-
sequently, there is no adequate way to
acknowledge all of them and their re-
spective contributions. For the loan of
specimens I am grateful to Walter Auf-
fenberg, William Duellman, Richard
Etheridge, John Lynch, Peter Meylan,
José Rosado, Richard Thomas, Al
Schwartz, Ernest Williams, George Zug,
and Richard Zweifel.

A number of persons contributed
substantive discussion, although my in-
terpretation of their comments at times
may have deviated from their intended
meaning. These people are James Bee,
Thomas Berger, Janalee Caldwell, David
Cannatella, Ronald Crombie, Richard
Etheridge, Richard Estes, Darrel Frost,
Jessica Harrison, Philip Humphrey,
Jacques Gauthier, Robert Hoffmann,
John Lynch, Michael Maher, Gary Mor-
gan, Jaime Péfaur, Rebecca Pyles, Carol
Terry, and E. O. Wiley.

I thank Storrs Olson, National Mu-
seum of Natural History, Smithsonian
Institution, for making much of the fos-
sil material available to me and allow-
ing free access to field notes and ideas.
Olson was assisted in the field by Fred-
erick V. Grady, Noel Snyder, and J.
Philip Angle. Noel Snyder was instru-
mental in locating caves and collecting
fossils, and Fred Grady made a tireless
effort in the initial sorting of the ma-
terial. The hospitality and field assis-
tance of Richard Thomas exceeded
generosity.

This paper is the essence of a doctoral
dissertation written at The University of
Kansas with the guidance of William E.
Duellman, Larry Martin, E. O. Wiley,
and Linda Trueb. These individuals
were always available when I needed
them. I thank Linda Trueb for an espe-

PLEISTOCENE PUERTO RICAN HERPETOFAUNAS 3

cially large contribution to this work.

Field work was supported in part by
The University of Kansas Graduate
School and The University of Kansas
Endowment Association’s Watkins Fund.
I received additional support from the
The University of Kansas Graduate
School as a Summer Fellow and from
the Smithsonian Institution as a Short-
term visitor.

METHODOLOGY

All fossils were prepared, identified,
and catalogued from unsorted material.
Fossils were grouped into bones of the
same element according to species. Cat-
alogue numbers were assigned to lots
of the same element for each species
except for holotypes, paratypes, and
those shown as figures or specifically
referred to in the text. These were as-
signed individual numbers. Each species
account includes material, description,
and comments. Material includes all the
fossils for that species arranged by abun-
dance in a particular locality and in a
skeletal sequence of cranial, appendicu-
lar, and vertebral elements. The descrip-
tion is a discussion of osteological fea-
tures pertinent to the identification of
the material. Each description is fol-
lowed by comments on the morpho-
logical implications of the fossils, sys-
tematic and taxonomic notes, and the
distribution and habits, when known, of
the extant animals to which the fossils
are referred. Snout-vent length (SVL)
estimates were made for the fossil in-
dividuals whenever possible. This was
done by multiplying the measurements
of separate fossil elements against the
ratio of body proportions for that skele-
tal element from known, whole indi-
viduals. Ontogeny, allometric growth,
and individual variation produce differ-
ent estimates according to which ele-
ment is used; for example, dentaries,
frontals, or limb bones. Thus, an aver-
age, minimum and maximum estimate is
given for each bone used in the compu-

tation when bones were sufficient in
quantity.

Museum acronyms appearing in this
paper are as follow:

AMNH American Museum of Natural

History

Albert Schwartz, Miami-Dade

Community College

FSM Florida State Museum, The
University of Florida

GKP Gregory Pregill, The Univer-
sity of Kansas

JDL John Lynch, The University of
Nebraska

KU The University of Kansas Mu-
seum of Natural History

ASFS

KUVP University of Kansas Verte-
brate Paleontology
MCZ Museum of Comparative Zool-

ogy, Harvard University
REE Richard Etheridge, San Diego
State University
RT Richard Thomas, The Univer-
sity of Puerto Rico
National Museum of Natural
History, Smithsonian Institu-
tion

USNM

Osteological terminology for lizards
follows McDowell and Bogart (1954),
Oelrich (1956) and Etheridge (1959);
for snakes McDowell (1975) and Pregill
(1977); and for frogs Trueb (1973) and
Lynch (1971). Fossils were measured
with dial calipers read to the nearest 0.1
millimeter. All osteological measure-
ments in the text are expressed in milli-
meters unless indicated otherwise.

Photographs were made using a
Cannon FTb camera mounted on a Wild
microscope. First, fossil specimens were
placed on pin heads and dusted with
ammonium chloride. A pin-mounted
specimen then could be inserted into a
styrofoam block draped with a black
velvet background. Negatives were shot
with Kodak Plus-X Pan film (ASA 125).
Prints were reproduced on Kodabromide
F5 high contrast single weight paper
and then arranged together by bone and
species.

4 MISCELLANEOUS PUBLICATION MUSEUM OF NATURAL HISTORY

The number of names applied to
island groups of the West Indies often
results in confusion as to their geo-
graphic limits. In this study, the West
Indies include all islands in the archi-
pelago between Florida and northern
South America. The Greater Antilles
comprise five major islands or island
groups and their satellites—The Ba-
hamas, Cuba, Jamaica, Hispaniola (Haiti
and Dominican Republic), and Puerto
Rico. The Lesser Antilles include all
West Indian islands in the archipelago
south of Puerto Rico. The name Puerto
Rico only refers to the island itself. The
term “greater Puerto Rico” or the “Puerto
Rican Bank” comprises Puerto Rico and
its offshore islets, and the British and
US. Virgin islands exclusive of St. Croix.
A base map of greater Puerto Rico and
the islands of the West Indies is pre-
sented as Figure 1.

Puerto Rico’s herpetofauna was last
treated comprehensively by Schmidt
(1928), who provided identification keys
of the species known up to that time.
Mostly I have relied on the check list of
Schwartz and Thomas (1975) for much
of the taxonomy and distributional data
of the West Indian species.

PHYSICAL DESCRIPTION OF
PUERTO RICO

Geography.—Centrally located in the
West Indies, Puerto Rico lies at the east-
ern extreme of the Greater Antilles at
18° north latitude approximately half
way between the southern tip of Florida
to the north and the Caribbean coast of
Venezuela to the south (Fig. 1). The
island is rectangular with an area of
8,897 km?. Topographically it varies
from coastal flat lands to peaks of the
Cordillera Central which transects the
island from east to west. The three prin-
cipal geographic regions of Puerto Rico
are the mountains and foothills of the
Cordillera, a discontinuous fringe of
more or less flat coastal plain, and the
rugged limestone and karst regions oc-

curring primarily in the north central
and northwestern part of the island (Fig.
2). The Cordillera Central dominates
from Mayagiiez on the west coast to
Aibonito in the eastern third of the
island. Maximum elevation is 1,338
meters at Cerro de Punta in the south
central part of the island. The range is
met in the east by the Sierra de Luquillo
and the Sierra Cayey. The mountains
are deeply eroded by streams and valleys
to several hundred meters, and slopes of
30° to 40° are common. Rocks are mostly
volcanic and include lava, sedimentary
rocks derived from basalt, intrusive ma-
terial and discontinuous beds of lime-
stone. These rocks date from Early Cre-
taceous to middle Eocene (Monroe,
1976).

The flat, coastal plains slope gently
upward from the shore to the foothills.
They are characterized by low relief,
lagoons, and mangrove swamps and are
sporadically peppered by projecting
cones of volcanic or intrusive bedrock.
The coastal plains grade into alluvial
plains of the larger rivers of the island.
Except for isolated hills, rocks are Qua-
ternary and consist of sand, clay and
gravel from alluvial fans of rivers and as
dunes of beach sand.

As in other areas of the Greater
Antilles, the limestone and karst regions
of Puerto Rico are geomorphic features
unique in the Neotropics. The name
“karst” is taken from the geologically
similar regions of southern Europe. Karst
landscape is best developed in a broad
limestone belt extending 140 km from
the Rio Grande Loiza, just south of San
Juan, to Aguadilla on the northwest
coast. Its widest point is about 22 km
near Arecibo. Six formations are recog-
nized in the karst region dating in age
from late Oligocene to middle Miocene
(Monroe, 1976).

Ecological Life Zones.—According to
Ewel and Whitmore (1973) Puerto Rico
occupies the Subtropical Latitudinal Re-
gion of the Holdridge Life Zone system.
Mean annual temperatures at sea level

PLEISTOCENE PUERTO RICAN HERPETOFAUNAS

‘spuelsy UIsIIA YsHug
pue “sf, oy} puke OONY OWENg :JasUT ‘spULIST JoyJO oY 0} UOHR[EI UT CONY OJON Jo uONTsod orydess0a8 oy} Burmoys Sorpuy ySeAA OY} JO AeW—'T “O1J

v9 89 el OL 08

TVYINIZ9 Ib

e [—~] OO1Y ae
2 2° =@ { ee)

SANDIIA

x
NHOr‘is OOl8 OLYANd

Sn x . °

a C7aq SVYWOHL YS

GS yavoanv

MISCELLANEOUS PUBLICATION MUSEUM OF NATURAL HISTORY

‘QLG1 “SOUOJ 10}Ze) SuoIder o1ydeasoosorsKyd jediound oy} Burmoys oony oyiong jo deW—s ‘Iq

SUle|d |Je1seoj ease snoulejuno; 1/9q Issey

OS Sia] awo|!»y O

ee ——————— ee $a

00,99 ,0€,99 00,29

| eoiuend |

uenr ues oqiaiy

,00
eA Gik

zanbedew

0€
BL

PLEISTOCENE PUERTO RICAN HERPETOFAUNAS 7

are less than 24°C, the lower limit of
the Tropical Latitudinal Region. This
distinction between the geographic trop-
ics—The Tropic of Cancer to the Tropic
of Capricorn (23°27’)—+reflects differing
physiognomies and species composition
of sea level forests north or south of 12°
to 15° latitude. Six climax forest types
occur on Puerto Rico (Fig. 3). These
range from Subtropical Dry Forest in the
southwestern corner of the island where
the mean annual precipitation is be-
tween 600 and 1000 mm, to Subtropical
Lower Montane Rain Forest in the
Luquillo Mountains, which has an an-
nual rainfall of over 4,500 mm (Ewel
and Whitmore, 1973).

The more important differences in
natural vegetation types are especially
impressive considering the short dis-
tances over which they occur. However,
60% of the island is covered by Sub-
tropical Moist Forest including the cave
regions in the Barahona Valley system
from which most of the fossils were ob-
tained (Fig. 4). Subtropical Wet Forest
exists at higher elevations of the Cordil-
lera Central and Sierra de Luquillo and
accounts for 24% of the island area. Sub-
tropical Rain Forest, Subtropical Lower
Montane Wet Forest and Subtropical
Lower Montane Rain Forest are re-
stricted practically to the Luquillo moun-
tains and account for less than 3% of the
life-zone areas.

Subtropical Moist Forest in the lime-
stone hills presents a continuum of vege-
tational associations owing to the north-
east-southwest orientation of the hills.
The hills are humid and moist on the
northeasterly facing (windward) slopes
and even more so on the southwesterly
(leeward) slopes, but they are xeric on
top. The average-diameter growth rates
of trees on the leeward slopes is nearly
twice that of trees on the windward
slopes and growth rates at the top of
the hills are significantly slower than
those on the bottom (Ewel and Whit-
more, 1973). Pasture is the predominant
land use in the Subtropical Moist Forest,

but sugar cane and coffee occupy large
areas; pineapple is grown in the drier
portions along the north central and
northwest coastal area of the island. The
effect of agriculture on the natural habi-
tat undoubtedly has had a significant
impact on the distribution of the island’s
herpetofauna, the ramifications of which
are discussed in the section on the
paleoenvironment.

The other life zone from which fos-
sils were collected is the Subtropical Dry
Forest in the southwest corner of the
island near Guanica. The serpentine-
derived soil of this area harbors a num-
ber of endemic trees and shrubs. Trees
are slender, open-crowned and usually
less than 12 meters tall (Fig. 5). The
forest floor is open because the exces-
sively drained soil supports little her-
baceous growth. Most plant species are
sclerophyllus and evergreen (Ewel and
Whitmore, 1973).

CAVE LOCALITIES AND
FOSSIL DEPOSITS

Pleistocene vertebrates from Puerto
Rico have been obtained almost entirely
from cave deposits. The present section,
describing the location and nature of
these deposits, particularly those newly
discovered, is intended to suffice not only
for the fossil herpetofaunas detailed be-
low, but also for future publications by
other researchers on the birds and mam-
mals from the same sites. To this end,
Storrs L. Olson of the Smithsonian In-
stitution has contributed a substantial
portion of the following from his field
notes and recollections.

Only a fraction of the hundreds of
caves present in the karst regions of
Puerto Rico have been explored paleon-
tologically. Caves range in size from
small rooms, barely head high, to im-
mense caverns with antechambers and
side rooms. Some are entered from
ground level (Fig. 6), whereas others
are reached only by scaling cliff faces or
by climbing down through collapsed

MISCELLANEOUS PUBLICATION MUSEUM OF NATURAL HISTORY

“PULISI 9} FO SoUOZ OFT] [POISO[0D9 94} SUIMOYsS OoNY O}1ong Jo de|—¢ ‘oI

jsauo} ules }SasO} 1AM ySou0} ule 1S910} JOM JS910} YS1ow 1Sa10} Aip
aueyuoW JAMO] aueluoW JAMO] |eos1doiiqns jeo1idoiyqns jeo1doiqns Jedidouqns

mT

l

OS Si9JQAWO|!> O

00,99 0€ 599 00,29

| eoiuens

zanbedew

sajelg
@

uenf ues oqiay

PLEISTOCENE PUERTO RICAN HERPETOFAUNAS 9

a

mae

Fic. 4.—The Subtropical Moist Forest near the cave region south of Barrio de Barahona, Puerto Rico.
Much of this area was cleared for farming in the last century and is now feral pasture and forest.

Fic. 5.—The Subtropical Dry Forest in the vicinity of Guanica Bat Cave of southwest Puerto Rico.

10 MISCELLANEOUS PUBLICATION MUSEUM OF NATURAL HISTORY

ceilings. The Commonwealth Depart-
ment of Natural Resources (DNR) main-
tains a cave inventory and numbering
system compiled through the efforts of
various speleological groups, but the list
is incomplete due to the tremendous
number of caves. Reference will be
made to the DNR numbers whenever
these are known to be applicable.

Purely for descriptive purposes and
ease of reference, names are provided
for the major fossiliferous caves. These
are often whimsical appellations that
were applied for convenience in the
field, mostly by the Smithsonian expedi-
tion. Local names for caves are unre-
liable for subsequent relocation of sites.
Different individuals often have different
names for the same cave, the names may

Fic. 6.—A ground level entrance to a cave near
Barrio de Barahona, Puerto Rico. Dr. Richard
Thomas poses in the foreground.

change through time, and there is ex-
tensive duplication of names, for example
the long list of caves called “Clara” or
“Golondrinas.”

Most fossils were collected in soil
from the floors of the caves in dry, loose,
unconsolidated substrate. The bones ac-
cumulating as owl pellets are deposited
on the floor usually next to a wall, pre-
sumably below the owl’s roost. When
undisturbed, the fossils and soil form a
cone-shaped earthen mass resting against
the wall. Bones often can be found to
depths of several feet, but usually they
disappear at the level of the underlying
breccia. The cones are excavated and
the material is passed through a screen
mesh. Typically, cave soil is very fine,
and the matrix nearly sifts itself leaving
the fossils behind. Vertebrate bone is
found in association with a variety of
snail shells and claws of land crabs. In
some West Indian caves, vertebrate bone,
including Rattus, is associated with hu-
man artifacts in kitchen middens. These
deposits are obviously more recent in
age than the others (e.g., see Wing,
Hoffman and Ray, 1968). Fossils are
disassociated and often fragmentary. It
is remarkable, however, how many are
intact, particularly smaller elements of
lizards and frogs. Most fossils are not
difficult to prepare save for their small
size. The matrix is largely sand and the
material can be removed with a sharp
probe. Some bones are encrusted with
travertine and are permineralized; acid
preparation will remove some of this
kind of matrix.

The first collections of fossil verte-
brates in Puerto Rico were made in 1916
by H. E. Anthony of the American Mu-
seum of Natural History. A copy of
Anthony’s field notes was kindly pro-
vided to Storrs Olson by Karl F. Koop-
man. Three of Anthony’s major sites
(Cueva Catedral, Cueva Clara and
Cueva San Miguel) were located in the
center of the island near the town of
Morovis (erroneously accented “Morvis”
by Wetmore, 1922) and immediately

PLEISTOCENE PUERTO RICAN HERPETOFAUNAS 11

south of the present village of Barahona.
In September 1957, James W. Bee of
the University of Kansas Museum of
Natural History collected fossils in the
same valley system south of Barahona in
three caves, all apparently different from
any of Anthony’s (although one of these
was also called “Cueva Clara”). Bee’s
specimens, including the reptiles and
amphibians used in this study and the
bats and insectivores described by
Choate and Birney (1968), are housed
in the University of Kansas Museum of
Natural History collection of vertebrate
paleontology.

In May 1976, an expedition under
the auspices of the Smithsonian Institu-
tion collected vertebrate fossils in a
variety of localities, but mainly in the
above mentioned valley system south of
Barahona. Collecting was done by Storrs
Olson, Noel Snyder, and Frederick V.
Grady, who were assisted in the early
stages by Herbert Raffaele, Melvin Ruiz,
and John Taapken. Later in 1976 (June
15-22), Snyder, along with Dwight
Smith, Carlos Delannoy, José Riveira,
Herbert Raffaele, Melvin Ruiz, and Joan
Duffield, made additional collections
from some of the same caves and also
discovered a number of new sites. Sub-
sequently, I visited the island in January,
1978, and traced some of the Smith-
sonian expeditions and performed addi-
tional prospecting, occasionally accom-
panied by Richard Thomas of the
University of Puerto Rico. All of the
specimens collected on these expeditions
are housed at the National Museum of
Natural History, Smithsonian Institution
(USNM).

Caves in the Barahona Area.—Prob-
ably the most productive area for fossils
in Puerto Rico is just north of the center
of the island and lies at the southern
edge of the karst belt as depicted by
Monroe (1976, fig. 2). This is a valley
system immediately south of the village
of Barahona, 2 km NE of Ciales and
4.5 km NW of Morovis (Ciales Quad-
rangle, U.S. Geol. Surv. 7.5 minute topo-

graphic map, 1957). The approximate
locations of major fossil sites are shown
in Fig. 7. From Anthony’s (1916, 1918)
descriptions and photographs, Noel Sny-
der was able to locate Cueva San
Miguel, but Cueva Catedral and An-
thony’s Cueva Clara were never identi-
fied confidently. Using the maps and
descriptions in Bee’s field notes, Snyder
and Olson were able to correlate Bee’s
sites with caves they encountered in their
explorations. A brief description of the
most productive sites is given below.

Blackbone Cave.—In terms of species
diversity, this is by far the most pro-
ductive fossil site yet discovered in
Puerto Rico. It was found by the Smith-
sonian parties who collected there in the
spring of 1976. Olson and J. Philip Angle
returned to this site in 1977 and removed
approximately 135 kg of previously
screened (% inch mesh) dirt, which was
later passed through finer mesh with the
result that many very small vertebrates
were recovered that were not repre-
sented elsewhere. A sampling bias may
have been introduced by this procedure
(see Discussion).

Apparently, Blackbone Cave is known
locally as “Cueva del Infierno,” the name
Blackbone having been applied to it by
the Smithsonian expedition on account
of the exceptional coloration and preser-
vation of the fossils found there. The
cave is located 1.2 km due south of
Iglesia Ascension (18° 20’ 57” N; 66° 26’
47” W), as marked on the Ciales Quad-
rangle topographic map, and comprises
an extensive system that runs completely
through a ridge of limestone. There are
two entrances. The southern entrance is
situated in the west side of a small rock
shelter concealed in the wall of a large,
steep-sided sink. This entrance leads
through several dark chambers that ulti-
mately empty into a large, well-lighted
room with a collapsed roof. A passage-
way leads from this room to the second
entrance, which is at the edge of a culti-
vated field. On the northern side of the
rubble formed from the rocks of the

12 MISCELLANEOUS PUBLICATION MUSEUM OF NATURAL HISTORY

_ @.
So

“@™ Dump

Ciales

Escuela
Barahona

Perro @ | e@

Blackbone

Nesophontes
e

@ Bee's Clara
@san Miguel

To Morovis

Fic. 7.—Map of the Barahona area, Puerto Rico, with dark circles indicating the approximate
locations of the major fossiliferous caves. See text for details.

collapsed roof is a subchamber with a
small subsidiary passage at floor level,
and another one about 2.5 m off the floor.
Beneath the latter was a cone 25 cm
deep of loose, fossil-bearing sediment
located in a slight pocket in the cave
wall. There was no detectable stratifica-
tion in this cone, which rested on a layer
of similar, loose, but nearly barren ma-
trix about 20 cm deep. In tum, this
second layer lay on a hard, waxy, red
and black clay that was sterile except
for a few sloth ribs.

Thousands of bones were obtained
here. Their preservation is unusual
among West Indian cave deposits in that
the bones are black in color, heavily
permineralized, hard, dense, unleached
and perfectly preserved. In some in-
stances, ossified tendons were preserved
on the avian bones. That this deposit

was principally the work of owls is at-
tested to by the fossilized pellets that
were present, although most bones were
found loose and unassociated. This de-
posit was designated “Blackbone 1”
(Fig. 8).

About two to three meters to the
right of Blackbone 1, near the mouth of
the small, floor-level tunnel, additional
bones were found in the sediment on the
cave floor. This assemblage was called
“Blackbone 2.” There were fewer bones
here than in Blackbone 1, and _ their
preservation was different also, being
lighter in color and usually encrusted
with lime.

The collapsed roof of the cave must
be of some antiquity, as owls could not
have gained access to the large room
through either of the long, dark passage-
ways leading to it. Owl roosts in cave

PLEISTOCENE PUERTO RICAN HERPETOFAUNAS 13

situations, fossil or not, are almost always
near an entrance with easy access to the
outside.

Nesophontes Cave.—tThis is another
extensive cave system that runs entirely
through a hill located about .4 km SSW
of Blackbone Cave. Discovered by the
Smithsonian expedition, it received its
name for the tremendous numbers of
bones of the extinct insectivore Neso-
phontes that were recovered there. There
are at least three entrances to the cave

Pay phe

and the fossils were found immediately
inside the more southerly of these. Here,
about 2 m above the cave floor, there
was a deposit on one side of the entrance
that sloped up for more than a meter to
the outside. This deposit consisted of
loose, yellowish, dusty soil with an ex-
tremely dense concentration of bones.
Preservation was entirely different from
Blackbone Cave in that the bones were
light in weight and buffy colored with a
leached appearance. Compared to Black-

:
4
a

Fic. 8.—Storrs Olson (left) and Fred Grady screening fossil matrix inside Blackbone Cave. Photo-

graph by Noel Snyder.

14 MISCELLANEOUS PUBLICATION MUSEUM OF NATURAL HISTORY

bone Cave, there was a much greater
proportion of Nesophontes and bats, and
correspondingly fewer birds, reptiles,
and amphibians.

There was some evidence of former
excavation in this cave, as inside the
entrance there was a large, rocky pile of
what appeared to be tailings. It seems
unlikely that this was one of Anthony's
or Bee’s sites, however, because the cave
does not fit any of their descriptions and
neither collector obtained Nesophontes
in such quantities as were found here.
The site may have been dug by pot
hunters or others looking for indian arti-
facts. Also, many bat caves were a
source of fertilizer for local farmers. The
most accessible caves were mined out
long ago with the result that other sur-
face contents (including fossils) either
have been removed or greatly disturbed.

Bee’s Cueva Clara (Fossils collected
by Bee from this and his other two sites
described below were given University
of Kansas locality numbers; for Cueva
Clara it is KU-ZCA-02).—Directly across
from the south entrance of Nesophontes
Cave, across a small valley, is the en-
trance to another cave with a very high
ceiling. Inside, the cave slopes down for
a great distance, becoming danker and
danker and increasingly populated by
bats. As one progresses, there is no small
feeling of apprehension. Appropriately,
it was dubbed “Horrible Bat Cave” by
the Smithsonian expedition. Snyder’s
party in 1976 found another cave above
this one that became “Cave on Top of
Horrible Bat Cave.” Snyder believes
that it is the same as Bee’s Cueva Clara
(but not the Cueva Clara of Anthony).
Bee’s field notes describe this cave as
being high on a slope and having two
entrances. Guano diggers had been there
previously and removed most of the soil.
Snyder’s group recovered a few bones
here, preserved in the same way as those
from Bee’s Cueva Clara.

Cueva San Miguel.—This was one of
Anthony’s localities (see 1918, Plate 58,
Fig. 2), relocated and worked by Noel

Snyder's party in 1976. The site is lo-
cated just south of Bee’s Cueva Clara.

Cueva del Perro (KU-ZCA-01).—
This was one of Bee’s most productive
sites and was relocated by the Smith-
sonian parties in 1976. At this time noth-
ing of interest remained. The cave is
situated about .3 km WNW of Blackbone
Cave on the western side of a fairly deep
valley with a footpath and stream at the
bottom. The cave has two entrances.
The material Bee removed came from
a small, side chamber, with the greatest
amount of fossils being found against
the wall to a depth of 1.2 m.

Cueva de Silva (KU-ZCA-03).—This
is the third of Bee’s localities and is lo-
cated across the valley (east) from
Cueva del Perro, on a wall overlooking
the sink containing the south entrance
to Blackbone Cave. This cave was seen,
but not explored by me or the Smith-
sonian expedition. Bee’s notes indicate
that the cave soil here was different from
that at his other two sites and that he
did not collect here extensively.

Barahona IV Cave (18° 21’ 04” N;
66° 27’ 51” W).—The Smithsonian party
of May 1976 explored a number of caves
about 1 km WSW of Barahona in the
vicinity of a fairly large refuse dump.
These included, caves 26, 27, 29, and 30
of the DNR inventory. Little of signifi-
cance was recovered from any of these.
Farther to the west was a more pro-
ductive site, designated “Barahona IV,”
and believed to correspond to cave 32
of the DNR inventory. This is an open,
airy cave, actually more of a tunnel, with
two entrances that run through a hill.
Across the valley to the southeast is a
similar “through” cave believed to be
cave 31 of the DNR inventory. Fossils
at Barahona IV were concentrated mostly
in dry, reddish soil a few meters in from
the north entrance, where the cave be-
comes constricted and passes into the
larger south chamber. Olson, Snyder,
and Grady dug and screened here exten-
sively before the more productive Black-
bone and Nesophontes Caves were dis-

PLEISTOCENE PUERTO RICAN HERPETOFAUNAS 15

covered. Bones of extinct rodents were
the most frequently encountered fossils,
but also present were remains of sloths,
birds, and other vertebrates.

There are many lesser caves in the
vicinity of Barahona, in some of which
the Smithsonian expeditions and myself
found usually insignificant samples of fos-
sil bones (“High Cave,” “Nancy Robles
Cave,” “Dwight’s Living Heart Sacrifice
Cave,’ “My Cave,” “Condor Cave,” etc.).
These have not been mapped in Fig. 7.

OTHER FOSSIL SITES

Cueva Arcilla—This is cave 39 of
the DNR inventory, located east of
Morovis near a small road leading off
of Route 159 just beyond the 3.8 km
marker. The cave is in a wall at the
edge of a banana patch and immediately
below a house. It is high and relatively
open, sloping up and back for about 18
m. The Smithsonian party of May 1976
obtained a few bones in the red clay
along one wall, mostly of the extinct
rodent Elasmodontomys.

Toranio Cave (18° 18’ 05” N; 66° 44’
38” W; Utuado Quadrangle ).—This
was another of Anthony’s (1916, 1918)
sites located by the Smithsonian expedi-
tion, which found only a few fossils
there. It is cave 65 on the DNR inven-
tory and is situated in a mogote in pas-
tureland, 5.7 km NW of Utuado near
the settlement of Cayuco. The entrance
is a very small opening in a hillside,
about 18 m above the surrounding pas-
ture. The cave drops off steeply into a
sink, necessitating the use of a rope to
descend into it. Anthony’s material came
from dry earth just inside the entrance,
whereas the few bones obtained by the
Smithsonian party came from a moist
pocket farther inside.

Cueva Soto, Rio Abajo (18° 20’ 13”
N; 66° 42’ 25” W; Utuado Quadrangle).
—This is cave 157 of the DNR inventory,
located 7.6 km north of Utuado in the
Rio Abajo Forest Reserve. The Smith-
sonian expedition dug in the white and
gray powdery sediments at the front of
the cave, but obtained only crab claws,

bat bones, and a few fossils of the ex-
tinct flightless rail, Nesotrochis debooyi.

Rio Camuy Cave (18° 19’ 23” N;
66° 49’ 27” W; Bayaney Quadrangle ).—
This is one of a series of caves men-
tioned by Monroe (1976) along the east
bank of the Rio Camuy in the northwest
part of the island, about .3 km SE of the
point where the river goes underground
off old Route 129. It is cave 60 of the
DNR inventory. The cave is an open
shelter with an entrance at either end.
The Smithsonian expedition found some
specimens near the floor surface adjacent
to the walls. The bones were relatively
recent in appearance and in association
with potsherd and other signs of human
occupation. Evidently, this was an ar-
chaeological site and was the only lo-
cality in which the extinct rodent Isolo-
bodon was encountered.

Rosario River Cave.—A very few
bones were recovered by Snyder’s party
from a cave at the western end of the
island near Mayagiiez, at the south end
of the Rio Rosario.

Gudnica Bat Cave (17° 57’ 40” N;
66° 50’ 53” W; Punta Verraco Quad-
rangle ).—Located .75 km north of Playa
de Tamarindo, this is the only significant
site discovered outside of the Barahona
area. It differs in being situated in the
dry coastal scrub of the southwestern
part of the island. The cave, hidden in
particularly dense thorn scrub, is entered
through a collapsed ceiling that opens
into a large, horseshoe-shaped cavern,
25 m high in some places. A number of
fossils were collected from a_ pocket
along the east wall in soft, powdery,
reddish-brown earth. The bones appear
to be more recent in age than from most
other sites and are not heavily mineral-
ized. Specimens were collected by Sny-
der’s group in 1976 and by me in 1978.

Age of the Deposits—One of the
major problems in West Indian paleon-
tology is dating fossiliferous cave ma-
terial. Precise ages of deposition are
poorly known. Stratigraphic sequences
in a given cave may be present, for ex-

16 MISCELLANEOUS PUBLICATION MUSEUM OF NATURAL HISTORY

ample a surface layer of 20 cm may be
lighter in color than a 20 cm layer below
it, but correlation of layers from differ-
ent caves is meaningless. Hence, fossils
have usually been assigned ages such as
latest Pleistocene, pre- or post-Colum-
bian, or SubRecent. In some deposits a
surface layer of several inches may con-
tain remains of post-Columbian exotics
like Rattus, Mus, and Herpestes. Deeper
layers may be characterized by color or
“index fossils” peculiar to the cave; one
may discuss the lizard layer, Rattus
layer, Nesophontes layer and so forth.
However, I think that this assessment is
often artifactual because some collectors
in the past removed bones selectively.

Hitherto, there have been no attempts
made to obtain dates radiometrically or
by other sophisticated modern tech-
niques. Difficulties do arise from the
lack of appropriate material for dating.
Shells of land snails collected from the
more productive deposits sampled by
the Smithsonian expeditions were sub-
mitted by Olson for C14 dating to the
Smithsonian Radiation Biology Labora-
tory, where they were analyzed by Dr.
Robert Stuckenrath. Snail shells present
a special problem in that the carbon used
in analysis comes from calcium carbo-
nate of the shells, which may be con-
taminated by carbonate from extrinsic
sources, especially in a limestone en-
vironment.

As detailed in the Discussion, there
is faunal evidence that the cave deposits
from the Barahona area are not all con-
temporaneous, and if one had to choose
a particular deposit that seems unques-
tionably older than the others it would
be Blackbone 1. The other sites are
more similar to one another faunally as
is the preservation of the fossils, which
is not markedly different from that seen
in other West Indian deposits. Of two
samples of snail shells from Nesophontes
Cave, one gave a C1 date of 28,300 +
530 years B.P., whereas the other yielded
a date greater than 43,000 years B.p., be-
yond the effective range of C'* dating.

Note that these dates are not consistent
with each other. Snail shells from Cueva
San Miguel gave a C!* date of 35,000 +
850 years B.e. These all seem widely off
the mark and it should be remembered
that the bones from the deposits in these
two caves appear to be leached.

Three samples of snail shells from
Blackbone 1 yielded more believable re-
sults. Two samples from the upper 20
cm of the deposit gave dates of 17,030
+ 160 years B.p. and 18,690 years B.P.,
respectively. One sample from the sec-
ond 20 cm layer gave a C' date of
21,400 + 330 years B.p. These dates at
least have some stratigraphic consis-
tency, and it is possible that because of
the unique preservation of the fossils at
this site, the carbonate in the snail shells
had not been contaminated. A date of
17 to 20 thousand years before present
would mean that the Blackbone 1 de-
posits were formed during the Wisconsin
glaciation when climates and habitats
were altered in the West Indies, such
that conditions were generally drier.
This, too, accords well with the fauna
found in the Blackbone 1 deposits (see
Discussion).

Tooth fragments consisting mainly of
enamel of the rodent Elasmodontomys
from Barahona IV Cave yielded a C14
date of 13,080 + 335 years B.e. Probably
this cannot be relied upon very confi-
dently, however, as it has been deter-
mined that enamel, dentine, and ce-
mentum from the same tooth can each
give different C' dates.

Amino acid racemization analysis can
provide useful information on relative
ages, but samples of bone and snail shells
from the above sites, submitted by Olson
to Dr. Ed Hare of the Carnegie Institu-
tion of Washington, were too old to be
within the effective resolution of the
technique.

In summary, the fossil deposits from
Puerto Rico indicate that there can be
considerable variation in age of deposi-
tion and that these deposits may be con-
siderably older than many such deposits

PLEISTOCENE PUERTO RICAN HERPETOFAUNAS 17

are thought to be. As a tentative work-
ing base, the Blackbone 1 fossils are
probably about 20,000 years old, whereas
the other Puerto Rican deposits are
younger.

SYSTEMATIC PALEONTOLOGY
Amphibia
Anura
Leptodactylidae

Two genera of Leptodactylid frogs
occur on Puerto Rico—Leptodactylus
with one species (L. albilabris) and Eleu-
therodactylus with 16 species. Both gen-
era are represented by fossils of cranial
and post-cranial elements. The diag-
nostic features of each genus are pro-
vided by Lynch (1971) and Heyer
(1969b ), and fossils of the two are easily
distinguishable in this sample. The bones
referred to L. albilabris are so attributed
on the basis of their similarity to the
same elements in modern individuals,
and the fact that L. albilabris is the only
representative of the genus on the island.
Because osteological descriptions are fa-
cilitated by comparison, L. albilabris is
described with occasional reference to
Eleutherodactylus coqui, a frog nearly
comparable in size.

Leptodactylus albilabris Ginther

Material—Blackbone 1: otoccipitals
(4—USNM 259011); sphenethmoids (5
—USNM 259012); maxillae (4 right, 5
left—USNM 259017); mandibles (9
right, 5 left—USNM 259018); scapulae
(6 right, 7 left—USNM 259019); ilia
(24 right, 19 left—USNM 259013-14);
atlases (4 + 1 fragment—USNM 259016);
presacral vertebrae (66—USNM 259016).

Blackbone 2: otoccipitals (1—USNM
259020); mandibles (1 partial left—
USNM 259021); ilia (2 right, 2 left—
USNM 259022); urostyles (2—USNM
259022).

Nesophontes Cave: ilia (5 right, 9
left—USNM 259025).

Barahona IV: scapulae (1 left—

USNM 259024); ilia (1 left—USNM
259023).

San Miguel Cave: pelvic girdle (1
complete—USNM 259026).

Description.—Otoccipital: the four
otoccipitals from Blackbone 1 are com-
plete; the specimen from Blackbone 2
is a left half. In this species the exoc-
cipitals and cristae prooticae are narrow,
whereas in Puerto Rican Eleutherodacty-
lus they are proportionately twice as
wide. The occipital condyles are ellip-
tical and widely spaced. In three of the
five fossils, a T-shaped parasphenoid is
fused to the ventral surface of the proo-
tic. The exoccipital crests are robust and
best developed in the larger specimens.
The measurements of the four complete
otoccipitals at their widest points, across
the tips of the cristae prooticae, are
GH al 2aliayien No.8!

Sphenethmoid: five fossil spheneth-
moids range in size from 5.2 to 7.0
(x = 64) measured midsagitally. The
posterodorsal edge of the sphenethmoid
is deeply emarginate with a small pos-
terior projection at the center of the
edge. The bone is a hollow box, smooth
and flat on the top and bottom. The
sides are flared anterolaterally to the
level of the palatine articulation. The
anterior end tapers abruptly to a blunt
point. A bony septumnasi divides the
sphenethmoid anteromedially. Each na-
sal cavity is partially divided by a me-
dially curving septum that arises from
the ventral floor. This septum is absent
in Puerto Rican Eleutherodactylus.

Maxilla: none of the fossil maxillae
is complete, however, each bears teeth
which distinguish it from the maxilla of
Puerto Rican Eleutherodactylus. The
teeth of L. albilabris are larger and more
widely spaced. Also, the premaxillary
process of the maxilla is expanded dor-
sally as a thickening of the pars facialis
and bears a short peg in contact with
the lateral side of the premaxillary bone.
In Eleutherodactylus coqui the anterior
end of the maxilla is simply rounded.

Mandible: 14 mandibles, most of

18 MISCELLANEOUS PUBLICATION MUSEUM OF NATURAL HISTORY

them missing the symphyseal end, are
referred to this species on the basis of a
conspicuous coronoid-like process just
anterior to the articular surface on the
back of the bone. In Eleutherodactylus
this process is indistinct from the articu-
lar surface. The mandibles of both spe-
cies are smooth and otherwise nearly
identical.

Scapula: the scapula of L. albilabris
is wide and not nearly as constricted in
the middle as those of other Puerto
Rican frogs. The coracoid articular facet
is broadly expanded proximally and
raised ventrally. The total length of 14
scapulae ranges between 4.6 and 6.0
(#=49).

Ilium: numerous ilia are assigned to
L. albilabris on the basis of the shape
of the dorsal prominence. This structure
is a large, semicircular expansion at the
proximal end of the dorsal crest (Fig. 9).
The crest and prominence are com-
pressed in a vertical plane. A rugose,
dorsal protuberance covers the posterior
half of the prominence on the lateral
surface. The ilial shaft is broken at
various places distally on most of the
fossils, although when complete the shaft
is grooved over most of the length of
the medial surface and rounded below
the groove. Overall, the shaft is gently
recurved. The ventral acetabular ex-
pansion is smaller than in Eleuthero-
dactylus coqui and the preacetabular
zone (sensu Lynch, 1971:61) is very
narrow.

Vertebrae and vertebral column: L.
albilabris has eight imbricate presacral
vertebrae and all but Presacral VIII bear
a distinct, posteriorly directed neural
spine (Fig. 10). Neural arches are nearly
as long as wide. Transverse processes
are wide and expanded on Presacrals II,
III and IV, but are narrow and unex-
panded on the remaining presacrals. The
sacral diapophyses are elliptical in cross
section, and dialated distally. The sacral
diapophyses are directed caudally from
the midline of the vertebral column. The
two cotyles that receive the urostyle are

Fic. 9.—Fossil ilia of Leptodactylus albilabris
(top, USNM 259014); Eleutherodactylus sp.
(middle, USNM 259035); Peltophryne lemur
(bottom, USNM 259052). Scale equals 5 mm.

separate from one another, whereas in
E. coqui the cotyles usually are medially
confluent. The smallest of eight sacral
vertebrae is 5.0 across the diapophyses
and the largest is 8.0 (x = 6.7).

Comments: Leptodactylus is poorly
represented in the West Indies. Schwartz
and Thomas (1975) included five species
in their checklist, but only two of these
occur in the Greater Antilles—L. domini-
censis (Hispaniola) and L. albilabris
(Puerto Rican Bank). Three species,
L. fallax, L. insularum and L. wagneri
range erratically throughout the Lesser
Antilles. The relationships of Antillean
Leptodactylus with other members of
the genus is unclear. Heyer (1969a,
1970) placed L. fallax in the pentadac-
tylus group, L. insularum in the ocellatus

PLEISTOCENE PUERTO RICAN HERPETOFAUNAS 19

group and L. wagneri in the melanonotus
group. Heyer referred to the two Greater
Antillean forms as L. mystaceus, thereby
implying conspecificity of L. albilabris
with L. dominicensis, but he did not dis-
cuss the reasons for his decision. On
Puerto Rico and the Virgin Islands, L.
albilabris is ubiquitous, occurring any-
where suitable moisture exists.

A minimum of 33 individuals is rep-
resented by fossils and 24 of these are
from Blackbone 1 alone. The most
curious feature of the fossil frogs is their
large size. Adults may have reached
sizes up to 50% larger than any individual
known today. The mean and range of
estimated snout-vent lengths of the fossil
individuals, based on five different skele-
tal elements, are as follow: otoccipital
width 67.3 (53.5-73.0); sphenethmoid
length 74.9 (61.0-82.0), sacral vertebra
width 45.2 (34.0-58.5), acetabular diam-
eter 57.0 (38.0-79.0), scapular length
56.0 (46.0-67.0). Small sample size of
some elements and individual variation
and allometry undoubtedly influence
these estimates. For example, in other
anurans, Trueb (1977) concluded that
sphenethmoid length is a rather variable
character in a population of Hyla lanci-
formis. Nonetheless, at least some of
the fossil individuals obtained snout-vent
lengths well in excess of modern indi-
viduals. Of 50 specimens examined by
Schmidt (1928), the largest was 49.0
SVL. One fossil otoccipital came from
an individual over 80.0 SVL.

Eleutherodactylus spp.

Material—Blackbone 1: otoccipitals
(6—USNM 259027); sphenethmoids (6
—USNM 259028); maxillae (30—USNM
259030); mandibles (42—USNM 259029);
scapulae (8 right, 12 left—USNM
259040); ilia (45 right, 39 left—USNM
259031-6); vertebrae (6 atlases, 204 pre-
sacrals, 36 sacralk—USNM 259037-8);
urostyles (6—USNM 259039).

Blackbone 2: ilia (1 left—USNM
259042).

San Miguel Cave: vertebrae (1 pre-
sacral—USNM 259043).

Description—tThe fossils referred to
Eleutherodactylus are quite similar in
general morphology to those of Lepto-
dactylus, and the features distinguishing
the two are pointed out in the descrip-
tion of the latter. However, a few per-
tinent points remain. For example, on
the average the ilia of Eleutherodactylus
are much smaller than those of L. albi-
labris. The ilia lack a conspicuous dor-
sal prominence, but retain a small
rounded dorsal protuberance (Fig. 9).
The distal ends of the ilial shafts are
broken from most of the fossils, but
otherwise they are unremarkable. The
eight nonimbricate presacral vertebrae
lack neural spines (Fig. 10). The diapo-
physes of the sacral vertebrae are straight
and not dilated distally. A transverse
ridge is present over the neural arch of
the sacral vertebrae, but it does not ex-
tend onto the diapophyses as it does in L.
albilabris. Measurements of some of the
Eleutherodactylus fossils are listed in
Table 1.

Comments.—The remarkable osteo-
logical similarity of Puerto Rican Eleu-
therodactylus precludes specific identifi-
cation of the fossils. Perhaps with large
skeletal series of both sexes of all species
this would be possible to some degree.

Puerto Rico’s 16 species of Eleuthero-
dactylus inhabit the island from sea level
to the rain forests at higher altitudes.
With the exception of the widespread
E. coqui, most species are localized in
distribution. They are mostly small frogs
and as adults have a snout-vent length
from 30.0 to 35.0. Eleutherodactylus
coqui reaches a maximum snout-vent
length of 60.0 (Thomas, 1965) as does
the rain forest stream denizen, E. karl-
schmidti. The snout-vent lengths of the
fossil individuals vary considerably. For
the following skeletal elements the esti-
mates are: sphenethmoid 47.0 (21.0-
60.0), ilia 30.0 (18.0-51.0), otoccipitals
42.0 (28.0-55.0), sacral vertebrae 32.0
(15.0-53.0). It is difficult to interpret

20 MISCELLANEOUS PUBLICATION MUSEUM OF NATURAL HISTORY

Fic. 10.—The vertebral columns of three Puerto Rican frogs: A. Peltophryne lemur; B. Lepto-
dactylus albilabris; C. Eleutherodactylus coqui.

these values because of the small size of
the bones and the error inherent in such
estimates. The largest individuals are
probably E. coqui or E. portoricensis.
The smaller bones may represent com-
mon lowland species such as E. antillen-
sis or E. cochranae. It is possible that all
of the fossils represent one species, al-
though this seems unlikely. In the field,
four or five size classes are encountered
with nearly equal frequency in winter
populations of E. coqui (pers. observ.)
and these size classes encompass the
range of snout-vent length estimates of
the fossils. However, larger, adult indi-
viduals are more likely to be seen and
preyed upon than juveniles and smaller
individuals and, hence, a fossil assem-
blage accumulated through predation
would not be likely to contain equal

numbers of juveniles, subadults, and
adults.

Bufonidae
Peltophryne lemur Cope

Syn: Bufo lemur Cope.

Material.—Blackbone 1: neurocrania
(5 right, 6 left—USNM 259044); sphe-
nethmoids (4—USNM 259045); nasals
(3 right, 2 left—USNM 259046); squa-
mosals (4 right, 1 left—USNM 259047);
maxillae (11 right, 9 Jeft—USNM
259048); mandibles (2 right, 6 left—
USNM 259049); scapulae (6 right, 12
left—USNM 259050); suprascapulae (2
right—USNM 259053); ilia (10 right, 21
left—USNM 259051-2); presacral verte-
brae (65—USNM 259054-5); sacral ver-
tebrae (9—USNM 259056); urostyles
(9—259057).

Blackbone 2:

neurocrania (1 right,

PLEISTOCENE PUERTO RICAN HERPETOFAUNAS 21

TABLE 1.—Selected measurements (mm) of cer-
tain fossil elements of Eleutherodactylus spp.

nm & Range

Otoccipital width’ 6 8.7 5.8-11.5
Sphenethmoid: 6

Midsagittal length 4.0 1.8- 4.5

Anterior width 46 2.0- 6.3
Tlia: 76

Acetabular height? Bi I= ALO
Sacrum: 28

Width across diapophyses 49 2.3- 8.1

* Measured across tips of cristae prooticae.
* Measured between tips of dorsal and ventral
acetabular expansions.

1 left—USNM 259058); parasphenoids
(2—USNM 259060); maxillae (4 right,
1 left—USNM 259059); mandibles (1
right, 3 left—USNM 259061); scapulae
(1 left—USNM 259062); coracoids (1
left—USNM 259062); ilia (9 right, 12
left—USNM 259063-4); presacral verte-
brae (9—USNM 259066); urostyles (6—
USNM 259065).

Guanica Bat Cave: maxillae (1 left
—USNM 259067); ilia (1 right, 1 left—
USNM 259068); presacral vertebrae (2
—USNM 259070); sacral vertebrae (2—
USNM 259070); urostyles (1—USNM
259069).

San Miguel Cave: maxillae (3 left—
USNM 259071); presacral vertebrae (2
—USNM 259072); urostyles (1—USNM
259073).

Rio Camuy: maxillae (1 right—
USNM 259075); ilia (1 right, 1 left—
USNM 259076); presacral vertebrae (2
USNM 259076).

Nesophontes Cave: maxilla (1 right
—USNM 259074).

Description—Skull: the skull of Pel-
tophryne lemur is well ossified (Fig. 11)
and, as a result, many cranial elements
are preserved as fossils. The roofing
bones are easily distinguishable from
those of any other Puerto Rican anuran
by their thickness, sculpturing, and
prominent canthal, supraorbital and su-
pratympanic crests. Moreover, the max-
illary arch bears a pair of “rostral bones”
anteriorly which exclude the premaxillae
from the tip of the snout. These rostral
structures, unique to P. lemur and the

other eight species of endemic West In-
dian bufonids, are described in detail
elsewhere along with the rest of the
skull (Pregill, 1981).

Pectoral girdle: scapulae are abun-
dant as fossils. They are robust struc-
tures, proximally bicapitate and approxi-
mately the same length as the clavicles.
A single coracoid from Blackbone 2 and
two suprascapulae from Blackbone 1 are
the only other pectoral girdle elements
found in the remains. The suprascapula
is a more or less rectangular, ossified
bone with the cleithrum indistinguish-
ably fused to the posteroventral edge.

Ilia: the ilial shaft of P. lemur is a
recurved, flattened cylinder bearing a
shallow groove on the distal three-fourths
of the dorsomedial surface. The ventral
acetabular expansion is nearly 90° to
the shaft, but the smaller dorsal acetabu-
lar expansion is at a less acute angle
(Fig. 9). Both the dorsal prominence
and dorsal protuberance are well devel-
oped. Most of the shafts are broken
distally on the fossils. The largest com-
plete ilium is 31.3 measured from the
dorsal acetabular expansion to the end
of the shaft.

Vertebrae and vertebral column: Pel-
tophryne lemur has eight imbricate pre-
sacral vertebrae (Fig. 10). The trans-
verse processes are wide on Presacrals
II, III and IV, and thin and narrow on
Presacrals V, VI, VII and VIII. Neural
arches are broad and emarginate on the
posterior edge. Low, short neural spines
are present on the neural arches of the
posterior four or five presacrals. On the
anterior presacrals except the atlas, the
neural spines terminate as wedge-shaped
caps that cover the neural arches. The
sacral vertebra has broadly expanded
diapophyses and a bicondylic articula-
tion with the urostyle.

Comments.—The abundance of Pel-
tophryne lemur as a fossil is striking by
contrast to its present scarcity. Schmidt
(1928) secured only five specimens for
his synopsis, and Grant (1932) reported
that only 28 individuals had been col-

22 MISCELLANEOUS PUBLICATION MUSEUM OF NATURAL HISTORY

lected since the species was first de-
scribed by Cope in 1868. As a result,
nothing is known of its habits. The toad
occurs island-wide at low elevations on
Puerto Rico; although there are records
of this species from Virgin Gorda, British
Virgin Islands, it has not been seen out-
side of Puerto Rico for many years. The
toad is not abundant anywhere in its
range; Richard Thomas informs me, how-
ever, that the animal is probably more
common than currently believed because
of inappropriate collecting methods. Its
presence in five of the fossil localities
suggests that formerly P. lemur was more
abundant than it is today. Probably the
animal leads a secretive, semifossorial
existence, a lifestyle which could ac-
count for its rarity in museum collec-
tions. Obviously, it must spend some
time above ground exposed to predation
(e.g. while breeding) in order to account
for the frequency of its occurrence as a
fossil. Schmidt (1928) suggested that
the introduction of Bufo marinus to the
island in the early 1920’s might account
for the rarity of the species today. Bufo
marinus is well established in disturbed
and undisturbed areas over most of the
island and, although it is somewhat
larger, by Schmidt's reckoning B. ma-
rinus may be displacing the endemic.
This seems unlikely to me.

TasLe 2.—Selected measurements (mm) of cer-
tain fossil elements of Peltophryne lemur.

n x Range

Neurocranium:

Greatest length* 9 14.8 13.0-16.8
Maxilla:

Greatest length 19 22.3 12.2-33.7
Nasal:

Greatest length 2 16.4 16.4-16.4
Mandible:

Greatest length 5 20.1 15.7-22.0
Scapula:

Greatest length 19 91 £7.1-12.8
Sacrum:

Width across

diapophyses Il 13.7 9.7-19.5
Acetabular diameter 44 68 4.8— 9.2

*Measured along medial border.
* Measured from tips of dorsal and ventral
acetabular expansions.

The phylogenetic relationships of the
nine endemic species of West Indian

Fic. 11—The skull of Peltophryne lemur

(USNM 27150) shown in dorsal view (top),

ventral view (middle), and lateral view (bot-
tom). Scale equals 5 mm.

PLEISTOCENE PUERTO RICAN HERPETOFAUNAS 23

toads were unclear until recently when
they were shown to be a natural, mono-
phyletic group, united by a uniquely
derived suite of characters that have re-
sulted from dermal co-ossification of the
skull, notably the rostral elements at the
front of the maxillary arch, and the di-
rect articulation of the zygomatic ramus
of the squamosal with the maxillary bone.
Consequently, these nine species, catau-
laciceps, empusa, gundlachi, longinasa,
peltocephala and taladai (Cuba), flu-
viatica, guntheri (Hispaniola) and lemur
were given generic recognition for which
the name Peltophryne (Fitzinger, 1843)
was available (Pregill, 1981).

Peltophryne lemur is one of the larg-
est species of Antillean toads. The two
largest species (exclusive of Bufo ma-
rinus) occur on Cuba. Peltophryne pel-
tocephala attains a maximum size of 170
snout-vent length and P. taldai reaches
147 (Schwartz, 1960; Ruibal 1959).
Schwartz (1972) gave a figure of 60 for
the maximum size of P. lemur, although
one female that I measured from the
National Museum of Natural History
(USNM 27148) has a snout-vent length
of 80. The snout-vent length estimates
of the fossil individuals range from 43
to 115 and average 80. The estimates
are computed from several skeletal ele-
ments as follow: neurocrania 67 (59-
76), maxillae 76 (43-106), mandibles 83
(64-90), scapulae 88 (73-115), ilia 86
(61-115), sacral vertebrae 74 (52-105).
Thus, some members of the fossil popu-
lations attained a greater maximum size
than any individuals known today. Table
2 lists other measurements of some of
the fossils.

Reptilia
Testudines
Emydidae

Chrysemys cf. decussata Gray

Material—San Miguel Cave: hypo-
plastron (1 right—USNM 259077).

Description—The badly worn plas-
tron element referred to this turtle is

more or less square and missing most of
the lateral margin of the ascending
bridge. The bone is 68 long and 64
wide. It came from an individual with
an estimated plastron length of 220.

Comments.—Turtles are perhaps the
least-studied vertebrates in the West
Indies. Their systematic relationships
were last treated comprehensively by
Barbour and Carr (1940) who recog-
nized six species then assigned to Pseu-
demys. Later, Williams (1956) briefly
discussed the West Indian terrapins and
placed them in the ornata subseries.
Five species now are listed as indigenous
to the Antilles (Schwartz and Thomas,
1975), but recently Schwartz (1978)
suggested that all of them may be con-
specific.

Chrysemys decussata stejnegeri oc-
curs throughout Puerto Rico in suitable
aquatic habitat. The species also in-
habits Cuba, the Cayman Islands, and
Hispaniola. The animal is active on
land, but its occurrence as a fossil is
probably accidental rather than a result
of predation.

Sauria
Gekkonidae

Sphaerodactylus sp. indet.

Material—Blackbone 1: frontals (2
—USNM 259078).

Description—Two complete frontal
bones from Blackbone 1 are the only
gekkonid skeletal elements present from
the entire fossil assemblage. They are
identified by their shape and very small
size. The descending laminar processes
are fused medially to enclose the olfac-
tory canal as in other gekkos. The two
frontals are 2.3 and 2.5 in length, re-
spectively. The supratemporal processes
diverge posterolaterally from the parietal
border and are 1.8 and 1.9 from tip to
tip in each bone. Both specimens are
constricted in the middle to a width of
0.5, but expand to 1.1 and 1.2 across
their respective anterior borders. The
nasal processes are short pegs projecting

24 MISCELLANEOUS PUBLICATION MUSEUM OF NATURAL HISTORY

craniad from the lateral edges of the
anterior border.

Comments.—Currently, five species
of Sphaerodactylus inhabit Puerto Rico
—S. macrolepis, S. nicholsi, S. klauberi,
S. roosevelti, and S. gaigae (Thomas and
Schwartz, 1966). Two of these, S. mac-
rolepis and S. klauberi, are fairly wide-
spread. Commonly, all are collected near
underbrush, and by raking leaf litter on
forest floors. These lizards are small,
ranging in size from 23 to 35 as adults.
The snout-vent length of the individu-
als represented by fossils is approxi-
mately 30.

Iguanidae
Anolis

The genus Anolis is distinctive osteo-
logically in ways that make identifica-
tion of fossils to the generic level un-
complicated in most instances. However,
at the species level the task is anything
but routine. Skeletal differences of
closely related species are often subtle
and qualitative when they exist at all.
Thus, identification requires repeated
comparison of skeletal series. Size classes
when discrete are a convenient and usu-
ally reliable means of reducing the num-
ber of potential species to which a given
fossil bone could be referred. Accord-
ingly, the many fossils of this lizard are
sorted into three groups—bones from
lizards with a snout-vent length over 100,
those from 60 to 70 and those less than
60, typically between 45 and 55. Anolis
cuvieri is the only Puerto Rican anole to
exceed a snout-vent length of 100. The
intermediate size class includes Anolis
cristatellus, Anolis cooki (Anolis cris-
tatellus cooki Grant) and Anolis gund-
lachi. The fossils of this size-group are
referred to A. cristatellus for reasons
pointed out below. Six of Puerto Rico’s
ten species of Anolis are less than 60 as
adults and are the most difficult to dis-
tinguish osteologically. Certain features
distinguish some of these, but most of
the material from small anoles is lumped
as Anolis spp.

Anolis cuvieri Merrem

Material—Nesophontes Cave: den-
taries (10 right, 10 left—USNM 259092-
4); articular + surangular (1 right, 7
left—USNM 259095); maxillae (12 right,
10 left—USNM 259096-7); prefrontals
(1 left—USNM 259099); frontals (4—
USNM 259098); basales (3—USNM
259100); pelves (4 right—USNM 259101);
vertebrae (6 dorsal, 1 sacral, 10 caudal—
USNM 259102).

Cueva del Perro: dentaries (2 right,
3 left—KUVP 11472); maxillae (4 right,
3 left—KUVP 11473); frontals (1—
KUVP 11473); parietals (3—KUVP
11473); basales (2—KUVP 11472); ptery-
goids (1 left—KUVP 11473); pelves (12
right, 11 left—KUVP 11475); vertebrae
(3 dorsal, 2 caudal—KUVP 11475).

Barahona IV: dentaries (2 right, 3
left—USNM 259111); articular + suran-
gular (1 right, 1 left—USNM 259113);
maxillae (1 right, 3 Jleft—USNM
259112); pyrefrontals (2 left—USNM
259115); frontals (2—USNM 259114);
basale (1—USNM 259118); quadrates
(2 left—USNM 259115): jugals (1 left
—USNM 259115); scapulae (3 left—
USNM 259116); pelves (1 left—USNM
259117); vertebrae (9 dorsal, 11 caudal
—USNM 259119).

San Miguel Cave: dentaries (1 left
—USNM 259104); articular + surangular
(2 right, 1 left—USNM 259105); maxil-
lae (1 left—USNM 259106); frontals (1
—USNM 259107); parietals (1—USNM
259108); jugals (1 right—USNM 259108);
basioccipital (1 fragment—USNM
259109); pelves (2 left—USNM 259110).

Blackbone 1: dentaries (3 right frag-
ments—USNM 259079); articular + su-
rangular (1 right, 1 left—USNM 259080);
maxillae (1 right, 1 left—USNM 259081);
prefrontals (1 left—USNM 259082);
postorbitals (1 right—USNM 259083);
jugals (1 left—USNM 259082); pelves
(1 right, 3 left—USNM 259084); dorsal
vertebrae (22—USNM 259085); caudal
vertebrae (6—USNM 259086).

Cueva Clara: dentaries (5 right, 7

PLEISTOCENE PUERTO RICAN HERPETOFAUNAS 25

left—KUVP 11526); articular + surangu-
lar (1 partial left—KUVP 11526); coro-
noids (1 left—KUVP 11526).

High Cave: dentaries (1 right, 1 left
fragment—USNM 259122); quadrates
(1 right—USNM 259123); jugals (2 left
—USNM 259123); pterygoids (1 partial
left—USNM 259124); pubis (1 left—
USNM 259124); vertebrae (3 dorsal—
USNM 259125).

Blackbone 2: dentaries (1 right frag-
ment—USNM 259087); articular + su-
rangular (1 right fragment—USNM
259087); maxillae (2 right—USNM
259088); pelves (1 right—USNM 259089).

Guanica Bat Cave: mandibles (1
right—USNM 259090); dentaries (2
right, 1 left—USNM 259091).

Cuevo de Silva: dentaries (1 right,
1 left—KUVP 11573); maxillae (1 right
KUVP 11573).

Rio Camuy: dentaries (2 left—
USNM 259120); maxillae (1 left—
USNM 259121).

Cueva Torano: dentaries (1 right—
USNM 259126); articular + surangular
(1 left—USNM 259126).

Rosario River: dentaries (1 left—
USNM 259103).

Robles’ Cave: dentaries (1 left frag-
ment—USNM 259126).

Description—Anolis cuvieri is dis-
tinguished from all other Puerto Rican
anoles by its larger size. Among other
West Indian giant anoles, it is the only
species lacking a splenial (Etheridge,
1959). Sexual dimorphism is pronounced
in A. cuvieri and in other giant anoles
as well. Rugose ornamentation covers
virtually all dorsal roofing bones of adult
males, whereas in females these bones
are smooth. The parietal of adult males
develops a horizontal triangular plate
fused to the underlying parietal crests.
The base of the triangle abuts the pos-
terior border of the frontal bone. In
adult females, the plate is absent and
the parietal is simply crested as in other
species of Anolis.

Dentaries of both sexes are long and
tapering, but in males the labial face is

scored by irregular folds and grooves
(Fig. 12). The posterior teeth of the
dentary and maxilla have truncated
crowns with weakly developed lateral
cusps in adults and juveniles of both
sexes. The anterior teeth are simple and
pointed. The angular and retroarticular
processes are broad and continuous with
one another, and extend laterally and
posteriorly well beyond the articular sur-
face for the quadrate. A small, hooklike
process is present on the lateral side of
the articular surface. Measurements of
the more abundant fossil elements are
listed in Table 3. The basale and frontal
are shown in Figures 12 and 14.

Comments.—Anolis cuvieri is an ar-
boreal lizard confined to Puerto Rico in
widely scattered localities from sea level
to 1200 m (Schwartz and Thomas, 1975).
Snout-vent length estimates made from
the fossils indicate that some individuals
from nearly all fossil localities achieved

Fic. 12.—Fossils of Anolis cuvieri: mandible

(top and middle, USNM 259090); basale in

occipital view (bottom, USNM 259107). Scale
equals 5 mm.

26 MISCELLANEOUS PUBLICATION MUSEUM OF NATURAL HISTORY

sizes greater than 140—the maximum
snout-vent length of Recent individuals.
The means and ranges of snout-vent
length based on five skeletal elements
are, as follow: dentaries 145 (103-175),
maxillae 138 (133-161), frontals 135
(127-161), basales 135 (113-165), ace-
tabula 180 (124-212).

Anolis cf. cristatellus
Dumeéril and Bibron

Material—Blackbone 1: dentaries
(5 right, 3 left—USNM 259127-8); ar-
ticular + surangular (4 right, 2 left—
USNM 259129); maxillae (1 left—USNM
259130); jugals (3 left—USNM 259132) ;
basales (3—USNM 259131); pterygoids
(1 right—USNM 259132).

Barahona IV: dentaries (1 right—
USNM 259134); articular + surangular
(1 right, 1 left—USNM 259134); maxil-
lae (1 left—USNM 259135); frontals (1
fragment—USNM 259135).

San Miguel Cave: mandibles (1 right
—USNM 259139); dentaries (3 left—
USNM 259138).

Nesophontes Cave: dentaries (2 left
—USNM 259136); maxillae (1 right, 1
left—USNM 259137).

TaBLe 3.—Selected measurements (mm) of cer-
tain fossil elements of Anolis cuvieri.

n x Range

Dentary:

Tooth row length 26 242 17.5-31.9

Number of teeth 26 25.9 22.0-31.0
Articular + surangular:

Greatest length 9 241 20.1-27.7

Length of retro. process 9 7.3 6.6— 9.2
Maxilla:

Tooth row length I] 22.4 21.5-26.1

Number of teeth Il 22.0 19.0-25.0
Frontal:

Midsaggital length 7 13.3 12.5=15.9

Posterior width 7 12.7 11.8-15.1
Parietal:

Greatest length 4 118 9.4-13.8
Basale:

Length’ 6 87 7.9- 9.2

Width? 6 14.6 12.3-17.6
Acetabular diameter 25 58 4.0- 6.8

*Measured from apex of angle formed by
pterygoid processes to the lip of the occipital
condyle.

? Measured across paraoccipital processes.

Cueva del Perro: mandibles (1 right
—KUVP 11473); dentaries (1 right, 1
left—KUVP 11473).

Blackbone 2: dentaries (2 right, 1
left—USNM 259133).

Description.—This fossil material be-
longs to a moderate-sized (60-70) anole,
but only the dentaries can be referred
to A. cristatellus with a high degree of
confidence. The dentary of A. cristatel-
lus is robust and convex midventrally.
In males of this species, the ventrolabial
face is sculptured with swollen irregular
bumps and excavations which are re-
placed by horizontal grooves posteriorly
(Fig. 13). The dentaries of females are
smooth. Teeth are straight, robust, and
closely spaced, numbering about 24. The
anterior teeth terminate as simple points,
whereas those posteriad have short, tri-
cuspid crowns. The transition from sim-
ple to tricuspid teeth occurs gradually
over three or four teeth between the
ninth and seventeenth teeth. Of 19 den-
taries referred to this species, the largest,
from Nesophontes Cave, has a tooth row
13.0 long and the smallest, also from
Nesophontes Cave, is 8.9. They came
from individuals with snout-vent lengths
of approximately 80 and 55, respectively.
The average tooth row length of all
fossil dentaries is 10.1—individuals with
a snout-vent length of about 62.

Maxillary teeth are like those on the
dentary, but the smallest teeth are in the
middle of the dental row. The basisphe-
noids of A. cristatellus are distinctive in
that the anterior border is curvilinear
and has a sharp median indentation. The
lateral edges of the pterygoid processes
continue caudally as ridges bordering
the sphenoccipital tubercles.

Comments.—I was unable to discern
any osteological features that would dis-
tinguish A. cristatellus from A. gundlachi
other than the dentary sculpturing and
shape of the basisphenoid. Adult males
of A. gundlachi also have a sculptured
lower jaw described and figured by
Etheridge (1959:123) as a row of from
7 to 10 transverse, semilunar notches on

PLEISTOCENE PUERTO RICAN HERPETOFAUNAS 27

the ventrolabial surface. This type of
sculpturing is found also on A. pulchel-
lus. None of the fossil dentaries referred
to A. cristatellus have this pattern, but
about half are intermediate between the
cristatellus type and the gundlachi type.
That is, semilunar notches are present
as in A. gundlachi, but the notches are
thrown into irregular folds and grooves
as in A. cristatellus. The dentary shown
in Figure 13 is of this intermediate type,
and in fact, the sculpturing on the fossils
is in general more robust and extensive
than on any recent skeleton that I ex-
amined.

One can only speculate as to the sig-
nificance of the sculpturing on the fossils.
Anolis cristatellus and A. gundlachi are
ecomorphic equivalents at low and high
altitudes respectively (Rand, 1964; Gor-
man and Hillman, 1977). The intermedi-
ate sculpture pattern on the fossils may
be from an ancestral population from
which A. cristatellus and A. gundlachi
only recently differentiated. Or possibly,
the pattern may occur in a population of
one of these species today, but repre-
sentatives were not present in my sample
of modern skeletons.

Anolis cooki ( Anolis cristatellus cooki
Grant ) is indistinguishable osteologically
from A. cristatellus.

Anolis evermanni Stejneger

Material—Blackbone 1:
(2 left—USNM 259140-1).

Description—The two dentaries re-
ferred to this species are 9.8 and 10.3
along the tooth row and have 24 and 27
teeth or empty alveoli, respectively. They
came from individuals with snout-vent
lengths between 55 and 60. Features of
the dentary which identify this lizard
are the elongate, linguilabially com-
pressed shape of the bone and the medial
constriction of the dental shelf (Fig. 13).
Anteriorly, the shelf is produced into a
thickened, convex border. The shelf nar-
rows posteriad gradually to the middle
of the tooth row where it dips ventrally.
Here the bases of the teeth are flush

dentaries

Fic. 13.—Fossil dentaries of 3 species of Anolis.

From top to bottom: A. cristatellus (lingual

and labial view, USNM 259128), A. evermanni

(lingual view, USNM 259140), A. occultus

(lingual view, USNM 259144). Scale equals
5 mm.

with the lingual edge of the shelf. The
dental shelf widens again posteriorly
and arches dorsad to the coronoid. In
other Puerto Rican anoles the dental
shelf is uniform and essentially hori-
zontal. Teeth of Anolis evermanni are
high and narrow, and closely spaced.
They have sharp, tricuspid crowns ex-
cept for the anteromost eight or ten.
These teeth are pointed.

Anolis krugi Peters
Material.—Blackbone 1: frontals (1
—USNM 259142).

Description—A single frontal bone
6.2 in length belongs to this species. The

28 MISCELLANEOUS PUBLICATION MUSEUM OF NATURAL HISTORY

frontal of A. krugi is unique among
Puerto Rican Anolis in that the parietal
border is convex (Fig. 14). The convex
border is received by a corresponding
indentation in the frontal border of the
parietal. The frontal-parietal suture is
straight in other Puerto Rican anoles.

Anolis occultus Williams and Rivero

Material——Blackbone 1: dentaries
(7 right, 5 left, 9 fragments—USNM
259143-4); maxillae (1 right, 3 left—
USNM 259146); frontals (2—USNM
259145).

Description —tThe diminutive size of
this lizard is one of its most obvious
characteristics. The very small dentaries
are long and narrow (Fig. 13); the 12
fossils have an average tooth-row length
of 5.6 and range from 3.6 to 7.2. The
smallest specimen has 16 teeth or empty
alveoli and the largest 23. The dental
shelf is broad and horizontal throughout
its length, with negligible dorsal curva-
ture at the coronoid end, as is the case
of other Puerto Rican species of Anolis.
The lower jaw of juvenile Anolis stratu-
lus is similar in size and shape, but the
bone of this species is more tubular and
the teeth are higher and narrower with
better developed tricuspid crowns. Teeth
of Anolis occultus are small, somewhat
constricted at their bases, and inflated
in the middle. Most teeth terminate in
simple blunt points, although the pos-
terior third of the dental series have a
faint suggestion of tricuspid crowns.

Williams et al. (1965) noted the ab-
sence of a canthal ridge in this species.
This feature is evident on the maxilla by
its low, medially arching nasal process.
Maxillary teeth are similar to those on
the dentary.

The frontal bone of Anolis occultus
is diagnostic. The bone is nearly as wide
as long, has little medial constriction,
and the supratemporal processes are re-
duced (Fig. 14). The average ratio of
midsaggital length divided by the nar-
rowest width (the medial constriction)
is 1.8 (1.5-2.1) for fossil and recent

frontals that I examined. The value is
considerably greater in other Anolis. For
example, in A. cristatellus the ratio is
3.4 to 3.8, in A. gundlachi, 3.4 to 3.6, in
A. stratulus, 3.8 to 4.2, in A. evermanni,
2.8 to 3.2 and in A. krugi, 3.5 to 3.9.

The largest specimen of A. occultus
reported by Williams et al. (1965) had
a snout-vent length of 42. The largest
fossil dentary came from an individual
estimated at 47. The average was 35.

Anolis spp.

Material—Blackbone 1: dentaries
(13 right, 10 left—USNM 259147); ar-
ticular + surangular (1 right, 2 left—
USNM 259147); coronoids (3 right—
USNM 259150); maxillae (20 right, 22
left—USNM 259148-9); premaxillae (4
—USNM 259149); frontals (14—USNM
259152); parietals (3—USNM 259153);
basales (65—USNM 259151); quadrates (1
right, 2 left—USNM 259150); jugals (1
right, 3 left—USNM 259153); pterygoids
(1 right—USNM 259151); scapulae (2
right, 2 left—USNM 259154); pelves (7
right, 6 left—USNM 259154); vertebrae
(several hundred—USNM 259155-7).

Comments.—Much of the fossil ma-
terial of Anolis is unidentifiable to spe-
cies and comes from individuals with
snout-vent lengths between 45 and 60.
This size class includes A. pulchellus,
A. krugi, A. evermanni, A. stratulus and
A. poncensis. For the most part the
fossils are fragmentary, and because of
ontogenetic and individual variation as
well as structural similarity of a given
bone from one species to another, an
attempt at specific allocation would be
irresponsible. Anolis pulchellus and A.
stratulus could be expected in this sam-
ple because these two forms presently
are common at low elevations on the
island (Gorman and Harwood, 1977).
No doubt further material of A. ever-
manni, A. krugi and A. occultus are
among these remains as well. Anolis
evermanni is found in mesic situations
in the interior, higher altitude areas of
Puerto Rico. Anolis occultus is a canopy

PLEISTOCENE PUERTO RICAN HERPETOFAUNAS 29

dweller that sleeps on exposed perches,
also mostly in the higher reaches of the
island. Like A. cristatellus and A. gund-
lachi, Anolis pulchellus, and Anolis krugi

Fic. 14.—Fossil frontal bones of Anolis cuvieri

(top, USNM 259107), A. krugi (middle, USNM

259142), A. occultus (bottom, USNM 259145).

All are shown in dorsal view. Scale equals
2 mm.

are ecomorphic equivalents at low and
high elevations, respectively.

Cyclura pinguis Barbour

Syn: Cyclura mattea Miller; Cyclura
portoricensis Barbour.

Material—Blackbone 1: dentaries
(1 right in two pieces, 1 left in two pieces
—USNM 259158-9); surangular (1 left
—USNM 259162); maxillae (2 left—
USNM 259160-1); parietals (I—USNM
259163); quadrates (1 left—USNM
259164); scapulae (1 right, 1 partial left
—USNM 259163); clavicles (1 right, 1
left—USNM 259165); humeri (1 left—
USNM 259166); ulnae (1 complete left,
1 proximal left—USNM 259167); radii
(1 right, 1 proximal left—USNM 259168);
femora (1 right, 1 left—USNM 259169) ;
fibulae (1 left—USNM 259170); tibiae
(2 right, 1 left—USNM 259171); tarsals
(4—USNM 259172); metatarsals (5—
USNM 259172); phalanges (83—USNM
259172); vertebrae (34—-USNM 259173).

Blackbone 2: articular (1 left—
USNM 259177); premaxillae (I—USNM
259176); postfrontal + postorbital (1
right—USNM 259178); parietals (1—
USNM 259175); basales (I—USNM
259174); jugals (1 left—USNM 259178) ;
clavicles (2—USNM 259179); humeri
(1 distal right—USNM 259181); radii
(1 right—USNM 259181); metacarpals
(2—USNM 259181); ilia (1 right—
USNM 259180); femora (1 partial right
—USNM 259181); vertebrae (3 dorsal,
1 sacral, 5 caudal—USNM 259182).

San Miguel Cave: pterygoids (2
right—USNM 259184-5); palatines (1
right—USNM 259186); humeri (1 left—
USNM 259187); pubis (1 left—USNM
259188); femora (1 proximal left—USNM
259187); tarsals (1 astragalus—USNM
259188); vertebrae (1 dorsal, 3 caudal
—USNM 259188).

Cueva Clara: parietals (1—KUVP
11564); ulnae (1 proximal left—KUVP
11564); tibiae (2 proximal right—KUVP
11564); fibulae (2 right—KUVP 11564);
vertebrae (3 dorsal, 3 caudal—KUVP
11564).

30 MISCELLANEOUS PUBLICATION MUSEUM OF NATURAL HISTORY

Cueva del Perro: dentaries (1 partial
right—KUVP 11472); vertebrae (3 cau-
dal—KUVP 11472).

Nesophontes Cave: maxillae (1 par-
tial left—USNM 259183); vertebrae (1
dorsal, 2 caudal—USNM 259183).

Cueva de Silva: femora (1 right, 1
left—KUVP 11572).

Barahona IV: vertebrae (2 dorsal,
1 caudal—USNM 259189).

Description—Dentary: fossil den-
taries of Cyclura pinguis are not com-
mon. From Blackbone | there are two
fragments each of a left and right den-
tary which appear to be from jaws
broken approximately in half. The pieces
consist of the tooth-bearing portions of
the bone; the ventral faces are missing.
The anterior portion of the right dentary
is 30.5 from the rounded symphysis to
the broken posterior edge. The posterior
half is 34.4 from the broken anterior end
to the posterior tip of the articular sur-
face of the coronoid. Thus, the two
pieces represent a jaw at least 65 in total
length. I doubt that much of the middle
portion of the dentary is missing, al-
though the two halves do not join neatly
together (Fig. 15). Sixteen teeth or
empty alveoli are present on the an-
terior piece and 14 on the posterior.
Avery and Tanner (1971) gave a tooth
count of 22 to 28 for Cyclura. A recent
skeleton of C. pinguis (ASFS V21995)
with a dentary tooth row of 34.4 has 26
teeth. Thus, accounting for the 2 or 3
teeth missing from the middle piece of

_ the fossil dentary, its tooth count would
be about 32—a figure not unreasonable
considering the size of the jaw.

The two pieces of a left dentary are
proportionately the same size as the
aforementioned right hand pieces, but
they are smaller and can not be identi-
fied as confidently as a single jaw in
two pieces. The anterior portion is 23.1
in length. The posterior fragment, bear-
ing an articular facet for the coronoid,
is 35.6 in total length. Overall, the jaws
are smooth and penetrated by seven
foramina anteriorly.

A small right dentary fragment from
Cueva del Perro is missing the posterior
fourth of the bone. The fossil is 26.5
measured ventrally from the symphysis
to the broken posterior edge. The dental
shelf extends back only half the length
of the bone thus exposing Meckel’s canal
posteriorly. Six teeth are present out of
13 alveoli on the shelf.

The teeth of C. pinguis are high and
narrow. The first five or six are simple
and pointed. The crowns of the pos-
terior teeth have two small anterior
cusps, a large central cusp, and a single
small posterior cusp. The crowns also
are compressed and oriented obliquely
so that adjacent teeth overlap one an-
other. The crowns are moderately ex-
panded beyond the width of the bases,
but not to the degree obtained by some
species, for example, C. carinata and C.
cychlura. In C. cornuta and C. nubila
the crowns are multicuspate and ap-
proach the serrated condition of Iguana
and Ctenosaura. The tooth crowns of
C. pinguis are in a horizontal plane
throughout the length of the dentary,
but the bases sit in a concave shelf so
that the tallest teeth are in the middle
of the dental series. The sides of the
teeth are parallel and the shafts are
compressed distomesially.

Maxilla: a large, right maxilla from
Blackbone 1 could be from the same indi-
vidual represented by the right dentary
discussed above. The bone is broken near
the midpoint. The anterior piece is 17.1
in length. It bears the premaxillary proc-
ess, the anterior half of the nasal process,
and a dental row with implacements for
eight teeth of which numbers 2, 3, 5 and
7 are present (Fig. 15). The premaxil-
lary process is broad dorsally, 9.0 wide,
and has articular surfaces for the pre-
maxilla anteriorly and the vomer antero-
medially. The nasal process rises sharply.
At its basal curvature is a deep pit lead-
ing to the anterior inferior alveolar
foramen. The nasal process rises 23.0
above the parapet of the jaw.

The posterior piece is 39.7 from its

PLEISTOCENE PUERTO RICAN HERPETOFAUNAS 31

Fic. 15.—Fossil jaws of Cyclura pinguis. Lin-
gual view of broken dentary (top, USNM
259158) and labial view of partial maxilla
(bottom, USNM 259160). Scale equals 10 mm.

broken anterior edge back to the tip of
the pterygoid process. Part of the nasal
process is present, but it is missing the
dorsal margin. The labial surface is
vertical and penetrated by three large
and closely spaced foramina located in
a row above the parapet of the jaw.
The dental shelf is concave with spaces
for 16 teeth, 9 of which are present. The
lingual edge of the dental shelf produces
a medially directed process for articula-
tion with the palatine. The dental shelf
narrows fore and aft to either side of
the process.

Maxillary teeth are similar to those
on the dentary. If, in fact, the two
halves are from the same maxilla, then
the bone would have had a total length
of approximately 59 and a tooth count
of 27 or 28.

A small left maxilla, also from Black-
bone 1, is missing the anterior edge of
the premaxillary process and the pos-
terior expansion of the nasal process.
Evidently it is from a juvenile. The bone
is 23.5 in length and has 17 teeth or
empty alveoli. The teeth are proportion-
ately shorter than those of adults. A

fragment of a left maxilla from Neso-
phontes Cave consists of the premaxillary
process and a dental shelf bearing four
teeth and three empty alveoli. It is simi-
lar to the large specimen from Black-
bone 1.

Premaxilla: a single premaxillary
bone from Blackbone 2 is complete, but
for the end of the right maxillary proc-
ess. The bone is large and robust, 31.3
from the anteroventral edge to the tip
of the nasal process. The width is 16.6
accounting for the missing right maxil-
lary process. The bone is rounded an-
teriorly and perforated by numerous
small foramina. Four teeth alternate
with four empty alveoli. The teeth are
simple and pointed. The nasal process
is long and thin relative to some species
such as C. cychlura and C. nubila in
which it is short and wide. Ventrally,
the caudal half of the nasal process is
compressed laterally into a keel which
articulates with the nasal bones.

Articular and surangular: the left ar-
ticular from Blackbone 2 and the left
surangular from Blackbone 1 are the only
bones present from the post-coronoid sec-
tion of the lower jaw. Both bones are
impressively large and came from man-
dibles probably exceeding 120. The ar-
ticular is broken anteriorly, but otherwise
complete. It is 46.5 long and 24.3 across
the tips of the angular and retroarticular
processes. The angular process is thick
and extends medially 12.7 from the rim
of the articular condyle. The large suran-
gular is missing part of the ventral edge.
It is 44.8 in length and 12.2 in height.
On the anterolabial side is a faint V-
shaped depression marking the site of
dentary overlap.

Palatine: a right palatine was re-
covered from the rear wall of San Miguel
Cave. The bone is broad, elliptical in
shape, and has a well developed vomer-
ine process directed craniad from the
anteromedial side (Fig. 16). The length
of the bone is 36.5 as measured from the
tip of the vomerine process to the curved,
posterior border. The greatest width is

32 MISCELLANEOUS PUBLICATION MUSEUM OF NATURAL HISTORY

19.5 from the lateral edge of the maxil-
lary process straight across to the medial
side of the bone. There is a crescent-
shaped excavation on the anterolateral
surface. Posteriorly, an oblique scar
marks the articular surface of the ptery-
goid.

Pterygoid: two right pterygoids re-
ferred to this lizard are both from San
Miguel Cave. Each is missing a small
anterior portion of the palatine process.
They are nearly equal in size, but the
larger of the two is 45.0 in length (Fig.
16). The pterygoids of C. pinguis are
distinguishable from other members of
the genus by two features. First, the
medial border anterior to the transverse
maxillary process is nearly straight with
little or no emargination. Secondly, the
pterygoid teeth occur in double rows
that begin in the middle of the bone at
the level of the maxillary process. The
rows curve inward along the medial bor-
der of the bone and continue craniad as
a single row that ends near the apex of
the palatine process. More than 20 sock-
eted teeth may be present. The ptery-
goid tooth pattern of C. pinguis closely
resembles that of Iguana or Ctenosaura,
but in other species of Cyclura the
pterygoid teeth occur in a small patch
or row and are confined to the middle
of the bone. They are usually 10 or less
in number. Caution is advised in using
this character in iguanine lizards be-
cause pterygoid teeth may be variously
lost or increased in number with on-
togeny. However, the pattern in C.
pinguis is so distinctive relative to other
members of the genus as to be diagnostic.

Parietal: three fossil parietals of dif-
ferent sizes exhibit the ontogenetic vari-
ation typical of this bone in many igua-
nid lizards (see Etheridge, 1959; Oelrich,
1956). As measured across the frontal
border, the largest (from Blackbone 2)
is 39.5, the smallest (from Blackbone 2)
is 27.6, and the intermediate size (from
Cueva Clara) is 32.4. As parietals in-
crease in size, the midsagittal ridge
heightens and the sides lateral to the

ridge become constricted. The angle
formed by the supratemporal processes
increases as the processes themselves
lengthen. In all three fossils the anterior
face of the dorsal surface is moderately
rugose.

Basale: a large braincase from Black-
bone 2 consists of fused exoccipitals,
basioccipital and basisphenoid (Fig. 16).
The distance is 36.6 between the ends
of the exoccipital processes. The generic
differences in braincases among iguanine
lizards was first pointed out by Boulenger
(1890) and later used by Etheridge
(1964a) to identify a fossil basale from
Barbuda, British West Indies. In Iguana
and Cyclura the basisphenoid is much
wider than long and slightly constricted
behind the pterygoid processes. In these
lizards, the basisphenoid length as meas-
ured from the bone’s posterior border to
the apex of the excavation between the
parasphenoid process and the pterygoid
process divided by the narrowest width
behind the pterygoid process, gives a
ratio of 0.40 to 0.72. The ratio of the
Blackbone 2 specimen is 0.50 (Fig. 16).
In C. pinguis the pterygoid processes are
broad and short such that the depres-
sions immediately posterior to the proc-
esses are shallower than in other species
of the genus.

Quadrate: the left quadrate from
Blackbone | is 23.5 from the apex of the
cephalic condyle to the bottom of the
mandibular condyle. The bone is rec-
tangular and 14.4 between the lateral
and medial borders. Its most unusual
characteristic is a prominent shelf which
overhangs the quadrate foramen. I have
not seen this feature in other species of
Cyclura.

Limb bones and girdle elements:
many appendicular components of vary-
ing sizes are present as fossils. There is
nothing remarkable about these bones
that would distinguish them from the
same elements of other species of Cy-
clura. They are more useful for estimat-
ing animal size and minimum number
of individuals per locality. Most of these

PLEISTOCENE PUERTO RICAN HERPETOFAUNAS 33

fossils are missing proximal or distal
ends, epiphyses, pieces of the shafts or
combinations of these. Measurements of
the longest and shortest complete long
bones are, as follow: humeri 75.9, 33.9;
ulnae 61.8; radii 57.2, 54.3: femora 80.0,
53.2; tibiae 69.8, 25.8; fibulae 72.0, 63.9.
A single ilium and pubis have lengths of
49.3 and 15.9, respectively.
Comments.—The occurrence of rock
iguanas on Puerto Rico was first revealed
by Barbour (1919:97) from a descrip-
tion of fossils found “. . . in a large cave
near Ciales.” The fossils were collected
two years previously by G. M. Allen and
J. L. Peters. Barbour chose the broken
extremities of a left humerus as the holo-
type (MCZ 1008) and named the species
Cyclura portoricensis. He used limb
bones in order to compare them with the
holotype of C. mattea, a left humerus
(USNM 59358-9) described by Miller
(1918) from Indian middens at Magens
Bay, St. Thomas, Virgin Islands. Cyclura
portoricensis was distinguished from C.

mattea by being larger and more mas-
sive. Barbour thought that C. portori-
censis and C. mattea were more closely
related to one another than either was
to the living C. pinguis on Anegada
Island at the eastern extreme of the
Puerto Rican Bank. Barbour also be-
lieved that C. portoricensis resembled
C. cornuta in having a shallower radial
fossa than that of C. mattea. Limb ele-
ments of iguanine lizards are poor indi-
cators of species-level taxonomic differ-
ences because of their relative simplicity,
ontogenetic change, individual variation,
and sexual dimorphism by size. I ex-
amined the holotypes and paratypes (all
limb bones) of both C. portoricensis and
C. mattea and, except for size differences,
found nothing significant that would dif-
ferentiate them from C. pinguis or from
any other species of Cyclura.

The presence of three species of Cy-
clura on the Puerto Rican bank is, and
has been, considered unlikely (Schwartz

Fic. 16—Fossil skull bones of Cyclura pinguis. Left, basale (top, occipital view, bottom, ventral
view, USNM 259174); right top, right palatine in ventral view (USNM 259186); right bottom,
right pterygoid in ventral view (USNM 259184). Scale equals 10 mm.

34 MISCELLANEOUS PUBLICATION MUSEUM OF NATURAL HISTORY

and Thomas, 1975; Schwartz and Carey,
1977), but until now adequate fossil ma-
terial from the islands that could verify
or refute the contention was unknown.
The fossils discussed here clearly are re-
ferrable to the species of Anegada Island.
Of the three available names, portori-
censis, mattea, and pinguis, the latter
has priority. Bones with osteological
characters that are considered diagnostic
for the species include the pterygoid,
basisphenoid, dentary, and maxillary
teeth, and the palatine, quadrate, and
nasal processes of the premaxilla. In
their review of iguanine lizards, Avery
and Tanner (1971) suggested that C.
carinata, C. cornuta, and C. nubila (= C.
macleayi) are typical of the genus. My
examination concurs in that most species
of Cyclura are difficult to separate osteo-
logically except for C. pinguis. Exter-
nally, Cyclura pinguis is also one of the
most distinctive members according to
Schwartz and Carey (1977), who re-
viewed the genus and relegated several
Bahaman forms to subspecies, reducing
the previously recognized number of
species from 14 to 8.

Cyclura pinguis is the most southerly
occurring species of this West Indian
genus. (A fossil basale from Barbuda,
British West Indies may belong to Cy-
clura; see Etheridge, 1964a.) Cyclura
pinguis is restricted to xeric habitats in
and around rock outcrops (Carey, 1975)
like its northern congeners, C. nubila
(Cuba, Cayman Islands), C. cychlura
(Bahamas), C. carinata (Turks and Cai-
cos Islands, Bahamas), C. collei (Ja-
maica), C. cornuta (Hispaniola, Mona
Island), C. ricordi (Hispaniola), and
C. rileyi (Bahamas). Cyclura pinguis is
also one of the largest members of the
genus—males to 539, females to 476
snout-vent length. Fossil individuals
ranged from juvenile to adult in size.
The smallest, based on a femur was 130-
140, whereas the largest exceeded 560
based on skull bones, vertebrae, and
limbs.

Leiocephalus

The iguanid lizard genus Leiocepha-
lus is endemic to the West Indies and
almost restricted to the Greater Antilles.
Cuba, Hispaniola, and the Bahamas har-
bor 20 of 21 living species. Leiocephalus
herminieri is known only from four speci-
mens with questionable locality data on
Martinique. It may be extinct. Two of
three species known only as fossils ex-
tend the former range of the genus to
Jamaica (L. jamaicensis, Etheridge,
1966a), and Barbuda in the Lesser An-
tilles (L. cuneus, Etheridge, 1964a). The
third extinct form, L. apertosulcus, was
described from Hispaniola (Etheridge,
1965).

Most iguanid lizard genera exhibit
multicuspid tooth crown morphology,
but flared, fan-shaped tricuspid teeth are
diagnostic of Leiocephalus. The large
iguanines, Iguana, Cyclura, Ctenosaura,
Conolophis, and Amblyrhynchus differ
in having deeper jaws and detailed tooth
crowns. Anolines lack fan-shaped tooth
crowns and an anterior opening for
Meckel’s canal. Leiocephalus and Leio-
lemus are the only lizards possessing fan-
shaped tricuspid teeth, a closed Mec-
kelian groove open anteriorly, and an
anterior extension of the coronoid on the
labial surface of the dentary (Etheridge,
1966b; Estes, 1963). In most species of
Leiolemus, the lateral cusps are not as
well delimited from the central cusp as
they are in Leiocephalus. Also, in some
species of Leiolemus such as L. elongata,
L. multiformis, and L. nitidus, the
splenial extends farther anterad than in
any Leiocephalus, spanning seven or
eight teeth, whereas the maximum in
Leiocephalus is three or four.

Finding fossil Leiocephalus on Puerto
Rico was not wholly unexpected because
of its past and present distribution, al-
though none is living there now. Two
extinct species are present as fossils.
One is represented by skull and post-
cranial elements and the other species
by two dentaries. Both possess the char-
acters of the genus. Richard Etheridge

PLEISTOCENE PUERTO RICAN HERPETOFAUNAS 35

first recognized the differences between
insular and mainland lizards previously
referred to Leiocephalus. Moreover, by
his contribution to the herpetology of
the West Indies, and our friendship
which partly inspired this study, it is
gratifying to name the first of these
two lizards:

Leiocephalus etheridgei, new species

Holotype—A right dentary No.
259190 in the vertebrate paleontology
collection of the National Museum of
Natural History, Smithsonian Institution,
Washington, D.C.

Type locality and horizon.—Black-
bone 1 Cave, 1.2 km S Barrio de Bara-
hona, Municipio de Morovis, Puerto Rico
(18° 20’ 57” N, 66° 26’ 47” W). Late
Pleistocene.

Material—Blackbone 1: dentaries
(7 right, 5 left, 5 fragments—USNM
259190-1); maxillae (1 right, 1 fragment
—USNM 259192); frontals (3—USNM
259193-4); basales (4—USNM 259195-
7); basisphenoids (1 fragment—USNM
259198); pterygoids (1 left—USNM
259202); dorsal vertebrae (10—USNM
259199); sacral vertebrae (8—USNM
259200); caudal vertebrae (2—USNM
259201).

Cuevo del Perro: dentaries (1 right,
6 left—KUVP 11473).

Diagnosis.—Leiocephalus_ etheridgei
is distinguished from all known species
of the genus by the presence of an acute
convex ridge on the anterior, labial face
of the dentary below the mental fora-
mina, and by the anterior opening of
Meckel’s groove, which extends from the
level of the sixth tooth forward to the
symphysis of the jaw. In other species
of Leiocephalus the labial face of the
dentary is smooth and convex, and but
for L. apertosulcus, Meckel’s groove is
open anteriorly as a small pore beneath
the first or second tooth. In L. aperto-
sulcus Meckel’s groove is open and un-
fused throughout the length of the den-
tary (Etheridge, 1965).

Description —The holotype (Fig. 17)

is a nearly complete right dentary meas-
uring 13.6 mm from the symphysis pos-
terior to the broken tip of the angular
process. It came from an individual with
an estimated snout-vent length of 110-
115 mm. The tooth row has a straight-
line measurement of 11.8 mm and con-
tains 20 pleurodont teeth or empty
alveoli. Teeth missing from front to back
are numbers 1, 2, 6, 8, 11, 16, and 19.
The height of the dentary just posterior
to the last tooth is 3.0 mm. The anterior
upper half of the labial face is smooth
and concave. Posteriorly the labial face
is slightly convex. The anterior upper
half of the labial face is demarcated
from the lower half by a narrow, convex
ridge arching gently from the symphysis
posteriorly to the level of the sixteenth
tooth. Nine foramina of unequal size
are spaced irregularly on the labial face
between the second and thirteenth teeth.
Posteriorly there is a faint wedge-shaped
coronoid scar that extends forward to the
level of the last tooth. The ventral sur-
face of the dentary is flat anteriorly, but
slightly convex posteriorly. Meckel’s
groove is open at the front of the jaw
as a lanceolate cleft extending from the
level of the sixth tooth forward to the
symphysis. Lingually, the bone is smooth
and convex below the tooth row except
for a shallow, longitudinal depression
preceding the posterior opening of Mec-
kel’s groove, which remains open pos-
teriorly to the level of the sixteenth tooth.

The crista dentalis has a ventral cur-
vature deepest at the midpoint of the
jaw. Accordingly, the teeth are largest
in the center of the dental row and be-
come smaller fore and aft. The bases of
the teeth are cylindrical except for the
posterior three which are mesodistally
compressed. The tricuspid crowns are
broader than the bases and have sharp
linguilabially compressed cutting edges.
The central cusps are pointed, higher,
and larger than the lateral cusps, and
separated from them by grooves extend-
ing to the base of the crown. Weakly
developed lateral cusps are present on

36 MISCELLANEOUS PUBLICATION MUSEUM OF NATURAL HISTORY

the third tooth and are fully developed
on the fourth tooth and all others but
the last. The last tooth is nearly conical
because the lateral cusps are rudi-
mentary. About 40% of each tooth is
exposed above the jaw parapet.

The other dentaries from Blackbone
1 and Cueva del Perro are similar to the
holotype. Tooth row lengths of 16 fossils
vary from 9.3 to 12.5 mm (x = 11.0).
Teeth missing from the holotype are
variously present in the referred speci-
mens, except for the second tooth, and
are similar to others in the tooth row.
Small, pointed replacement teeth are
present in several dentaries.

Maxilla: one maxilla and one maxil-
lary fragment from Blackbone | are re-
ferred to this species. The more com-
plete specimen (Fig. 18) is missing the
dorsal end of the nasal process. Its total
length is 12.4 mm and it measures 11.4
mm along the tooth row. The bone con-
tains 18 teeth or empty alveoli. Missing
from front to back are numbers 7, 12,
15 and 17. Parts of the bases of teeth
3, 5, and 9 are missing also. The first
tooth is simple and pointed and the sec-
ond tooth is weakly tricuspid. All teeth
posterior to number 2 are flared and
tricuspid like those of the dentary. The
crowns are oriented at an oblique angle
such that the anterior cusp of each tooth
overlaps the lingual surface of the pos-
terior cusp on the preceding tooth. The
labial face of the maxilla is smooth and
penetrated by an irregular series of 15
foramina.

The maxillary fragment is an anterior
piece bearing the premaxillary process,
part of the nasal process, and a section
of the dental row with seven teeth or
empty alveoli. The third, fifth, and
seventh teeth are replacement series,
smaller and narrower than those that are
fully developed.

Frontal: two frontal bones, one large

and one small, are typical of Leiocepha-
lus in being broad at the parietal border
and constricted medially. The larger
specimen is missing the tip of the left
supratemporal process (Fig. 18). It is
12.7 mm as measured midsagitally, 11.7
mm across the parietal border, 4.8 mm
across the narrow midsection, and 6.1
mm at the anterior end. The pineal
foramen is present on the parietal
border as a V-shaped notch. The bone
has a slight dorsal curvature and an
irregularly sculptured surface. There
are two dorsal, semicircular depressions
for articulation with the nasal bones
on the anterior border. Laterally the
frontal is thick and rounded and bears
prefrontal scars extending back half
the length of the bone. The smooth
ventral surface has two low descending
laminar processes and a notch anteriorly
for articulation with the nasals.

The smaller frontal bone is missing
the tips of the supratemporal processes
and the ends of the nasal processes. The
bone is 9.0 mm long, 6.7 mm across the
parietal border, 2.9 mm across the mid-
dle and 4.2 mm wide at the anterior end.
It is otherwise similar to the larger speci-
men.

Basale: Leiocephalus etheridgei also
is represented by four braincases each
composed of fused exoccipitals, a bas-
ioccipital and basisphenoid. The two
most nearly complete specimens are 12.3
mm and 11.8 mm across the exoccipitals
and 6.5 mm and 6.3 mm from the an-
terior border of the parasphenoid to
the posterior lip of the occipital condyle.
The other two basales are missing por-
tions of the basisphenoid and_basioc-
cipital. They are 12.2 mm and 10.6 mm
across their respective exoccipital proc-
esses. The pterygoid processes on the
complete basales are long, and flare out
laterally so that their distal ends are in
a line with the sphenoccipital tubercles.

Fic. 17.—Holotype of Leiocephalus etheridgei, new species (top, dentary in lingual and labial views,
USNM 259290), and holotype of L. partitus, new species (bottom, dentary in lingual and labial
views, USNM 259203). Scale equals 5 mm.

PLEISTOCENE PUERTO RICAN HERPETOFAUNAS

38 MISCELLANEOUS PUBLICATION MUSEUM OF NATURAL HISTORY

The sutures are fused between the re-
spective bones of each braincase. One
fossil is shown in Figure 18.
Vertebrae: vertebrae are referred to
this species on the basis of their size
and similarity to those of other species
of Leiocephalus. They are from various
regions of the vertebral column. The
largest midbody vertebra has a centrum
2.7 mm in length, including the condyle,
and bears a midventral ridge. The neural

spine is robust and angles posteriad to
a height of 19 mm. Zygosphenes are
moderately developed within the an-
terior rim of the neural arch. Hypopo-
physes are present on the cervical verte-
brae. Eight sacral vertebrae of nearly
equal size have an average width across
the diapophyses of 9.5 mm. Two pre-
autonomic caudal vertebrae are 6.1 mm
and 6.4 mm across their diapophyses.
They have well-developed zygosphenes.

Fic. 18.—Fossils of Leiocephalus etheridgei, new species. Maxilla (top left—lingual and labial
view, USNM 259192); basale (bottom left—occipital view, USNM 259196); frontal bone (top
right—dorsal view, bottom right—ventral view, USNM 259193). Scale equals 3 mm.

PLEISTOCENE PUERTO RICAN HERPETOFAUNAS 39

Leiocephalus partitus, new species

Holotype——A right dentary No.
259203 in the vertebrate paleontology
collection of the National Museum of
Natural History, Smithsonian Institution,
Washington, D.C.

Type locality and horizon.—Guanica
Bat Cave, Reserva Forestal Guanica, 6
km E Barrio de Guanica, Municipio de
Guayanilla, Puerto Rico (17° 57’ 40” N,
66° 50’ 53” W). Late Pleistocene-Sub-
Recent.

Material—Guanica Bat Cave: den-
taries (1 right—USNM 259203).

Cueva del Perro: dentaries (1 right
—KUVP 11473).

Diagnosis.—A large species of Leio-
cephalus distinguished from all other
members of the genus by the presence
of a vertical intramandibular septum
traversing the length of Meckel’s groove
and fused to the ventral, internal surface
of the dentary. An intramandibular sep-
tum of this proportion is not present in
any other species of Leiocephalus in
which Meckel’s groove is closed. Leio-
cephalus partitus is further characterized
by blunt, weakly flared tricuspid teeth,
the transition to tricuspid teeth taking
place at a more posterior position in the
tooth row and by Meckel’s groove open-
ing anteriorly below the seventh tooth.

Etymology.—partitus, L. divided; in
reference to the intramandibular septum.

Description.—The holotype (Fig. 17)
is a nearly complete right dentary that
came from an individual with an esti-
mated snout-vent length of 125-130 mm.
The bone is missing the posterior border
behind the coronoid scar. The dentary
is 16.2 mm in total length and has a
tooth row of 13.5 mm measured in a
straight line. There are 24 teeth or empty
aveoli of which the following teeth are
missing from front to back: numbers 2,
4, 14, 15, 16, the base of 17, and the
crowns of 21 and 24. The height of the
dentary is 3.5 mm just posterior to the
last tooth. The first tooth is oriented
anterodorsally as a simple shaft with a
pointed crown. The third tooth is larger

and missing the crown, but otherwise is
similar to the first. The still larger fifth,
sixth, and seventh teeth are also simple
and pointed. The transition to tricuspid
teeth begins with the eight tooth which
has rudimentary lateral cusps adjacent
to a blunt median cusp. The ninth and
all posterior teeth have tricuspid crowns.
The tooth bases are mesodistally com-
pressed. The crowns have a slight lateral
flare only and are not appreciably wider
than their bases. Lateral cusps are sepa-
rated from the median cusps by shallow
furrows extending to the base of the
crown. The median cusps are blunt
points; thus, the shearing surface of the
tooth row is not so acute as it is in other
species of Leiocephalus. In occlusive
view, the tooth row presents a reversed
sigmoidal outline; the row curves later-
ally from the symphysis, back medially
towards the middle of the row and then
laterally again posteriorly.

The upper half of the labial face is
smooth and convex except for an irregu-
lar surface anteriorly. A horizontal row
of five mental foramina penetrates the
labial surface between the second and
fourteenth teeth. There is a large >-
shaped scar for the coronoid overlap on
the upper posterior surface. The scar
extends forward to the level of the twen-
tieth tooth. The lower half of the labial
face is convex anteriorly and flat pos-
teriorly below the coronoid scar. The
ventral surface is rounded. Meckel’s
groove opens as a narrow pore below
the level of the seventh tooth on the
ventral surface and continues forward to
the symphysis as a deep, curving sulcus.
The dental shelf is produced lingually
as a horizontal ledge upturned at the
front of the bone. The shelf curves dor-
sally at the level of the twentieth tooth
with the widening of Meckel’s canal.
At this widening an intramandibular sep-
tum is exposed as a vertical partition
within the canal. The ventral border of
the septum is fused to the dentary and
extends back to the last tooth.

The only other bone referred to this

40 MISCELLANEOUS PUBLICATION MUSEUM OF NATURAL HISTORY

species is another right dentary re-
covered from Cueva del Perro, estimated
to have come from a lizard with a snout-
vent length of 120 mm. It is very simi-
lar to the holotype, but with the fol-
lowing measurements: total length 14.7
mm; tooth row length 13.3 mm; and
height 2.9 mm posterior to the last
tooth. The dentary has 24 teeth or
empty alveoli of which numbers 5, 7,
15, 22 and 24 are missing. The first
four teeth as before are simple and
pointed like the first tooth of the holo-
type. The transition to tricuspid teeth
begins at the tenth tooth and is com-
pleted at the eleventh. A well-developed
intramandibular septum also is present
in this specimen.

Comments.—There are no extant spe-
cies of Leiocephalus with dentary char-
acters approaching those of L. etheridgei
or L. partitus. The Cuban lizard L.
macropus has a labial ridge like that of
L. etheridgei, but it is poorly developed
and the dentary has fewer mental foram-
ina (5 versus 9). In fact, the dentaries
of living species of Leiocephalus are uni-
form and morphologically comparable.
Variations and similarities of features
such as tooth-crown shape probably re-
flect dietary specializations, whereas
other characters may have phylogenetic
significance such as the anterior condi-
tion of Meckel’s groove. The opening is
wide and uniform to the end of the jaw
in L. etheridgei, whereas in L. partitus
the opening is small and continues cra-
niad as a narrow sulcus. But in both
Puerto Rican species the opening is at
about the same place on the jaw—much
farther posterior than in other Leioceph-
alus. This probably reflects incomplete
fusion of Meckel’s groove, and to this
extent L. etheridgei and L. partitus ap-
proach the condition of L. apertosulcus
in which the groove is open altogether.
Complete fusion of Meckel’s groove oc-
curs in a number of iguanids, but spo-
radically among lizards in general. Thus,
the fusion of Meckel’s groove is derived
for Leiocephalus and in other iguanids

where it occurs, and the open condition
of L. apertosulcus and incomplete fusion
in L. etheridgei and L. partitus is primi-
tive for Leiocephalus and not necessarily
indicative of relationships amongst these
three species. Most species of Liolaemus
show fusion of Meckel’s groove with an
anterior opening similar to Leiocephalus,
but there is also incomplete closure and
fusion in several species, for example,
Liolaemus lineomaculatus, L. fitzingeri,
L. kingi, and L. wiegmanni.

An intramandibular septum is absent
from Leiocephalus etheridgei and L.
jamaicensis, but is prominent in the other
fossil species L. apertosulcus and L.
cuneus. In these two, the structure is
similar to that of L. partitus except that
the posterior edge is distinctly emargi-
nate. The septum is absent from other
representative tropidurines (sensu Eth-
eridge, 1964b:629) such as Tropidurus
torquatus, Plica umbra, Stenocercus var-
ius, Uranoscodon superciliosa, and Ura-
centron flaviceps. A short septum is pres-
ent in Ophryoessoides iridescens, several
species of Liolaemus (multiformis,
chilensis, nigriceps, lineomaculatus) and
some modern species of Leiocephalus
(carinatus, melanochloris, schrebersei,
punctatus). In these species the septum
is “short” in that it does not extend pos-
teriorly for more than three-fourths of
the length of the dentary, and the nutri-
ent canal it forms by fusing to the inner
labial wall has a very small diameter.
The septum may continue aft as a short
keel descending from the roof of Mec-
kel’s canal. There is an interesting se-
quence in the reduction of the septum
that corresponds with closure of Meckel’s
groove in Liolaemus. Closure of Meckel’s
groove increases along the length of the
dentary from L. multiformis to L. chilen-
sis to L. nitidus, and the septum becomes
fused more to the inner labial wall rather
than the ventral floor, in other words a
more horizontal as opposed to vertical
orientation within the dentary. Perhaps
then, the absence of an intramandibular
septum in other tropidurines is a loss

PLEISTOCENE PUERTO RICAN HERPETOFAUNAS Al

resulting from the closure and fusion of
Meckel’s groove, Leiocephalus cuneus
and L. partitus being exceptions.

There is more to the above argument,
however. The intramandibular septum
is not common in iguanids, whether
Meckel’s groove is open or closed, and
where present, it is reduced, extending
no more than half to three-fourths of the
length of the dentary. Among iguanines,
lizards in which Meckel’s groove is
fused, the septum is best developed
within Ctenosaura and least in Conolo-
phis. Iguana, Cyclura, and Dipsosaurus
are intermediate. A septum like that of
the iguanines is present in Sceloporus
and Urosaurus, the sceloporines that I
examined, and Meckel’s groove is either
open or mostly unfused in this group
(Etheridge, 1964b). The septum is ab-
sent from most anoles, but it is present
as a very thin structure in the giant spe-
cies A. cuvieri, A. ricordi, and A. eques-
tris; Meckel’s groove is completely fused
in all Anolis.

In anguimorph lizards, Estes (1964:
124) correlated the intramandibular sep-
tum with dentary/post-dentary articula-
tion, but the morphology of the septum
in these lizards suggests that this septum
might not be homologous with that of
iguanids. However, the septum, or some
remnant of it, does play a part in den-
tary/post-dentary articulation in iguanid
lizards. In large iguanines (Iguana, Cy-
clura, Amblyrhynchus, Ctenosaura, Con-
olophus), Laemanctus, Basiliscus (basili-
scines), Morunosaurus, Enyaliosaurus,
and Enyalioides the articulation is a ver-
tical ziz-zag joint between the dentary
and surangular-angular on the labial side
of the jaw directly below the coronoid.
The intramandibular joint is a tongue
and groove arrangement in which the
surangular laterally and the anterolingual
arm of the coronoid medially form a slot,
into which fits the short, posterior keel of
the intramandibular septum descending
from the inner roof of the dentary. In
these lizards, except the basiliscines, the
joint is reinforced by a pronounced an-

teriorly directed overlap of the coronoid
onto the labial surface of the dentary,
forming a bracket on top of the jaw.
In Laemanctus and Basiliscus, where the
tongue and groove arrangement is more
robust, the labial arm of the coronoid
descends vertically onto the mandible,
mostly overlapping the surangular.

In tropidurines and Anolis the intra-
mandibular joint is a modification of the
above condition. The most conspicuous
difference is that in tropidurines and
Anolis the labial surface of the dentary
extends posteriorly beyond the middle
of the coronoid, about 10-12% of the den-
tary length on the average. In these
lizards the primary joint is provided, by
the wide overlap of the posterior end
of the dentary onto the surangular bone,
with an articulation which may fuse in
adults. On the lingual side of the jaw,
the anterolingual arm of the coronoid
fits up underneath the overhanging den-
tal shelf flush against the surangular.
Presumably, this articulation is a con-
dition derived in conjunction with the
reduction of the splenial in these lizards.
When the intramandibular septum is
present, as in those species mentioned
previously, the joint is still formed the
same way; the keel fits flush on top of
the anterolingual arm of the coronoid.
The coronoid overlaps the dentary on
the labial surface in Anolis, but not in
tropidurines except Leiocephalus, Lio-
laemus, and, to a slight extent, in Cteno-
blepharis.

In Sceloporus, the intramandibular
joint is structurally intermediate to the
“jguanine” type and the “tropidurine”
type. The surangular and the antero-
lingual arm of the coronoid form a slot
to receive the descending keel of the
intramandibular septum. There is no
coronoid overlap on the labial side of
the jaw, but the dentary does extend
posteriorly beyond the level of the coro-
noid to form a fairly wide overlap with
the surangular, though not as much as
in tropidurines and anoles.

We might conclude that whereas the

42 MISCELLANEOUS PUBLICATION MUSEUM OF NATURAL HISTORY

significance of the intramandibular sep-
tum in iguanids is unclear, the structure
is present primitively as a weakly de-
veloped feature in most members of the
family; its loss altogether is loosely cor-
related with closure and fusion of Mec-
kel’s groove. Thus, the hypertrophied
condition of the intramandibular septum
in Leiocephalus partitus, L. apertosulcus
and L. cuneus is a uniquely derived state
suggesting a relationship between these
three species.

The number of dentary teeth in-
creases with tooth row length in Leio-
cephalus, as it does in many lizards, but
for individuals of equivalent size the
number is fairly constant among species.
The fewest teeth occur in L. macropus
(18) and the greatest number in L.
melanochloris (25). Twenty-two is aver-
age. Likewise, the transition from sim-
ple to tricuspid teeth takes place rather
abruptly between the seventh and ninth
teeth except for the species L. cuneus,
L. etheridgei, and L. partitus. In the
former two, the transition is completed
at the fourth or fifth tooth and in L.
partitus at the tenth or eleventh tooth.
The more anterior transition to tricuspid
teeth was used by Etheridge (1964a) to
distinguish L. cuneus from other species
of Leiocephalus. Although L. cuneus
now shares this character with L. ethe-
ridgei, the former species has an intra-
mandibular septum, a deeper coronoid
scar which extends farther forward on
the labial surface of the jaw, and a
smaller anterior opening for Meckel’s
groove.

The anterior coronoid overlap does
not reach the last tooth in the species
barahonensis, macropus, pratensis, and
semilineatus. The overlap reaches the
last tooth in etheridgei, inaguae, green-
wayi, loxogrammus, raviceps, stictogas-
ter, and vinculum and extends to the
penultimate tooth or beyond in carnatus,

cuneus, cubensis, lunatus, melanochloris,

psammodromus, punctatus, and partitus.
There is little correlation with this char-
acter and the anterior extent of the

splenial on the lingual side of the jaw.
The splenial does not reach the level of
the last tooth in L. macropus, but spans
the posterior three or four teeth in cari-
natus, melanochloris, personatus, punc-
tatus, psammodromus, and raviceps. The
splenial overlaps one or two teeth in all
other species. The anterior extent of the
splenial does not appear to be size-re-
lated because in the small Cuban species,
L. raviceps, the bone spans three pos-
terior teeth as it does in the larger L.
carinatus. The splenial is absent from
the fossils, but the scar remains on the
posterodorsal edge of Meckel’s groove.
The scar extends for three teeth in L.
partitus and for four in L. etheridgei.
Finally, the anterior inferior alveolar
foramen and the mylohyoid foramen are
joined as a single opening in about half
the species. However, this feature may
be subject to ontogenetic change.

Examination of dentary characters re-
veals features that easily distinguish the
fossil species L. etheridgei, L. partitus,
L. apertosulcus, and L. cuneus from the
extant forms and, in fact, the latter three
extinct species may be related on the ba-
sis of the enlarged intramandibular sep-
tum. However, there is more difficulty in
distinguishing the living forms from one
another. Osteological features can be
useful for identifying species, but appar-
ently they are too subtle to show rela-
tionships within the genus. This point
is demonstrated further by lizard fossils
recently discovered from the middle
Miocene of Nebraska, which currently
are being described as an extinct species
of Leiocephalus by Carl Wellstead who
kindly provided me with illustrations and
descriptions for this discussion. Although
this North American species is indeed
distinct from any fossil or recent West
Indian member of the genus, the differ-
ences are within expected morphological
variation for these lizards. It appears
from the available material, though, that
among the trends in the evolution of
Leiocephalus are: 1) reduction of the
angular, 2) narrowing and shortening of

PLEISTOCENE PUERTO RICAN HERPETOFAUNAS 43

the splenial in its placement within Mec-
kel’s groove, 3) posterior extension of the
dentary onto the surangular and 4) loss
of the intramandibular septum and fu-
sion of Meckel’s groove. That these
transformations are more evident and
complete in other tropidurine genera
suggest that, on these characters, Leio-
cephalus is the least derived member of
the group.

Scincidae

Mabuya mabouya Lecépéde

Material—Blackbone 1: dentaries
(1 right, 4 left—USNM 259205-6) ; parie-
tals (I—USNM 259207); basales (1—
USNM 259208); pelves (1 right—USNM
259209); dorsal vertebrae (31—USNM
259210); sacral vertebrae (1—USNM
259211).

Description—Two of the five den-
taries referred to this species are com-
plete. Their tooth rows measure 8.5 and
8.2 and they have 31 and 28 teeth or
empty alveoli, respectively. The larger
specimen is shown in Figure 19. The
dentaries of Mabuya mabouya have
smooth and convex labial surfaces.
Lingually, the dental row is set in a
shallow horizontal trough. Meckel’s
groove is closed anteriorly, but open for
the posterior third of the jaw. On the
fossils the articular surface for the
splenial is visible throughout most of the
dorsal edge of the Meckelian opening.
An intramandibular septum is fused in-
ternally to the ventral surface of Mec-
kel’s canal and extends for three fourths
of the dentary length. Teeth are straight
and simple with lingually directed, stri-
ated, conical cusps.

The other fossils are not remarkable
in their comparison with recent skeletal
material of Mabuya mabouya. The fossil
parietal is missing the right nasal proc-
ess. It measures 4.7 across the posterior
border and 4.9 along the left side. The
basale is very small, 3.2 in width. The
pelvis is missing most of the ischium,
but the ilium is complete and 7.2 in
length from its distal end to the center

of the acetabulum. All the dorsal verte-
brae are similar to one another—with
elongate centra and neural arches and
long sloping neural spines that extend
past the postzygopophyses. The average
length of the centra is 2.5.

Comments.—A minimum of five in-
dividuals is represented by the fossils,
the largest of which had an estimated
snout-vent length of 99. The largest of
155 alcohol specimens of this species in
the University of Kansas herpetology col-
lection is 105.

Mabuya mabouya occurs on all larger
islands and cays of the Puerto Rican
Bank and ranges throughout the Lesser
Antilles and South and Central America.
Little is known of its habits in the West
Indies, but the lizard is most often en-
countered at forest edges near logs and
fallen trees in sunny locations. Dunn
(1935) made an attempt at subspecific
recognition in the island populations.
Greer (1970) included the genus in his
review of scincid lizard subfamilies in
which he elaborated on the osteology of
the family.

Teiidae
Ameiva exsul Cope

Material—Blackbone 1: dentaries
(12 right, 9 left—USNM 259212-3); ar-
ticular + surangular (5 right, 4 left—
USNM 259214); coronoids (83—USNM
259218); maxillae (10 right, 10 left—
USNM 259215-6); frontals (83—USNM
259219-20); parietals (3—USNM 259221);
basales (6—USNM 259222-3); quadrates
(2 right, 5 left—USNM 259217); pala-
tines (2 right—USNM 259224); scapu-
lae (9 right, 5 left—USNM 259225);
pelves (3 right, 2 left—USNM 259226);
atlas (I—USNM 259227); dorsal verte-
brae (90—USNM 259227-30); sacral
vertebrae (1I—USNM 259231); caudal
vertebrae (21—USNM 259228).

Cuevo del Perro: dentaries (6 right,
3 left—KUVP 11473); maxillae (1 right
—KUVP 11476); parietals (1—KUVP
11473); pelves (2 right, 3 left—KUVP

44 MISCELLANEOUS PUBLICATION MUSEUM OF NATURAL HISTORY

11475); vertebrae (5 dorsal, 1 sacral, 1
caudal—KUVP 11472).

San Miguel Cave: dentaries (2 right,
2 left—USNM 259241); maxillae (2
right, 2 left—USNM 259242); parietals
(1—USNM 259243); vertebrae (1 dor-
sal, 1 sacral, 3 caudal—USNM 259244).

Blackbone 2: dentaries (1 right, 2
left—USNM 259232); articular + su-
rangular (1 right, 1 left—USNM
259234); pelves (1 right—USNM 259236);
interclavicle (I—USNM 259235).

Cueva Clara: dentaries (2 right—
KUVP 11526); maxillae (1 right—KUVP
11526); basales (1I—KUVP 11564); ver-
tebrae (1 sacral, 1 caudal—KUVP 11564).

Barahona IV: maxillae (1 left—
USNM 259245); frontals (1—USNM
259246); parietals (I—USNM 259247);
pelves (2 right—USNM 259248); verte-
brae (1 dorsal—USNM 259248).

Nesophontes Cave: dentaries (1
right, 1 left—USNM 259237); maxillae
(1 right, 1 left—USNM 259238); basales
(1—USNM 259239); vertebrae (5 dor-
sal—USNM 259240).

High Cave: dentaries (1 right—
USNM 259249); maxillae (1 right frag-
ment—USNM 259259); vertebrae (1
dorsal—USNM 259250).

Guanica Bat Cave: dentaries (1 right
—USNM 259250); articular + surangu-
lar (1 left—USNM 259251).

Description—Fossil dentaries of
Ameiva exsul are common in occurrence.
The bones from all localities are variable
in size and exhibit ontogenetic varia-
tion. In juveniles and smaller individuals
the dental shelf is horizontal throughout
the anterior two-thirds of its length. Pos-
teriorly the shelf slopes up with the
widening of Meckel’s groove. In adults,
the dental shelf is horizontal for the an-
terior half of the bone before Meckel’s
groove widens. This gives a distinctive
arc to the tooth row (Fig. 19). The
labial face of the dentary is smooth and
gently convex in juveniles, but in adults
the posterior margin on the labial side
bulges out in a prominent angle at the
bottom of the coronoid scar. Sculpturing

in the form of horizontal grooves is pres-
ent on the ventral surface of dentaries
over 20.

Anterior dentary teeth are small and
unicuspid. The larger posterior teeth
are bicuspid. The transition from uni-
cuspid to bicuspid teeth is variable, but
usually occurs between the fifth and
tenth teeth. The posterior teeth of juve-
niles and occasionally the ultimate and
penultimate teeth of adults may be tri-
cuspid. Blunt, molariform crowns are
presént on a few of the largest speci-
mens. In the fossils a tendency exists
for an increase in tooth count with
lengthening of the dental row up to 17.
Tooth counts ranged from 17 to 26 in
dentaries 13 to 23 in length. Tooth rows
over 17 in length have a mean and mode
of 23 teeth.

Pertinent dimensions of other fossil

Fic. 19.—Fossils of Mabuya mabouya (top,

dentary USNM 259205), Ameiva exsul (middle,

dentary USNM 259212; bottom, maxilla USNM
259216). Scale equals 5 mm.

PLEISTOCENE PUERTO RICAN HERPETOFAUNAS A5

skull bones are given in Table 4. All are
similar to the same bones of Ameiva
undulata parva described by Fisher and
Tanner (1970) and they need not be
discussed here. A dentary, maxilla,
basale, and frontal are shown in Figs.
19 and 20. Pelvic girdle elements are
common in the assemblages and it is
worth pointing out their most diagnostic
feature. The ilia are short and very ro-
bust with high, recurved iliac processes.
The acetabulum is deep and both the
pubis and ischium are widely expanded
distally.

Comments.—Ameiva exsul is a con-
spicuous component of Puerto Rico’s
modern herpetofauna. The lizard is pri-
marily a beach denizen, although it also
is found in sunny, open habitat below
150 m (Heatwole and Torres, 1967). It
occurs sympatrically with the other
Puerto Rican Ameiva, A. wetmorei,
which is restricted to the southwest and
Isla Caja de Muertos off the coast from
Ponce. I was unable to discern osteo-
logical differences between these two
species other than size, but admittedly
my comparative material of A. wetmorei
was meager. The average snout-vent
length of Ameiva exsul is 125, but some
individuals may reach 200 (Heatwole
and Torres, 1967). The average snout-
vent length of adult Ameiva wetmorei
is 50 to 60. The fossils are all from in-
dividuals over 60 and I have assigned
them to A. exsul on this basis. Snout-
vent length estimates from various skele-
tal elements are, as follow: dentaries
134 (92-165), maxillae 123 (94-152),
basales 151 (103-194), frontals 128 (116-
139), parietals 126 (104-152), scapulae
149 (97-190), sacral vertebrae 118 (102-
135).

Anguidae
Diploglossus pleei Duméril and Bibron

Material—Nesophontes Cave: den-
taries (8 right, 5 left—USNM 259258-9) ;
articulars (5 left—USNM 259260); max-
illae (3 right, 2 left—USNM 259261-2);
frontals (1—USNM 259263); parietals

(2—USNM 259264); basales (1 com-
plete, 2 fragments—USNM 259265-6);
pelves (2 right, 4 left—USNM 259267);
vertebrae (28 dorsal, 14 sacral, 38 caudal
—USNM 259268); osteoderms (15—
USNM 259299).

Blackbone 1: dentaries (1 left frag-
ment—USNM 259252); articular (1
right—USNM 259253); maxillae (1 right
—USNM 259254); frontals (1 right half
—USNM 259255); pelves (1 left—USNM

Fic. 20.—Fossils of Ameiva exsul. Basale (top

—occipital view, middle ventral view, USNM

259222), frontal bone (bottom—dorsal view,
USNM 259220). Scale equals 5 mm.

46 MISCELLANEOUS PUBLICATION MUSEUM OF NATURAL HISTORY

TaBLE 4.—Selected measurements (mm) of cer-
tain fossil elements of Ameiva exsul.

n x Range

Dentary:

Tooth row length 32 186 13.0-23.4
Articular + surangular:

Width* 12 66 44 9.1
Maxilla:

Tooth row length 17 14.7 11.0-17.8
Frontal:

Midsagittal length 4A 12.5 11.3-13.5
Parietal:

Greatest length’ 5 10.0 8.2-12.0
Basale:

Length® 8 OS Sag

Width* 15.2 10.4-19.6
Palatine:

Length 2 94 £8.3-10.4
Quadrate:

Height 7 A8 3.3- 6.5
Scapula:

Length® 12 123 7.92155
Sacrum:

Width across

diapophyses 14 105 9.0-12.0

‘Measured from lateral lip of condylar sur-
face to tip of angular process.

* Measured from frontal border to tip of
supratemporal process.

* Measured from apex of angle formed by
pterygoid processes to lip of occipital condyle.
“Measured across paraoccipital processes.

° Measured along anterior border.

259257); vertebrae (16 dorsal, 1 sacral,
1 caudal—USNM 259256).

San Miguel Cave: dentaries (1 left
—USNM 259269); articular (1 right—
USNM 259270).

Barahona IV: maxillae (1 partial left
—USNM 259271); pelves (1 right—
USNM 259272).

Description—Dentary: 14 complete
dentaries have tooth rows ranging in
length from 6.6 to 10.8 («=9.1). The
average height is 3.0 at the position of
the last tooth. Eighteen to 20 teeth or
empty alveoli are present. The first two
teeth are simple, terminating in conical
crowns. Numbers four, five, and six are
similar, but their crowns are more trun-
cate. The crowns of the remaining teeth
are more truncate still and slightly meso-
distally compressed. Four or five mental
foramina penetrate the labial face be-
tween the second and fourteenth teeth,
but this surface is otherwise smooth and

convex over the length of the dentary.
On larger specimens, the anterior end
often is irregularly scored. The coronoid
overlap is present posteriorly between
the surangular and coronoid processes
and extends forward to the penultimate
tooth. Meckel’s groove is open the
length of the dentary, but nearly closed
between the seventh and eleventh teeth.
An intramandibular septum with a fused
ventral border is visible posterior to the
fourteenth tooth. A fossil dentary is
shown in Fig. 21.

Articular + surangular: the largest of
five bones from the post-coronoid por-
tion of a mandible is 14.9 in total length
and 2.6 wide anterior to the articular
condyle. The quadrangular-shaped retro-
articular process extends 3.7 beyond the
posterior border of the condyle. The
angular process is a vertically oriented
knob on the medial side of the retroar-
ticular process.

Maxilla: the most complete of five
maxilla referred to this species is 9.1 from
the premaxillary process back to the
jugal process and 2.0 from the parapet
of the jaw to the top of the nasal process
(Fig. 21). This bone has fourteen teeth
similar to those on the dentary. The four
posterior teeth are the smallest and have
the most truncated crowns. The teeth
in the middle of the dental row are more
chisel shaped. Five foramina penetrate
the labial face in an irregular row be-
tween the third and ninth teeth. Three
to five other foramina penetrate the an-
terior end of the nasal process and a
single foramen is located high on the
posterolabial side of the premaxillary
process. Two scars occur on the medial
edge of the alveolar border for articula-
tion with the palatine and vomer.

Frontal: the unfused frontals of
Diploglossus pleei are nearly square. The
largest complete fossil specimen (Fig.
22.) is 7.9 in mid-sagittal length and 7.3
across the parietal border. Fused osteo-
derms cover the bone dorsally except on
the recessed nasal process. The subor-
bital processes are wide at their bases

PLEISTOCENE PUERTO RICAN HERPETOFAUNAS AT

Fic. 21.—Fossil jaws of Diploglossus pleei.

Lingual view of dentary (top, USNM 259258),

labial view of maxilla (bottom, USNM 259261).
Scale equals 3 mm.

and protrude medially as spoon-shaped
flanges. Each frontal articulates anteri-
orly with the nasal and laterally with
the prefrontal, posterolaterally with the
postfrontal and posteriorly with the
parietal. The pre- and postfrontals do
not make contact laterally, although the
prefrontal extends back over half the
length of the frontal.

Parietal: two nearly complete parie-
tals were recovered from Nesophontes
Cave. The larger is 6.3 across its frontal
border and 8.5 from the frontal border
to the end of the supratemporal process.
The ventral surface is concave for its an-
terior two-thirds, but rises thereafter to
a rounded summit bearing the parietal
fossa. Two lateral ridges converge to-
ward the fossa from the frontal border.

Basale: a single basale and two sepa-
rate otic elements are approximately of
the same size and proportions. The
basicranium consists of the otic capsules
and occipital bones (Fig. 22). The
basisphenoid is absent from the fossil.
The distance between the sphenoccipital
tubercles is 3.4.

Vertebrae: the vertebral centra of
this lizard are smooth and flat and the

condyles are depressed and elliptical.
Sacral vertebrae are fused at the zygapo-
physes as well as at the transverse proc-
esses distally. Fracture planes on au-
tonomic caudal vertebrae pass in front
of the transverse processes.

Comments.—The generic allocation
of West Indian diploglossine lizards has
fluctuated widely over the years. As
many as four genera have been proposed
—Diploglossus, Celestus, Sauresia and
Wetmorena (Cochran, 1941; Grant, 1940;
Schwartz, 1970), and as few as two—
Diploglossus and Wetmorena (Under-

Fic. 22.—Fossils of Diploglossus pleei. Basale

(top—occipital view, USNM 259265), frontals

(bottom—dorsal view, USNM 259263). Scale
equals 3 mm.

48 MISCELLANEOUS PUBLICATION MUSEUM OF NATURAL HISTORY

wood, 1959, Meszoely, 1970). Recently,
four genera were again proposed by
Strahm and Schwartz (1977), who resur-
rected Celestus and Sauresia in a com-
pelling analysis of osteoderm structure.
Retained in the genus Diploglossus are
the species anelpistus, delesarga, montis-
serati, pleei, and warreni. Species as-
signed to Celestus are barbouri, costatus,
crusculus, curtissi, darlingtoni, duques-
neyi, fowleri, hewardi, marcanoi, micro-
blepharis, occiduus, and stenurus. Sau-
resia includes two species, sepsoides and
agasepsoides while Wetmorena remains
monotypic (W. hatiana).

I have examined the skulls of most
of these species and found little to dis-
tinguish them from one another except
for size. The intramandibular septum is
one of the more variable characters in
that 1) the ventral border may be fused
entirely, 2) fused anteriorly but not pos-
teriorly or 3) free ventrally altogether.
It is completely fused in C. crusculus
and both species of Sauresia; free pos-
teriorly in W. hatiana, C. curtissi, C.
darlingtoni, C. hewardi, and C. costatus
and completely free ventrally in D. war-
reni, C. barbouri, C. stenurus, and D.
pleei. The fact that the intramandibular
septum is completely free in recent skele-
tons of D. pleei and fused in the fossils
is, at present, inexplicable. Moreover,
the function of the structure is unknown.
Perhaps it is associated with fossorial
adaptations—i.e., fusion of the ventral
border providing reinforcement to the
jaw. I found no evidence that a free or
fused ventral border was related to size
or ontogeny. The fossils of D. pleei are
the largest diploglossines in which I ob-
served the fused condition. Estes (1964)
discussed this structure in other anguino-
morph lizards.

One dentary character distinguishing
D. pleei from other West Indian diplo-
glossines is the angular process at the
back of the bone. In this species, the
angular process extends backward well
past the level of either the supra-angular
process or coronoid process above it. The

angular process is parallel with the
supra-angular process in Sauresia, D.
warreni and C. stenurus, and shorter than
the supra-angular process in C. curtissi,
C. darlingtoni, C. barbouri, C. hewardi,
C. crusculus, C. hewardi, and W. hati-
ana. It is absent from C. costatus.

The frontals of D. pleei are unusual
in that they are nearly square instead of
rectangular, a condition shared with
Wetmorena hatiana. There is little me-
dial constriction to the frontals of D.
pleei, and the suborbital processes are
shorter and more robust than in other
species.

Much more study of West Indian
and mainland diploglossine lizards will
be necessary before their relationships
are understood. For now, I have found
no skull characters indicating four genera
within the West Indian species. Certain
characters are useful in distinguishing
one species from another, but overall, the
similarity in skull morphology bespeaks
a close relationship amongst the group.
Further analysis of osteology and integu-
mentary structure is recommended.

Living Diploglossus pleei reach a
snout-vent length of about 110. The fos-
sil individuals had an average of 135
(98-160) based on dentary length and
135 (100-147) estimating from length of
the maxillae.

Amphisbaenia
Amphisbaenidae

Amphisbaena sp. indet.

Material—Blackbone 1: mandibles
(1 partial left—USNM 259273); verte-
brae (2—USNM 259274).

Description—Two vertebrae and a
nearly complete left mandible are simi-
lar to those of living Amphisbaena. The
mandible (Fig. 23) is 4.5 as measured
along its ventral border. The tooth row,
bearing eight pleurodont teeth, is 3.1.
The three anteromost teeth are pointed,
recurved, and larger than those that fol-
low. The five posterior teeth are equal
in size except for the last which is

PLEISTOCENE PUERTO RICAN HERPETOFAUNAS 49

smaller. The teeth arise from a narrow
dental shelf at an oblique, anteriorly
directed angle and terminate as blunt
cones. Overall, the mandible is high,
linguilabially compressed and has a steep
coronoid with a broken tip. Two fo-
ramina penetrate the labial side below
tooth 4 and tooth 6. A single foramen
is present anteriorly just lateral to the
symphysis.

The two vertebrae have centra 2.1
and 2.3 long including the condyles. The
neural arches are low, constricted in
their middle and the pre- and postzyga-
pophyses are broad and directed ob-
liquely away from the midsagittal line.
A rib tubercle is present on either side
of the cotyle inferior to the prezygapo-
physes.

Comments.—Four of the ten species
of Amphisbaena occurring in the West
Indies are endemic to Puerto Rico, and
a fifth, A. fenestrata, is restricted to the
Virgin Islands. On Puerto Rico A.
schmidti is known from the northwestern
limestone region west of Dorado and
A. caeca occurs throughout most of the
island. Conceivably the fossils could rep-
resent either of these species. Amphis-
baena bakeri is believed to be confined
to the mountains between Mayagiiez and
Lares, but formerly its range may have

Fic. 23.—Fossil dentary of Amphisbaena sp. in
lingual view (USNM 259273). Scale equals

2 mm.

included cave regions to the east. Am-
phisbaena xera is restricted to the arid
southwest corner of the island. Appar-
ently, the fossil individual was a sub-
adult with an estimated snout-vent length
of 130. As adults, the five species on
the Puerto Rican Bank exceed 200 ac-
cording to Thomas (1966), who re-
viewed the group. Other Antillean spe-
cies are treated by Gans and Alexander
(1962).

Serpentes
Typhlopidae
Typhlops sp. indet.

Material—Blackbone 1: vertebrae
(35 midbody, 1 caudal—USNM 259275-
6).

Description—Some of the snake ver-
tebrae from Blackbone | are referred to
Typhlops because of their small size,
depressed neural arches, rudimentary
neural spines, well-developed accessory
processes, and knoblike synapophyses for
rib articulation. An example is shown
in Figure 24. List (1966) described the
thoracolumbar vertebrae of typhlopid
snakes, pointing out the flattened centra
penetrated by one or two foramina on
the ventral surface. The foramina may
be laterally paired or single and median,
but they usually are present in most spe-
cies of Typhlops. The Puerto Rican
blind snake, Typhlops richardi, is a spe-
cies that List included among those
bearing a single median foramen. This
condition obtains in 28 of 35 fossil verte-
brae. Most of the fossils are equal in
size except for six which are one half as
large. The largest vertebra has a cen-
trum 2.4 in length including the condyle.
The single caudal vertebra is distin-
guished by an enclosed hemal arch.

Comments.—Three species of Ty-
phlops inhabit the island today, but only
T. richardi occurs throughout the Puerto
Rican Bank. Typhlops rostellata is wide-
spread on Puerto Rico, but restricted to
mesic situations. The third species, T.
granti, is confined to the xeric southwest

50 MISCELLANEOUS PUBLICATION MUSEUM OF NATURAL HISTORY

part of the island. Although I made no
osteological comparisons of these three,
from List’s (1966) generalizations I
would not expect discernible differences
in vertebral structure.

Schmidt (1928) gave a range of 210
to 310 total length for 14 individuals of
T. richardi. Seventy-eight specimens of
T. rostellata examined by Ruthven and
Gaige (1935) varied in length from 104
to 227, while the maximum size they
gave for T. granti is 154. There is prob-
ably a minimum of two individuals rep-
resented by fossils according to size, but
this can not be said with much certainty.
Most of the vertebrae came from an in-
dividual in excess of 300, which makes
T. richardi a reasonable guess as to its
identity.

Boidae

Epicrates inornatus Reinhardt

Material—Guanica Bat Cave: den-
taries (1 right—USNM 259278); maxil-
lae (1 right—USNM 259279); palatine
(1 left—USNM 259280); vertebrae (11
anterior, 17 anterior trunk, 24 midtrunk,
7 posterior trunk, 4 caudal—USNM
259281-2).

Cueva Clara: dentaries (1 right—
KUVP 11524); maxillae (1 right—KUVP
11524); pterygoids (1 right—KUVP
11524); vertebrae (6 anterior, 17 anterior
trunk, 8 midtrunk, 1 caudal—KUVP
11524).

Barahona IV: vertebrae (34 mid-
trunkK—USNM 259285).

Barahona Dump No. 1: vertebrae
(3 midtrunk—USNM 259286).

Blackbone 1: vertebrae (2 midtrunk
—USNM 259277).

Nesophontes Cave: vertebrae (2
midtrunk—USNM 259283-4).

Above Horrible Bat Cave: vertebrae
(1 midtrunk—USNM 259287).

Description —The two dentaries, one
each from Guanica Bat Cave (Fig. 25)
and Cueva Clara are very similar to one
another. Their tooth rows are 24.9 and
24.0 respectively. There are positions
for 17 teeth, although neither specimen

has a full complement. The anterior
three or four teeth are twice as large as
those that follow, but all are sharply re-
curved. Posteriorly on the labial surface
is a large, wedge-shaped emargination
extending craniad to the level of the
twelfth tooth. It is occupied in life by
the surangular process of the compound

Fic. 24-—Fossil vertebrae in dorsal view of

Typhlops sp. (top, USNM 259275); cf. Alsophis

portoricensis (middle, USNM 259289); cf.

Arrhyton exiguum (bottom, USNM 259296).
Scale equals 2 mm.

PLEISTOCENE PUERTO RICAN HERPETOFAUNAS 51

bone. A single mental foramen pene-
trates the surface below the fourth and
fifth teeth.

The maxilla from Guanica Bat Cave
is the section posterior to the medial
palatine process. The maxilla from Cueva
Clara is nearly complete. This fossil,
about two-thirds as large as the Guanica
specimen, has a tooth row 23.3 long and
positions for 19 teeth.

The single palatine bone is like that
of modern E. inornatus. It has 6 teeth
or empty alveoli and a thick, dorsome-
dially directed process forming a roof
over the pterygoid notch at the back of
the bone.

The expanded central portion of a
left pterygoid is missing both ends of
the bone, but it is otherwise unremark-
able by comparison to the same structure
in living E. inornatus. A row of small
teeth and alveoli extends forward from
the middle of the ventral surface at the
level of the maxillary articulation.

Most of the fossils referred to this
snake are vertebrae from various regions
of the body. The generalized vertebra
of E. inornatus has a neural arch that is
emarginate in back and swollen above
the zygantra. The zygapophyseal facets
are oval and lie in a horizontal plane.
The accessory processes are small and
the condyles are round. There are ap-
proximately 230 presacral vertebrae in
living individuals. The first 40 vertebrae
are slightly broader than long and have
thick, vertical neural spines and caudally
directed hypopophyses. The hypopophy-
ses on the next 20 vertebrae gradually
diminish in size to form a midventral
ridge along the centra. There are about
120 thoracic or midtrunk vertebrae that
follow these. The midtrunk vertebrae
increase in size and become much
broader than long. The midventral hy-
popophyseal ridge widens, and the neu-
ral spine terminates in a flared rectangu-
lar cap (Fig. 25). The hypopophyseal
ridge is narrow on the remaining pre-
sacral vertebrae, but the ridge expands
into the base of the centrum cotyle.

These posterior trunk vertebrae are
smaller than those at midbody and are
approximately as long as wide. The
caudal vertebrae have transverse proc-
esses that curve out and down. The
paired hemal spines are short and di-
rected backward.

Comments.—In their review of
Hispaniolan Epicrates, Sheplan and
Schwartz (1974) recognized two species
in greater Puerto Rico—E. inornatus on
Puerto Rico and E. monensis on Mona
Island and the U.S. and British Virgin
Islands. The authors stated (p. 100)
that “. .. no other complex within An-
tillean Epicrates is more puzzling than

Fic. 25.—Fossils of Epicrates inornatus. Den-

tary (top, USNM 259278); trunk vertebra

(middle—dorsal view, bottom—lateral view,
USNM 259281). Scale equals 7 mm.

52 MISCELLANEOUS PUBLICATION MUSEUM OF NATURAL HISTORY

those taxa associated with Greater Puerto
Rico.” Part of the reason for this enigma
is that the Mona Island subspecies, E. m.
monensis, is known from very few speci-
mens. It is most peculiar that the range
of E. inornatus divides that of E. monen-
sis. Sheplan and Schwartz accounted for
this by proposing that at one time Puerto
Rico was inhabited by two boas, a very
large one, E. inornatus, and a small one
E. monensis, that became extinct. If so,
and the idea makes sense, the small one
is not present in any of the fossil de-
posits. The fossils are from very large
snakes. The smallest had an estimated
snout-vent length of 1300, whereas the
largest exceeded 1650. The maximum
size known for E. monensis is 770,
whereas E. inornatus may reach 1860
(Sheplan and Schwartz, 1974). The au-
thors regarded E. monensis as quite dis-
tinct from either of the two small His-
paniolan species, E. fordii and E. gracilis.
Epicrates inornatus is believed to re-
semble closely E. subflavus of Jamaica.

Little is known of the habits of E.
inornatus, although it is widely distrib-
uted over the island, typically in wooded
situations from sea level to 1050 meters
at El Yunque. It may be associated with
caves and talus slopes. Epicrates angu-
lifer is partly cavernicolus. This species
has been taken near cave entrances in
Cuba and it is known to feed on bats.
If these habits are applicable to E. inor-
natus, the fossil assemblage could be
anticipated.

Colubridae

cf. Alsophis portoricensis
Reinhardt and Liitken

Material—Blackbone 1: vertebrae
(10 trunk, 3 caudal—USNM 259288).

Barahona IV: vertebrae (12 trunk,
6 caudal—USNM 259294-5).

Nesophontes Cave: vertebrae (7
trunk, 1 caudal—USNM 259293).

Cueva Clara: vertebrae (4 trunk—
KUVP 11524).

Blackbone 2: vertebrae (2 trunk—
USNM 259289-90).

Rosaria River: vertebrae (2 trunk—
USNM 259292).

Guanica Bat Cave:
trunk—USNM 259291).

Description—Vertebrae from seven
caves are tentatively assigned to the
Puerto Rican ground snake, Alsophis
portoricensis, because of their similarity
to this species. I have withheld defini-
tive assignment because of the current
systematic chaos among West Indian
xenodentine snakes. However, the fos-
sils can be distinguished from the only
other Puerto Rican colubrid, Arrhyton
exiguum, by several features (Fig. 24).
The most obvious difference is size. On
the average, Alsophis portoricensis is
twice as large as Arrhyton exiguum—
923 versus 438 maximum _ snout-vent
length (Schwartz, 1966). The fossil ver-
tebrae of A. portoricensis have an aver-
age centrum length of 5.9 and an inter-
prezygapophyseal width of 4.7. The
midbody vertebrae of A. portoricensis
have elliptical zygapophyseal facets with
accessory processes projecting anterolat-
erally. In A. exiguum, the zygapophyseal
facets are circular and the accessory
processes project laterally. The zygo-
sphenal arch has a more undulating bor-
der in A. exiguum because the zygo-
sphenes extend farther forward. The
neural spines of A. portoricensis are high
and narrow in comparison with those of
A. exiguum, in which the spines are low
and the neural arches themselves are de-
pressed. The condyles of both species
are round and oriented slightly dorsad.
A midventral ridge is present on the
centra of both species, but in A. exiguum
the ridge is constricted anteriorly, form-
ing a depression on either side of the
ridge at the base of the cotyle.

The comparative differences in verte-
bral structure between these two snakes
are subtle and intra-columnar variation
tends to obscure their distinctions. In
all, however, they are sufficient to differ-
entiate the two species.

vertebrae (1

Comments.—The fossil vertebrae re-
ferred to Alsophis portoricensis probably

PLEISTOCENE PUERTO RICAN HERPETOFAUNAS 53

represent single individuals from each
of the localities where they are found
except at Blackbone 1 where three dis-
crete sizes are present. Estimates of
snout-vent length are based on measure-
ments between pre- and postzygapophy-
ses, across the accessory processes an-
teriorly, and midventrally from the edge
of the cotylar lip to the end of the
condyle. The smallest individual from
Blackbone 1 had a snout-vent length of
325 to 350. The largest, from Cueva
Clara, was 780 to 850, whereas the aver-
age snout-vent length of snakes from all
caves was 700 to 725 (630-765).

The classification of West Indian
colubrid snakes has never been stable
and over the years the number of genera
and species has continued to expand and
contract. Dunn (1932) was first to make
a serious attempt at working up a large
segment of the group, mainly the Greater
Antillean forms, and he relied heavily
on a few superficial characters, for ex-
ample sensory pits. More recently, Mag-
lio (1970) addressed the problem using
skull osteology. He compared all An-
tillean species and several mainland
forms and proposed separate origins and
derivations for the two Puerto Rican spe-
cies. Maglio allied A. exiguum with the
funerus species group of Jamaica. Also-
phis portoricensis was placed in the
cantherigerus group whose ancestral
populations were to have island-hopped
southeasterly from Cuba. Maglio’s (1970)
treatment of West Indian colubrids re-
mains to be tested and current work on
other neotropical zenodontine snakes
suggests that major systematic revisions
are still in order (Alan Savitzky, pers.
comm. ).

Alsophis portoricensis ranges through-
out the Puerto Rican Bank in a multi-
tude of habitats. Although it is rare or
extinct on Vieques, St. Croix, Tortola,
and St. Thomas, it persists on smaller
cays and islets. The species is locally
common on Puerto Rico, but in recent
years no collector has been able to amass
large series from any one locality and

apparently its restriction coincides with
the introduction of the mongoose
(Schwartz, 1966).

cf. Arrhyton exiguum Cope

Material—Blackbone 1: vertebrae
(20 trunk, 3 caudal—USNM 259296-7).

Nesophontes Cave: vertebrae (2
trunk—USNM 259298).

Description—The general morphol-
ogy of the vertebrae is discussed under
the preceding description of Alsophis
portoricensis. The vertebrae of A. ex-
iguum are small; the largest fossil (Fig.
24) is 3.7 between pre- and postzygapo-
physes, the smallest 1.0. The anterior
trunk vertebrae have caudally directed
hypopophyses and the neural spines pro-
ject beyond the posterior border of the
neural arch. Their centra are shorter
and narrower than the vertebrae pos-
terior to them. The midbody vertebrae
are elongate and lack hypopophyses, but
retain a low hypopophyseal ridge. The
neural arches are depressed. Caudal ver-
tebrae have flat, anterolaterally directed
transverse processes with broad bases
that extend for over half the length of
the centra.

Comments.—Fossils of Arrhyton ex-
iguum from Blackbone | fall into three
sizes—individuals with estimated snout-
vent lengths of 115 to 125, 250 to 370,
and 420 to 440. The individual(s) from
Nesophontes Cave was (were) about
400. Schwartz (1967) gave maximum
sizes for this species of 438 for females
and 428 for males.

Arrhyton exiguum is a reclusive snake
not frequently met in the field. Its range
includes Puerto Rico, the Virgin Islands
(with the possible exception of St. John),
and many of the bank satellites. It is
found most often in trash and rubble of
fallen fronds and branches, and beneath
logs and rocks from sea level to the El
Yunque rain forest (Schwartz, 1967).

DISCUSSION

Comparison of the Cave Faunas.—
Although the fossil herpetofaunas de-

54 MISCELLANEOUS PUBLICATION MUSEUM OF NATURAL HISTORY

scribed here are not necessarily a com-
plete sampling of late Pleistocene spe-
cies, their diversity is remarkable and
deserves discussion. In all, 21 species
represent 16 genera and 12 families.
All modern genera of amphibians and
reptiles indigenous to Puerto Rico are
represented, with the exception of the
geographically restricted gekko, Phyllo-
dactylus. Two genera, Cyclura and Leio-
cephalus, no longer occur on the island.
The 47 species of amphibians and rep-
tiles living on Puerto Rico currently are
grouped into 13 genera and 12 families.
Six modern introductions are not includ-
ed: Hemidactylus brooki, H. mabouia,
Bufo marinus, Hyla cinerea, Osteopilus
septentrionalis and Rana _ catesbeiana
(Cochran, 1941; Trueb and Tyler, 1974;
Schwartz and Thomas, 1975).

Some fossil localities have more spe-
cies than others, and the number of in-
dividuals of any given species varies
among them also. Relative abundance
can be compared by determining the
minimum number of individuals (MNT)
of a species from each locality by count-
ing the most abundant skeletal element
from a particular side, either right or
left. For example, the single most abun-
dant skeletal element of Ameiva exsul
from Blackbone 1 is the right dentary,
of which there are 12. Thus, at least 12
individuals are represented as _ fossils.
The results of this tabulation are re-
corded in Table 5.

Anolis cuvieri and Ameiva exsul are
the most frequently occurring species as
well as the most abundant individuals.
Of the remaining 18 species, only 6 are
present in four or more of the localities.
These species are Epicrates inornatus
(7), Alsophis protoricensis (7), Anolis
cristatellus (6), Peltophryne lemur (6),
Leptodactylus albilabris (5), and Diplo-
glossus pleei (4). Evidently, the distri-
bution and abundance of species as fos-
sils is largely a consequence of prey
vulnerability and the predatory habits
of the owls, although the complexities of
predator-prey ecology are such that the

most abundant prey species in an area
may not necessarily be the most fre-
quently represented in the diet of the
predator. Tyto cavatica was obviously
an efficient predator, and barn owls in
general are opportunistic feeders (Craig-
head and Craighead, 1969), as demon-
strated below.

The North American barn owl, Tyto
alba, is common in the Greater Antilles,
and its diet has been reported from the
southern Bahamas, Hispaniola, and
Grand Cayman Island and recently sum-
marized by Morgan (1977). Morgan’s
comparisons of North American and
West Indian barn owl roosts showed
that, in general, the continental pellets
were composed almost exclusively of
mammals, whereas those from the West
Indies, both fossil and Recent, contained
birds, lizards, and frogs as well as small
mammals. This is a result of the lower
diversity and abundance of nonvolent
mammals in the West Indies, both now
and in the late Pleistocene. In both fos-
sil and Recent barn owl deposits studied
from the West Indies, birds are the most
commonly occurring class of vertebrates
and usually constitute a significant per-
centage of the total fauna. Bats are di-
verse in both fossil and Recent pellet
remains from West Indian barn owl
roosts, but they constitute a low per-
centage of the total number of indi-
viduals. Mammalian remains in modern
West Indian owl pellets are dominated
by Rattus and Mus, whereas the West
Indian fossil deposits contain a much
more diverse assemblage of all sorts of
vertebrates. On Puerto Rico, the only
mammal other than bats that was small
enough to be taken by Tyto cavatica was
the insectivore Nesophontes.

Reptiles and amphibians are ex-
tremely rare in North American barn
owl deposits, but they occur in all known
West Indian pellet remains in variable
numbers. Fossil pellets tend to have a
higher percentage of amphibian and rep-
tile species than do modern ones. The
newly collected birds and mammals from

55

PLEISTOCENE PUERTO RICAN HERPETOFAUNAS

jou SOY} SBUIpN[OUr seyeooy [Isso} [[e—|[V
vuoyeieg—A]eq *g suoqyor[g—Zgqq ‘ear

"1X9} 99g "poye[nqr}

‘RIZID BAMND—D ‘o1leg jep vAonp—Ieg ‘eAvH jeg voURNdg—uENy ‘faa"D jens ues—yug ‘AT
sojuoydoseNN—osen| *[ Suoqyorig—Tqq ‘S[eNpriArpur jo Jequinu winuUTMI—JNW ‘suonerAsiqqy ,

OTE
ik i
6T 6
61 9
90 G
€o0 T
ete TUL
VOL €€
vo YP
90 @&
Gy &T
Se Of
TOL Ge
Ly ST
GOI GS
90 G
€0 iT
€<Cl 68
671 LY
VOL €€
2 INIW
IV

GL
es oT
e8 tT
LOT @
es oT
ees 2

Z INW

ite)

r

9

1G

if,

VIG

9

TC

6

%

€
L

“CV

86

001
006

9 0°06
00r

GI 0'0¢

0'0Z

Or

I

6I
06
06
et IL 06
Tlé 7 081

GOI @ 06
SST € 06
Sct € 016
es tT

SSI € 06
eel TL

Sia

IT

T

oS Ti

(94) (|

(Si

€'8
8

00S
GV
€8

VG
6G
I 6G
6G

GL 6G
I
G ULI

ce
I
I
I

A oO

891
ST
ST
9'0
GT

Ano co

90

re

90

A
ret
AMMAArY TAA

Gol 16
896 SV
€VL VG

INN 4 INW

Jog

uens)

4 INN 2 INN

WUs AIP a

% INW 2 INW

odd

OSoN

2 INW
19a

[PIO],
wmnngixa UozhyLoy
sisuaa110j10d svydos;y
SsnyoUuLouL saqoinidy
‘ds sdojydhy,
soyxeus
‘ds nuangsiyduy
spruseqsiydury
vaajd snssojsojdiq
jnsxa DuaWwYy
pinoqow vinqgoyy
snqyiod “7
vadpiuayja snjpydasoiaT
smauid pimgohy
‘dds syouy
SN]]9}DIS1LO “VW
1ua1ano syouy
‘ds snjhzopposapyds
spiezry
Dyossnaap shwashiyD
sony,
mua, auhvydoyag
‘dds smphzopposayjnay 7
suqopiqio smhzoppojdaT
ss014q

“AyBOOT [Issoz red soloads Jo (%) souepunqe osANeier puke uoNNGEIsIqG—G aAIaV J,

56 MISCELLANEOUS PUBLICATION MUSEUM OF NATURAL HISTORY

the Puerto Rican caves have not been
described yet, and thus, the relative pro-
portion of the amphibians and reptiles
is unknown. Bird remains are plentiful,
however, and more plentiful in Black-
bone | than in the other localities (Storrs
Olson, pers. comm.). Both fossil and
modern deposits are similar in that the
herpetofauna often is dominated by one
or two species. For example, on Cayman
Brac Anolis sagrei accounted for nearly
47% of a fossil fauna, but seven other
lizard species were present, and on
neighboring Grand Cayman Island
Anolis conspersus constituted almost 65%
of a recent deposit (Morgan, 1977). On
Barbuda in the Lesser Antilles, three
species of Anolis composed 85% of the
lizard remains from one cave in which
seven lizard species were present (Ethe-
ridge, 1964a). The Puerto Rican faunas
follow this same pattern. Eleutherodac-
tylus dominates the highly diverse Black-
bone 1 fauna, and as noted previously,
only 8 amphibian and reptile species out
of 21 occur in 4 or more of 14 localities.
Overall, Anolis cuvieri and Ameiva exsul
account for over 26% of all individuals.
These two lizards are approximately
equal in size (125-150 SVL) and appar-
ently were optimum prey. They differ
in habits. A. cuvieri is a forest-canopy
dweller, whereas A. exsul is active on
the ground in open habitat. Both are
diurnal species, and Tyto cavatica may
have hunted for them in the early eve-
ning when A. exsul was still moving
about, but when A. cuvieri had stretched
out for the night on an exposed perch.
Cyclura pinguis and Epicrates inor-
natus are among the more frequent spe-
cies in the cave remains on Puerto Rico,
but they are represented by few indi-
viduals and constitute a low percentage
of the total fauna in any cave. Their
presence as fossils is probably not a re-
sult of predation, however, as adults of
Cyclura exceed 7,000 gm body weight
(Carey, 1975), whereas the average Tyto
alba weighs about 300 gm (Craighead
and Craighead, 1969). It is impossible

that Tyto cavatica, which was slightly
smaller than T. alba, preyed on lizards
of this size. However, juveniles may
have fallen prey as indicated by an in-
dividual with a snout-vent length of 150
in Blackbone 1. Juveniles would be ex-
posed to late afternoon predation, for
Cyclura pinguis reappears at this time
when ambient temperatures drop below
30°C (Carey, 1975). I have no data on
body weights of Epicrates inornatus, but
adults are too large to have been preyed
on by T. cavatica.

The total species diversity and abun-
dance of individuals from all localities
reflects an obvious bias from Blackbone
1 (Table 5). For example, frogs account
for 53.6% of the Blackbone 1 fauna and
37.6% for all localities combined. How-
ever, frogs account for no more than 20%
of the total fauna from any other cave
except Blackbone 2. Moreover, Eleu-
therodactylus spp. alone constitutes
nearly 27% of the Blackbone 1 herpeto-
fauna, but the genus is represented by
only one individual in each of two other
deposits. The absence of Eleutherodac-
tylus and other small species (Sphaero-
dactylus, Typhlops, small anoles) from
most of the caves is probably the result
of sampling bias. The fossil matrix from
Blackbone 1 was screened and picked
to a finer degree because of the large
quantity of bones and obvious antiquity
of the deposit. I did similar screening
of a smaller sample of fossil matrix from
Nesophontes Cave, but was unable to
recover any fossils that added to the
faunal list for that cave. The richness of
the Blackbone 1 fauna, whether a con-
sequence of sampling bias or not, affects
the sum total of diversity and relative
abundance in two ways. First, the Black-
bone 1 fauna greatly increases the total
diversity of species. Secondly, it alters
the total relative abundance of a species
in its favor, as for example Eleuthero-
dactylus. Thus, Blackbone 1 is exem-
plary of the potential diversity and abun-
dance of specimens in fossil owl pellet
deposits; the cave contains many species

PLEISTOCENE PUERTO RICAN HERPETOFAUNAS 57

that have a low probability of turning
up as fossils on account of small size or
secretive habits.

Some of the small, secretive species
might not have been prey of the barn
owl, but of the smaller Puerto Rican
screech owl, Otus nudipes, or to an ex-
tinct burrowing owl, Speotyto sp., re-
mains of both of which are common in
the Blackbone 1 deposits (Storrs Olson,
pers. comm.). These owls also appear
to have been preyed on extensively by
the barn owl. Although Wetmore (1922)
suggested that the screech owl may have
been responsible for the remains of small
birds in other Puerto Rican caves, both
this species and the forms of Speotyto
are largely insectivorous. Some of the
small, secretive reptiles could have been
ingested secondarily by the avian preda-
tors. Sphaerodactylus, Amphisbaena or
Typhlops may have been eaten by an-
other reptile, which, in turn, was con-
sumed shortly thereafter by an owl.
Alsophis portoricensis is known to be
partly ophiophagus (Schwartz, 1967)
and may have preyed on Amphisbaena
or Typhlops.

It is more difficult to compare the
faunas chronologically in the absence of
stratigraphic control, there being no cer-
tain means of establishing the relative
ages of the cave faunas nor the duration
of deposition in each case. Most deposits
probably overlap chronologically and
thus are partly contemporaneous. The
fossil material from Blackbone 1 may be
older than the other cave remains be-
cause of its discoloration and partial per-
mineralization. An approximate date of
20,000 y.b.p. is assigned cautiously to
the Blackbone 1 fossils and correlates
with climatological evidence discussed
in the next section. San Miguel, Neso-
phontes, Blackbone 2 and Barahona IV
Caves have similar species composition
and differ mainly in the relative abun-
dance of particular species. The color
and texture of the fossils from these
caves is similar and there are no obvious
morphological differences in the bones

that would suggest chronological trends.
Intraspecific variation is no greater be-
tween caves than within them, with re-
spect to tooth count, dermal sculpturing,
size, or other morphological features of
the fossils. Cueva del Perro may be close
in age to Blackbone 1 because it is the
only other locality in which the extinct
lizard Leiocephalus etheridgei occurs.
The absence of this lizard from the other
deposits suggests that the animal may
have become extinct prior to use of the
caves by owls. On the other hand, Cueva
del Perro also shares the extinct species
Leiocephalus partitus only with Guanica
Bat Cave, the fauna of which is rather
recent in appearance (color of fossils)
and composition. Cueva del Perro and
Blackbone 1 probably represent faunas
of some antiquity, and caves that re-
mained active as accumulation sites for
a long time.

Fossils of several species exhibit some
structural differences distinguishing them
from their living relatives. For example,
gigantism occurred within Leptodacty-
lus albilabris, Peltophryne lemur, Anolis
cuvieri, and perhaps Diploglossus pleei;
each achieved snout-vent lengths in ex-
cess of any individuals of their species
known today. The reasons for this are
unclear, but it has also been reported
in fossils of other Antillean lizards, such
as Aristelliger lar, Anolis ricordi, Anolis
sagrei, and Anolis bimaculatus (Ethe-
ridge, 1964a, 1965). Curiously, among
modern West Indian amphibians and
reptiles only the “giant anoles” (A. cu-
vieri, A. roosevelti, A. equestris, A. ri-
cordi), Diploglossus anelpistus, and D.
warreni still attain unusually large size.
Other morphological differences be-
tween fossil and Recent species include
the dentary sculpturing on male Anolis
cristatellus, the fused intramandibular
septum in Diploglossus pleei, and the
condition of Meckel’s groove in the two
new species of Leiocephalus. In each
instance, all fossils, regardless of locality,
differ from all modern specimens.

The Paleoenvironment.—lt is reason-

58 MISCELLANEOUS PUBLICATION MUSEUM OF NATURAL HISTORY

able to assume that the owls living in
Puerto Rican caves hunted near their
roosts rather than in distant areas of the
island. Their prey, therefore, was ob-
tained in the immediate vicinity, which
permits an estimate of the probable
habitat of the area based on ecological
requirements of the prey species. Al-
though there is a dearth of knowledge
on habits and life histories of most West
Indian amphibians and reptiles, the di-
versity of fossil species from the caves
of the Ciales region exceeds any pre-
viously described paleoherpetofauna in
the Antilles and invites interpretation in
light of current ideas on late Pleistocene
climatology.

Pleistocene glacial climatic fluctua-
tions have been given increased atten-
tion in recent years by geologists and
biologists. Current theory favors numer-
ous glacial advances and retreats during
the last half million years. However, the
latitudinal climatic gradient produced
by glaciations has yet to be determined,
and only the fact of severe world-wide
climatic fluctuations is established. Data
on fluctuations in paleotemperatures and
comcomitant changes in humidity have
been augmented by palynological studies
on both continents in the New World.
It is of particular interest that equatorial
as well as temperate regions were af-
fected by Pleistocene climatic fluctua-
tions.

In the West Indies, core samples
taken from the central Caribbean Sea
have yielded an undisturbed record of
late Pleistocene O18/O'® isotopes that
show fluctuations in sea water tempera-
ture indicative of climatic change (Emili-
ani, 1966; Emiliani and Rona, 1969).
Temperature correlations covering the
last 450,000 years show 17 major fluctua-
tions from about 22° to 28°C. Approxi-
mately 65,000 years ago the last full
interglacial ended and sea water began
cooling until a low of 22°C was reached
around 15,000 to 20,000 years ago. There-
after, a warming trend began which con-
tinues to the present. Emiliani’s data

were corroborated by Bonatti and Gart-
ner (1973) from an analysis of CaCOs-
free sections of the core showing fluctua-
tions in concentration ratios of kaolinite
to quartz. Kaolinite is a layered silicate
commonly produced by chemical weath-
ering during humid climatic conditions.
The kaolinite to quartz ratio is highest
during the warmer interglacials and low-
est during glacial maxima when com-
pared with Emiliani’s isotopic tempera-
ture curve. Thus, lower temperatures
also indicate periods of aridity because
cooler sea surface temperatures would
reduce the supply of water vapor to the
atmosphere.

Haffer (1974) discussed evidence
that, on land, Pleistocene glaciation in
northern South America produced alter-
nations in humid and dry climatic peri-
ods in the tropics that resulted in shrink-
age and expansion of lowland forests in
eastern Colombia and interior Guyana.
Indeed, Pleistocene pulsations appar-
ently had a profound influence on struc-
turing the modern Amazonian forest
biota (Simpson and Haffer, 1978; Haffer,
1979). Immediately north of the Greater
Antilles in Florida, pollen studies
strongly indicate dry climates in a period
from about 37,000 to 13,000 years ago.
Core samples taken from midpeninsular
lake sediments are dominated by pollen
characteristic of a savanna of dune scrub
and scattered stands of oak during that
period (Watts, 1975; Moran, 1975).

Let us relate this to the Antilles.
Lowland thorn-scrub characterizes the
West Indies, dominating areas with less
than four inches (10 cm) of rain per
month (Howard, 1973). Periodic flood-
ing coupled with poor drainage due to
serpentine and siliceous soils, alternates
with desiccation to produce retarded tree
growth, and hence, savannas. Savannas
and thorn-scrub occur on many islands
of low relief and also in rain shadows on
mountainous islands such as Puerto Rico
(Figs. 2 and 3). The cool, arid regime
65,000 to 15,000 years ago probably
favored a more extensive xeric flora on

PLEISTOCENE PUERTO RICAN HERPETOFAUNAS 59

Puerto Rico, as it did in northern South
America and Florida. Today, savannas
occur on the northwest coast of Puerto
Rico only at the cape region near Isa-
bella, but during the late Pleistocene
they probably were more widespread.
It is difficult to say just how far savannas
or a dry coastal flora reached into the
more interior regions of the island, or to
what elevations. The cave area around
Barahona is 180 m above sea level and
is characterized today by a mixture of
secondary, deciduous species with sea-
sonal flowering. The prevalence of fos-
sils of xerophilic amphibians and reptiles
supports a hypothesis that savannas or
a dry coastal flora extended to this re-
gion 10,000-20,000 years ago. Cyclura
and Ameiva, for example, are heliophilic,
essentially open country lizards that are
common as fossils from the Ciales region.
Cyclura is now extinct there. Ameiva
exsul is uncommon, being now practi-
cally restricted to coastal regions with
mostly xeric habitat. The high incidence
of Anolis cuvieri as a fossil from the cave
regions appears to contradict the argu-
ment because of the strictly arboreal
habits of this lizard. However, a closed
savanna adjacent to the more open areas
could account for its occurrence. More-
over, the lizard simply may have in-
habited lower elevations to compensate
for cooler montane temperatures. Anolis
cuvieri is not now restricted to upland
localities per se, but it requires dense
tree foliage and most of the habitat is
presently at higher elevations.

Few other of the commonly occurring
fossil species are directly helpful in de-
termining the paleoenvironment. For ex-
ample, Peltophryne lemur is known to-
day from widely scattered localities on
the island, and most of these are at ele-
vations no higher than the Barahona
Valley. The toad apparently has an af-
finity for more xerophytic habitats, but
this has not been confirmed. Lepto-
dactylus albilabris occurs over the island
in regions of suitable moisture for egg
laying, but it is a terrestrial species and

its presence as a fossil is not unexpected.
Diploglossus pleei also is fairly abun-
dant as a fossil. It is a shade-loving spe-
cies and a denizen of forest floors and
leaf litter. Today the lizard is found
principally in mesic situations in the in-
terior upland, although I have collected
it in the coastal lowlands north of Ciales
near Manati. Diploglossus pleei appar-
ently adapts to disturbed environments,
for it is found in the coffee plantations
on the mid-altitude slopes. Leiocephalus
is not especially abundant as a fossil, but
two extinct species are present. Most
Leiocephalus are xerophiles, but some,
such as the Hispaniolan L. melano-
chloris, are shade and forest dwellers
(Schwartz, 1965). Thus, it is difficult to
relate the Puerto Rican Leiocephalus to
a late Pleistocene xeric habitat, but I am
inclined to favor an arid regime for them.

After evaluating the composition of
the fossil herpetofauna, one is left with
the impression that the late Pleistocene
environment in the vicinity of the cave
region may have been in a transitional
zone between xeric, open habitat to-
wards the coast and a more closed, mesic
situation extending to the interior. Un-
fortunately, neither the flora nor the am-
phibians and reptiles presently occupy-
ing the Barahona Valley can be used to
test this hypothesis conclusively. For
example, Anolis cristatellus, Anolis pul-
chellus, and Eleutherodactylus coqui are
the most common species in the Ciales
region today, but they are equally com-
mon at low and middle altitudes in dis-
turbed and undisturbed habitat through-
out the island. In other words, these
species are not useful ecological indi-
cators because of their vagility and the
same is assumed for the Pleistocene.
(For further discussion of this point see
Rand and Williams, 1969.) Species such
as Ameiva exsul and Cyclura pinguis are
most helpful in interpreting the past en-
vironment because of their more special-
ized ecological requirements.

Interpretation of past environments
should account for any extinctions of the

60 MISCELLANEOUS PUBLICATION MUSEUM OF NATURAL HISTORY

animals that occupied them. The three
species in the fossil assemblages that
became extinct presumably were xero-
philic—Leiocephalus etheridgei, L. par-
titus and Cyclura pinguis. Their demise
may have coincided with a return to
humid conditions in the past 13,000
years. Cyclura pinguis could have per-
sisted into Subrecent times on Puerto
Rico because its remains on St. Thomas
are associated with kitchen middens
(Miller, 1918). Increasing humidity with
concomitant expansion of forest in the
Barahona Valley probably eliminated
Ameiva exsul from most of this area and
restricted populations to coastal areas.
Interestingly, southwestern Puerto Rico
in the vicinity of Guanica Bat Cave
boasts habitat seemingly suitable for the
arid forms that became extinct on the
other side of the island. Of the three,
only L. partitus occurs as a fossil from
the Guanica locality. The absence of
Leiocephalus etheridgei and Cyclura
pinguis from the Gudanica deposits is
enigmatic because the occurrence of L.
partitus suggests that the habitats were
reasonably similar on either side of the
island. Of course, the absence of fossils
does not prove that the lizards were not
in the vicinity.

The assumption that climatic changes
that took place in the past 13,000 years
were responsible for the extinction of
the three lizards does not explain how
the animals survived the previous inter-
glacials occurring over the last 450,000
years. Indeed, this requires that the spe-
cies were present and subject to the
rigors of changing environments with
potentially similar effects. Extinctions of
other vertebrates in the West Indies were
not uncommon in the late Pleistocene
and sub-Recent times and a general ex-
planation is desirable. A number of bird
species (many of them xerophyllic forms)
disappeared along with small mammals
(Olson, 1978). This is in contrast to the
North American continental fossil record
for the same period, where extinctions
mostly involved the mammalian mega-

fauna and associated vultures. The ma-
jority of the amphibians and reptiles
known as fossils from the late Pliocene
and Pleistocene are still with us to-
day (Gehlbach, 1965). Further data
on the North American fossil record
of amphibians and reptiles are being
reported rapidly (Holman, 1976 and
other works), and misidentification of
previously described fossils is not un-
common, which could mitigate this ar-
gument in either direction. One ex-
planation for the greater survival of
the North American herpetofauna dur-
ing the Pleistocene is that continental
species have broader distributions con-
taining multiple habitats and thus are
more likely to survive local extinctions.
Insular species with fewer and smaller
populations (i.e., absolute size) are not
as fortunate. Further documentation and
dating of West Indian fossils is essential
if this problem is to be resolved.
Zoogeographic Considerations—The
comparative recency of the fossil her-
petofaunas from Puerto Rico and other
West Indian islands precludes much in-
sight into their history. The fossils are
essentially modern, usually referable to
extant species and genera, and provide
little information for extending historical
biogeography into the Tertiary. The fos-
sil evidence does demonstrate, though,
that the modern herpetofauna of Puerto
Rico was well-established in the late
Pleistocene. Since then the distribution
of amphibians and reptiles has been sub-
ject to changing environments, by both
natural occurrence and human interfer-
ence, and to fragmentation of the Puerto
Rican Bank by eustatic rising in sea level.

Little is known of pre-Pleistocene sea
levels, but an estimate of 120 m has been
given for maximum Pleistocene lowering
(Donn et al., 1962). By 14,000 years ago
sea levels were 75 m below the present
level, though Puerto Rico and all of the
Virgin islands except St. Croix were still
a single land mass. Sea level rose to 15
m below present level after 8,000 years
ago, submerging small islands off the

MONA

MONA

MONA

PLEISTOCENE PUERTO RICAN HERPETOFAUNAS 61

PUERTO RICO

Maximum Pleistocene Lowering St CROIX

PUERTO RICO

14,000 - 12,000 B.P.

St. CROIX

PUERTO RICO

8,000 - 6,000 B.P. (ees.

St. CROIX

Fic. 26.—Fragmentation of the Puerto Rican Bank by eustatic rising in Pleistocene sea levels. The
islands were united as a single land mass during maximum Pleistocene lowering as shown at top.
Sea levels were about 75 m below present level in the middle diagram and fragmentation was es-
sentially complete when sea levels rose to about 15 m below present level in Holocene time
(bottom). The islands as they exist today are shown by the dashed lines. (Data taken from

Heatwole and MacKenzie, 1967.)

62 MISCELLANEOUS PUBLICATION MUSEUM OF NATURAL HISTORY

eastern shelf and separating all the Vir-
gin Islands from Puerto Rico and one
from another (Fig. 26). The configura-
tions of the islands were roughly the
same as today because of the steepness
of the slopes, but their areas were larger,
especially that of Anegada. For the past
6,000 years sea levels have fluctuated
slightly above or below present level
(Heatwole and MacKenzie, 1967).

Because the islands of the Puerto
Rican Bank were not separated until
8,000 years ago, it is not surprising that
none of the smaller islands has more than
two endemic species (less than 10% of
their herpetofauna). On the other hand,
Mona Island and St. Croix have a high
number of endemics (50% and 66%, re-
spectively) because these islands have
been separated from Puerto Rico for a
long time, if, in fact, they were ever con-
nected (Heatwole and MacKenzie, 1967).
Puerto Rico also has a larger number of
endemic species (60%), but the island is
much larger than any of the others on
the bank and the habitats are much more
diverse. Those species that persist on
the Virgin Islands do so in an environ-
ment of low relief and rainfall compara-
ble to their habitat on Puerto Rico, now
and during the late Pleistocene. Pre-
dictably, they include Eleutherodactylus
cochranae, E. antillensis, Ameiva exsul,
Anolis cristatellus, the grass/bush anoles
A. pulchellus and A. stratulus, Sphaero-
dactylus macrolepis, Typhlops richardi,
and both species of colubrid snakes. Cy-
clura pinguis persists as a relict on but
one island, Anegada. It is unlikely that
the more mesic-adapted species such as
Diploglossus pleei, the rain forest anoles
like A. occultus, A. krugi, and most of
the Eleutherodactylus were ever well es-
tablished on these “islands” in the Pleis-
tocene.

Some amphibians and reptiles have
disappeared from the bank islands in
modern times, for example, Anolis roose-
velti, the giant anole known only by a
few specimens from Culebra Island off
the east coast of Puerto Rico, has not

been seen since 1932 (Schwartz and
Thomas, 1975). Anolis roosevelti was
absent from all of the fossil deposits on
Puerto Rico and during the late Pleisto-
cene was probably excluded from the
west end of the bank by A. cuvieri.
Whether or not either of these lizards
was ever present on any of the other
bank islands is unknown. Peltophryne
lemur is another example of a species
whose range has been constricted. The
toad was last collected on Virgin Gorda
in 1931 and now may be extinct on that
island.

Insights into earlier events in Puerto
Rico’s zoogeographic history are seen in
the island’s geographically central posi-
tion in the West Indian archipelago. The
Puerto Rican herpetofauna is strikingly
Greater Antillean, in spite of the fact
that the island is no closer geographically
to the northern islands than to the Lesser
Antilles to the south. For example,
nearly all Puerto Rican Eleutherodacty-
lus are members of the auriculatus group
whose other Antillean members are
found mostly on Hispaniola and Cuba
(Schwartz, 1969). Over one hundred
species of Eleutherodactylus occur in the
Greater Antilles compared with five in
the Lesser Antilles. Peltophryne lemur
is a southern representative of the Cu-
ban-Hispaniolan complex of endemic
bufonids, which does not reach the
Lesser Antilles. West Indian Chrysemys
are restricted to the Greater Antilles ex-
cept for introductions on Guadelope.
The macrolepis complex of Puerto Rican
Sphaerodactylus probably resulted from
a single invasion from the north by the
notatus group (Schwartz, 1978). Puerto
Rican Anolis represent a radiation of the
Alpha Section that originally invaded
Hispaniola and subsequently dispersed
and differentiated as two groups—the
cristatellus and bimaculatus series. In-
terestingly, Williams (1976) accounted
for the entire diversification of West In-
dian Anolis by assuming only three main-
land invasions. Cyclura and Leiocepha-
lus are also Greater Antillean genera.

PLEISTOCENE PUERTO RICAN HERPETOFAUNAS 63

Most likely, Ameiva exsul is allied with
A. chrysolaema on Hispaniola rather
than with any northern Lesser Antillean
species, such as A. pleei (Baskin and
Williams, 1966). Diploglossus pleei
probably is a derivative from an an-
cestral species resembling D. warreni on
Hispaniola. The West Indian anguid
lizard fauna is exclusively Greater An-
tillean, except for Diploglossus monti-
serrati on Montserrat Island. Amphis-
baena is the only genus with its greatest
Antillean diversity on Puerto Rico; five
species occur in the Puerto Rican Bank,
Hispaniola has four, and Cuba has one.
The genus is absent from the Lesser An-
tilles. Epicrates also is restricted to the
Greater Antilles. The systematics of
West Indian xenodontine snakes is not
resolved, but the majority of species are
Greater Antillean.

There are two Puerto Rican genera
whose affinities may be in the Lesser
Antilles. Leptodactylus reaches the
Greater Antilles only as far as the south-
east corner of Hispaniola. Phyllodacty-
lus has been collected in Hispaniola, but
otherwise it is restricted to the Puerto
Rican Bank (P. wirshingi) and Barbados
(P. pulcher). Another lizard, Mabuya
mabouya, is an Old World derivative
now widespread in the Neotropics (a
number of species may be masquerading
as Mabuya mabouya) including the West
Indies. Only one other species of skink,
M. lineolata, occurs in the islands (His-
paniola).

Clearly, the Puerto Rican herpeto-
fauna is an assemblage of taxa on the
periphery of a Greater Antillean radia-
tion, and most of Puerto Rico's species
were derived from Hispaniola. The great
diversity of amphibians and reptiles on
Hispaniola (173 species) and the island’s
proximity to Puerto Rico has overshad-
owed chance invasion through the Lesser
Antilles. In addition, the distribution of
several fossil and living species indicates
that the Greater Antillean herpetofauna
extended farther south in “sub-Recent”
times. There are fossils of Leiocephalus,

Eleutherodactylus, and possibly Cyclura
on Barbuda (Etheridge, 1964a; Auffen-
berg, 1959; Lynch, 1966), and modern
representatives of Diploglossus on Mont-
serrat and Leiocephalus on Martinique.
Further documentation of this pattern
must await additional paleontological
work in the Lesser Antilles.

The antiquity of the Greater Antil-
lean herpetofauna is also demonstrated
by paleontological investigations of the
North American Tertiary. Leiocephalus
is known from the early Miocene of
Florida (Estes, 1963) and the middle
Miocene of Nebraska (Carl Wellstead,
pers. comm.), and a diploglossine lizard
was found recently in early Eocene de-
posits of Wyoming (Jacques Gauthier,
pers. comm.). These finds are important
because the probability now exists for a
North American origin of at least some
of the Antillean lizards heretofore as-
sumed to be South or Central American
derivatives. Wherever the source, under-
standing the origins of the Antillean her-
petofauna will require knowledge of
fragmentation events as well as dispersal.
Given the possibility of arid regimes pre-
dominating at periodic intervals through-
out the Pleistocene, any modern biogeo-
graphic analysis must now take into
account the possibility of a multitude of
restructured and relictual distributions
throughout the last half million years
alone, which substantially increases the
zoogeographic complexity. Continued in-
vestigation of the Antillean fossil record
is our best tool for understanding the
zoogeographic consequences of Pleisto-
cene climates.

SUMMARY
AND CONCLUSIONS

The known fossil record of West In-
dian amphibians and reptiles is no older
than late Pleistocene, and the few exist-
ing studies deal almost exclusively with
lizards from the Greater Antilles. This
study analyzes the first fossil herpeto-
faunas known from Puerto Rico, and the

64 MISCELLANEOUS PUBLICATION MUSEUM OF NATURAL HISTORY

most extensive yet reported from the
West Indies. Thousands of fossil bones
of amphibians and reptiles were col-
lected from 14 caves on the island; one
cave in particular, Blackbone 1, pro-
duced a unique abundance of species of
lizards, frogs, and snakes. In the total
fossil fauna are 21 species representing
16 genera and 12 families, many of which
are improbable finds for cave deposits.
The predatory role of the owl, Tyto
cavatica, that deposited the fossil pellets
biased the sample by favoring animals
of certain size and habits. The material
from Blackbone | includes fossorial spe-
cies, for example typhlopid snakes and
an amphisbaenid, as well as other small,
secretive species such as leaf-litter gek-
kos (Sphaerodactylus) and canopy-dwell-
ing anoles. Secondary ingestion may ac-
count for some of these, reflecting, in
part, the trophic structure of the commu-
nity. Apparently Tyto cavatica was an
efficient and opportunistic hunter. Also
among the remains are two previously
unknown species of the iguanid lizard
genus Leiocephalus, and fossils of the
previously enigmatic rock iguana Cy-
clura portoricensis (Barbour), which is
placed in the synonymy of C. pinguis
living today on Anegada Island.

The species composition of the fossil
faunas indicates that the environment of
the cave region was a xeric, savanna-like

habitat quite in contrast to the mesic
conditions that prevail there now. Cor-
roborative evidence from deep-sea drill-
ing projects in the Caribbean show that
arid and cool climates were present in
the West Indies 15,000-20,000 years ago,
the approximate time of fossil deposi-
tion. An environmental shift to more
mesic conditions subsequent to 13,000
years ago restructured the herpetofaunal
community and may have brought about
the extinctions of Leiocephalus and Cy-
clura on the island and restricted the
xerophiles to coastal habitat.

The herpetofauna of Puerto Rico is
Greater Antillean in its affinities and de-
rived largely from Hispaniola. As a
whole, the amphibians and reptiles of
the Greater Antilles are relictual forms,
and the Puerto Rican taxa are now at
the periphery of a Greater Antillean dis-
tribution that extended well into the
Lesser Antilles during the Pleistocene.

Future paleontological work is
strongly recommended for the West In-
dies. In cave faunas, particular atten-
tion should be directed towards potential
stratigraphic sequences of species and
continued efforts at confident dating of
the material. Pre-Quaternary fossils and
sediments should be sought in an effort
to extend our knowledge of West Indian
vertebrates further into the past.

PLEISTOCENE PUERTO RICAN HERPETOFAUNAS 65

LITERATURE CITED

AnTuony, H. E. 1916. Preliminary report on
fossil mammals from Porto Rico, with de-
scriptions of a new genus of ground sloth
and two new genera of hystricomorph ro-
dents. Ann. New York Acad. Sci., 27:
193-203.

Antnony, H. E. 1918. The indigenous land
mammals of Porto Rico, living and extinct.
Mem. Amer. Mus. Nat. Hist., 2(2):331—435.

Antuony, H. E. 1925-1926. Mammals of
Porto Rico, living and extinct. Sci. Survey
Porto Rico and Virgin Islands, New York
Acad. Sci., 9:1—238.

AUFFENBERG, W. 1959. A small fossil herpeto-
fauna from Barbuda, Leeward Islands, with
the description of a new species of Hyla.
Quart. Jour. Florida Acad. Sci., 21(3):248—-
254,

AUFFENBERG, W. 1967. Notes on West Indian
tortoises. Herpetologica, 23(1):34-44.

AUFFENBERG, W. 1974. Checklist of fossil land
tortoises (Testudinidae), Bull. Florida State
Mus., 18(3):121-251.

Avery, D. F., and W. W. TANNER. 1971. Evo-
lution of the iguanine lizards (Sauria, Igua-
nidae) as determined by osteological and
myological characters. Brigham Young Sci.
Bull., XII(3):1—79.

Barsour, T. 1919. A new rock iguana from
Porto Rico. Proc. Biol. Soc. Wash., 32:
145-148.

Barsour, T., and A. F. Carr. 1940. Antillean
terrapins. Mem. Mus. Comp. Zool., 54(5):
381-415.

BAskin, J. N., and E. E. WitiiaMs. 1966. The
Lesser Antillean Ameiva. Stud. Fauna
Curacao Carib. Islands, 89:143-176.

Bonatti, E., and S. Gartner. 1973. Carib-
bean climate during Pleistocene ice ages.
Nature, 244:563—565.

BouLencer, G. A. 1890. On the distinctive
cranial characters of the iguanoid lizards
allied to Iguana. Ann. Mag. Nat. Hist.,
6:412-414.

Carey, W. M. 1975. The rock iguana, Cyclura
pinguis, on Anegada, British Virgin Islands,
with notes on Cyclura ricordi and Cyclura
cornuta of Hispaniola. Bull. Florida State
Mus., 19(4):189-233.

CuoaTe, T., and E. Birney. 1968. Sub-recent
Insectivora and Chiroptera from Puerto
Rico, with the description of a new bat of
the genus Stenoderma. Jour. Mammalogy,
49(3):400-412.

Cocuran, D. M. 1941. The herpetology of
Hispaniola. Bull. U.S. Natl. Mus. 1977:
1-398.

CRAIGHEAD, J. J., and F. C. CraicHeap. 1969.
Hawks, owls and wildlife. Dover Publ. Inc.,
N.Y. xix + 443 p.

Donn, W. L., W. R. FAarraAnp, and M. Ewine.

1962. Pleistocene ice volumes and sea-level
lowering. Jour. Geol. 70:206-214,

Dunn, E. R. 1932. The colubrid snakes of the
Greater Antilles. Copeia, 1932:89-92.
Dunn, E. R. 1935. Notes on American
Mabuyas. Proc. Philadelphia Acad. Nat.

Sci., 87:533-557.

EmiiAnt, C. 1966. Paleotemperature analysis
of Caribbean cores P6304 and P6304-9 and
a generalized temperature curve for the past
425,000 years. Jour. Geol., 74(2):109-126.

EMILIANI, C., and E. Rona. 1969. Caribbean
cores P6304-8 and P6304-9: New analysis
of absolute chronology. A reply. Science,
166:1551—-1552.

Estes, R. 1963. Early Miocene salamanders
and lizards from Florida. Quart. Jour.
Florida Acad. Sci., 26(3):234-256.

Estes, R. 1964. Fossil vertebrates from the
late Cretaceous Lance Formation eastern
Wyoming. Univ. California Press, 180 pp.

ETHERIDGE, R. 1959. The relationships of the
anoles (Reptilia: Sauria: Iguanidae). An in-
terpretation based on skeletal morphology.
Unpubl. Ph.D. thesis Univ. Michigan, xiii
+ 236 pp.

ETHERIDGE, R. 1964a. Late Pleistocene lizards
from Barbuda, British West Indies. Bull.
Florida State Mus., 9(2):43-75.

ETHERIDGE, R. 1964b. The skeletal morphology
and systematic relationships of sceloporine
lizards. Copeia, 1964:610-631.

Eruerince, R. 1965. Fossil lizards from the
Dominican Republic Quart. Jour. Florida
Acad. Sci., 28(1):83-105.

Erueriwce, R. 1966a. An extinct lizard of the
genus Leiocephalus from Jamaica. Quart.
Jour. Florida Acad. Sci., 29(1):47—-59.

ErHeripcE, R. 1966b. The systematic relation-
ships of West Indian and South American
lizards referred to the iguanid lizard genus
Leiocephalus. Copeia, 1966:79-91.

Eruerince, R. 1966c. Pleistocene lizards from
New Providence. Quart. Jour. Florida Acad.
Sci., 28(4):349-358.

Ewe, J. J., and J. L. Wuirmore. 1973. The
ecological life zones of Puerto Rico and the
U.S. Virgin Islands. Forest Serv. Research
Paper ITF-8, USDA, 72 pp. + map.

FisHer, D. L., and W. W. Tanner. 1970.
Osteological and myological comparisons of
the head and thorax regions of Cnemidopho-
rus tigris septentrionalis Burger and Ameiva
undulata parva Barbour and Nobel. Brig-
ham Young Sci. Bull. Ser. XI(1):1-41.

Firzincer, L. 1843. Systema Reptilium. 106
p. Vindobonae.

Gans, C., and A. A. ALEXANDER. 1962. Studies
on amphisbaenids (Amphisbaenia, Reptilia).
On the amphisbaenids of the Antilles. Bull.
Mus. Comp. Zool., 128(1):65—-158.

66 MISCELLANEOUS PUBLICATION MUSEUM OF NATURAL HISTORY

Geuipacu, F. R. 1965. Amphibians and rep-
tiles from the Pliocene and Pleistocene of
North America: A chronological summary
and selected bibliography. Texas Jour. Sci.,
17(1):56-70.

Gorman, G. C., and R. Harwoop. 1977. Notes
on population density, vagility and activity
patterns of the Puerto Rican grass lizard,
Anolis pulchellus (Reptilia, Lacertilia, Igua-
nidae). Jour. Herp., 11(3):363-368.

Gorman, G. C., and S. Hitman. 1977. Phys-
iological basis for climatic niche partition-
ing in two species of Puerto Rican Anolis
(Reptilia, Lacertilia, Iguanidae). Jour. Herp.,
11(3):337-340.

Grant, C. 1932. Bufo lemur, a rare Puerto
Rican toad. Jour. Dept. Agric. Puerto Rico,
XVI(1):41—42.

Grant, C. 1940. The reptiles in Lynn and
Grant: The herpetology of Jamaica. Bull.
Inst. Jamaica Sci. Ser., 1:1—-65.

Greer, A. FE. 1970. A subfamilial classification
of scincid lizards. Bull. Mus. Comp. Zool.,
139(3):151-183.

Harrer, J. 1974. Avian speciation in tropical
South America. Publs. Nuttall Omitho-
logical Club, 14:1-390.

Harrer, J. 1979. Quaternary biogeography of
tropical lowland South America, pp. 107—
140 in W. E. Duellman (ed.), The South
American Herpetofauna: Its Origin, Evo-
lution and Dispersal. Mus. Nat. Hist. Univ.
Kansas Monograph No. 7, 485 pp.

Heatwo te, H., and F. Macxenzir. 1967. Her-
petogeography of Puerto Rico. IV. Paleo-
geography, faunal similarity and endemism.
Evolution, 21(3):429-438.

Heatwote, H., and F. Torres. 1967. Distri-
bution and geographic variation of the
Ameivas of Puerto Rico and the Virgin Is-
lands. Studies Fauna Curacao Carib. Is-
lands, 24:63-111.

Hecut, M. 1951. Fossil lizards of the West
Indian genus Aristelliger (Gekkonidae).
Amer. Mus. Nov., (1538):1-33.

Heyer, W. R. 1969a. Biosystematic studies on
the frog genus Leptodactylus. Unpub. Ph.D.
thesis, Univ. Southern California. 234 pp.

Heyer, W. R. 1969b. Studies on the genus
Leptodactylus (Amphibia, Leptodactylidae )
Ill. A redefinition of the genus Leptodacty-
lus and a description of a new genus of
leptodactylid frogs. Contrib. Sci. Los An-
geles Co. Mus., (155):1-14.

Heyer, W. R. 1970. Studies on frogs of the
genus Leptodactylus (Amphibia, Leptodac-
tylidae). VI. Biosystematics of the me-
lanonotus group. Contrib. Sci. Los Angeles
Co. Mus., (191):1-48.

Houtman, J. A. 1976. Paleoclimatic implication
of “ecologically incompatible” herpetological
species (late Pleistocene: southeastern United
States). Herpetologica, 32(3):290—-295.

Hooyer, D. A. 1963. Geochelone from the
Pleistocene of Curacao, Netherlands Antil-
les. Copeia, 1963:579-580.

Howarp, R. A. 1973. The vegetation of the
Antilles, pp. 1-38 in A. Graham (Ed.), Vege-
tation and Vegetational History of Northern
Latin America. Elsevier Scientific Publ. Co.,
Amsterdam xiii + 393 pp.

Koopman, K. F., and R. Rurar. 1955. Cave-
fossil vertebrates from Camagiiey Cuba.
Breviora, (46):1-8.

List, J. C. 1966. Comparative osteology of
the snake families Typhlopidae and Lepto-
typhlopidae. Illinois Biol. Mono., (36):
1-112.

Lyncu, J. D. 1966. The status of the tree
frog Hyla barbudensis Auffenberg from
Barbuda, British West Indies. Copeia, 1966:
524-530.

Lyncu, J. D. 1971. Evolutionary relationships
osteology, and zoogeography of Leptodac-
tylid frogs. Misc. Pub. Univ. Kansas Mus.
Nat. Hist., (53):1—-238.

Mactio, V. J. 1970. West Indian xenodontine
colubrid snakes: Their probable origin,
phylogeny, and zoogeography. Bull. Mus.
Comp. Zool., 141(1):1-53.

McDowe tt, S. B. 1975. A catalogue of the
snakes of New Guinea and the Solomons,
with special reference to those in the Bernice
P. Bishop Museum. Part II. Anilioidea and
Pythoninae. Jour. Herp., 9(1):1-80.

McDoweE tt, S. B., and C. M. Bocsrr. 1954.
The systematic position of Lanthanotus and
the affinities of the anguinomorphan lizards.
Bull. Amer. Mus. Nat. Hist., 105:1—142.

MeszogeLy, C. A. M. 1970. North American
fossil anguid lizards. Bull. Mus. Comp.
Zool., 139(2):87-149.

MittEer, G. S. 1918. Mammals and reptiles
collected by Theodore de Booy in the Virgin
Islands. Proc. U.S. Natl. Mus., 54:507-511.

Monroe, W. H. 1976. The karst landforms of
Puerto Rico. U.S. Geol. Survey Prof. papers,
(899):1-69.

Moran, J. M. 1975. Return of the ice age
and drought in peninsular Florida? Geology,
3:695-696.

Morean, G. 1977. Late Pleistocene vertebrates
from the Cayman Islands, British West In-
dies. Unpub. Master’s thesis, Univ. Florida,
260 pp.

OrLricH, T. M. 1956. The anatomy of the
head of Ctenosaura pectinata (Iguanidae).
Mise. Publ. Mus. Zool. Univ. Michigan,
(94):1-122.

Outson, S. L. 1978. A paleontological per-
spective of West Indian birds and mammals,
pp. 99-117 in F. B. Gill (ed.), Zoogeography
in the Caribbean. Acad. Nat. Sci. Phila-
delphia, Special Publ., (13):1-128.

PrEGILL, G. 1977. Axial myology of the racer
Coluber constrictor with emphasis on the

PLEISTOCENE PUERTO RICAN HERPETOFAUNAS 67

neck region. Trans. San Diego Soc. Nat.
Hist., 18(11):185-206.

Precitit, G. 1981. Cranial morphology and
the evolution of West Indian toads (Salientia,
Bufonidae): Resurrection of the genus Pel-
tophryne Fitzinger. Copeia, 1981:273—-285.

Ranp, A. S. 1964. Ecological distribution in
anoline lizards of Puerto Rico. Ecology,
45(4):745-752.

Ranp, A. S., and E. E. Wimuiams. 1969. The
anoles of La Palma: Aspects of their eco-
logical relationships. Breviora, (327):1-19.

Ray, C. E. 1964. A small assemblage of verte-
brate fossils from Spring Bay, Barbados.
Jour. Barbados Mus. Hist. Soc., 31(1):11—-22.

Rupa, R. 1959. Bufo gundlachi, a new spe-
cies of Cuban toad. Breviora, (105):1—-14.

Ruruven, A. G., and H. T. Gaicr. 1935. Ob-
servations on Typhlops from Puerto Rico
and some of the adjacent islands. Occ.
Papers Mus. Zool. Univ. Michigan, (302):
1-12,

Scumipt, K. P. 1928. Amphibians and land
reptiles of Porto Rico, with a list of those
reported from the Virgin Islands. Sci. Sur-
vey Porto Rico and Virgin Islands, New
York Acad. Sci., 10(1):1-160.

Scuwarrz, A. 1960. The large toads of Cuba.
Proc. Biol. Soc. Washington, 73:45—56.
Scuwartz, A. 1965. The Leiocephalus of
Hispaniola I. Leiocephalus melanochlorus

Cope. Jour. Ohio Herp. Soc., 5(2):39-48.

ScHwartz, A. 1966. Snakes of the genus
Alsophis in Puerto Rico and the Virgin Is-
lands. Stud. Fauna Curacao Carib. Islands,
23(90):177-227.

ScuHwartz, A. 1967. A review of the genus
Dromicus in Puerto Rico and the Virgin
Islands. Misc. Papers Mus. Biol. Univ.
Puerto Rico, Rio Piedras, (9):1—14.

Scuwartz, A. 1969. The antillean Eleuthero-
dactylus of the auriculatus group. Stud.
Fauna Curacao Carib. Islands, 30(114):
99-115.

Scuwartz, A. 1970. A new species of large
Diploglossus (Sauria: Anguidae) from His-
paniola. Proc. Biol. Soc. Washington, 73:
717-178.

Scuwartz, A. 1972. The native toads of His-
paniola. Jour. Herp., 6(3—4):217-231.
Scuwartz, A. 1978. Some aspects of the her-
petogeography of the West Indies, pp. 31-
51 in F. B. Gill (ed.), Zoogeography in the
Caribbean. Acad. Nat. Sci. Philadelphia,

Special Publ., (13):1-128.

ScHwartz, A., and M. Carry. 1977. Sys-
tematics and evolution in the West Indian
iguanid genus Cyclura. Stud. Fauna
Curacao Carib. Islands, (173):15—97.

Scuwartz, A.. and R. THomas. 1975. A
check-list of West Indian amphibians and
reptiles. Carnegie Mus. Nat. Hist. Special
Publ., 1:1-216.

SHEPLAN, B. R., and A. Scuwartz. 1974.
Hispaniolan boas of the genus Epicrates
(Serpentes, Boidae) and their Antillean re-
lationships. Ann. Carnegie Mus., 45(5):
57-143.

Stmpson, B. B., and J. Harrer. 1978. Spe-
ciation patterns in the Amazonian forest
biota. Ann. Rev. Ecol. Syst., 9:497-518.

StraHM, M. H., and A. Scuwartz. 1977.
Osteoderms in the anguid lizard subfamily
Diploglossinae and their taxonomic im-
portance. Biotropica, 9(1):58-72.

Tuomas, R. 1965. New species of antillean
Eleutherodactylus. Quart. Jour. Florida
Acad. Sci., 28:275-391.

Tuomas, R. 1966. Additional notes on the
amphisbaenids of greater Puerto Rico.
Breviora, (249):1—23.

Tuomas, R., and A. ScHwartz. 1966. Sphae-
rodactylus (Gekkonidae) in the greater
Puerto Rican region. Bull. Florida State
Mus. Biol. Sci., 10(6):194—260.

Trues, L. 1973. Bones, Frogs and Evolution,
pp. 65-132 in J. L. Vial (ed.), Evolutionary
Biology of the Anurans, Contemporary Re-
search on Major Problems. Univ. Missouri
Press, Columbia.

Trues, L. 1977. Osteology and anuran sys-
tematics: Intrapopulational variation in
Hyla lanciformis. Syst. Zool., 26(2):165-184.

Trues, L., and M. J. TyLer. 1974. Systematics
and evolution of the Greater Antillean Hylid
frogs. Occ. papers Mus. Nat. Hist., Univ.
Kansas, (24):1—-60.

UnDERWoop, G. 1959. A new Jamaican galli-
wasp (Sauria, Anguidae). Breviora, (102):
1-13.

Verona, L. S. 1974. Catalogo de los mamif-
eros vivientes y extinguidos de las Antillas.
Havana: Academia De Ciencias De Cuba,
139 pp.

Watts, W. A. 1975. A late Quaternary record
of vegetation from Lake Annie, southcentral
Florida. Geology, 3:344-346.

Wetmore, A. 1920. Five new species of birds
from cave deposits in Porto Rico. Proc.
Biol. Soc. Washington, 33:77-82.

Wetmore, A. 1922. Bird remains from caves
of Porto Rico. Bull. Amer. Mus. Nat. Hist.,
46(4):297-333.

WituiaMs, E. E. 1950. Testudo cubensis and
the evolution of western hemisphere tor-
toises. Bull. Amer. Mus. Nat. Hist., 95(1):
1-36.

WituraMs, E. E. 1952. A new fossil tortoise
from Mona Island, West Indies and a ten-
tative arrangement of the tortoises of the
world. Bull. Amer. Mus. Nat. Hist., 99(9):
541-560.

WituiaMs, E. E. 1956. Pseudemys scripta
callirostris from Venezuela with a general
survey of the scripta series. Bull. Mus.
Comp. Zool., 115(5):145-160.

68 MISCELLANEOUS PUBLICATION MUSEUM OF NATURAL HISTORY

WituiaMs, E. E. 1976. West Indian anoles:
A taxonomic and evolutionary summary 1.
Introduction and a species list. Breviora,
(440):1-21.

WituiaMs, E. E., J. A. Rivero, and R. THoMas.
1965. A new anole (Sauria, Iguanidae) from
Puerto Rico. Breviora, (231):1—-18.

Wine, E., C. HorrMan, Jr., and C. Ray. 1968.
Vertebrate remains from Indian sites on
Antigua, West Indies. Caribbean Jour. Sci.,
8(3—4):123-139.

Woop, R. C. 1972. A fossil pelomedusid turtle
from Puerto Rico. Breviora, (392):1-13.

APPENDIX I: Comparative MATERIAL
EXAMINED. (Museum acronyms are given
on page 3.)

Frogs

Peltophryne empusa, AMNH 61409;
P. fluviatica, AMNH 87235; P. gundlachi,
AMNH 99764; P. guntheri, AMNH
54743; P. lemur, RT 4075; USNM 27148,
27150; P. longinasa, AMNH_ 101835; P.
peltocephala, USNM 28021, 029484,
90888, 167528; Eleutherodactylus antil-
lensis, KU 124244: E. cooki, MCZ 18653;
E. coqui, MCZ 43221; KU 79950; GKP
261, 305; E. karlschmidti, FSM 24200;
MCZ 18711; E. locustus, KU 124249; E,
portoricensis, GKP 042, 254; MCZ 85212;
E. richmondi, MCZ 18922; E. wight-
manae, KU 124255; Leptodactylus al-
bilabris, GKP 302; MCZ 87339; L. wag-
neri, KU 125940.

Lizards

Amphisbaena fenestrata, KU 87856;
Anolis cooki, GKP 0301; A. cristatellus,
KU 49022-9, 49035, 49043-4; A. cuvieri,
MCZ 57898-9, 10960, 35967-71, R127119
(alcohol); A. evermanni, MCZ 132041;
A. equestris, KU 61391; USNM 167530;
A. gundlachi, GKP 0154; KU 79198; MCZ
131917; A. krugi, FSM 39764-5; MCZ
132087; A. occultus, MCZ 146683; A.
pulchellus, GKP 0290; FSM 29766-7;
MCZ 36073; A. ricordi, USNM 72632;
A. stratulus, GKP 0303; KU 49053-57;
Ameiva c. chrysolaema, REE 1459; A.
exsul, KU 045557; MCZ 131904; A. f.
festiva, KU 117468; A. lineolata, REE
1482; A. taeniura, REE 1456; A. undulata

parva, KU 117473; A. wetmorei, MCZ
36437; Basiliscus vittatus, USNM 71430;
Celestus barbouri, ASFS 16081; C. cos-
tatus, FSM 12308-1; C. curtissi, ASFS
V42502; C. crusculus, FSM 20935, 21743;
C. darlingtoni, ASFS V45037; C. hewardi,
ASFS V11534; Conolophis pallidus, REE
1382, 1447; C. subcristatus, USNM
165756; Ctenosaura quinquecarinata,
USNM 30564; C. sp., USNM no data;
Cyclura carinata, FSM 30206-8, 30407,
30425, 32674, 32676, 33291, 33293-4,
33265, 35122, 37024, 39770-1; USNM
88819; C. cornuta, FSM 33651, 34750;
REE 383, 1967; C. cychlura, FSM 19066;
C. figginsi, KU 174796; C. mattea, USNM
59358-9; C. macleayi, REE 228; FSM
21910; C. pinguis, ASFS V21995; C. por-
toricensis, MCZ 1008-13; Diploglossus
pleei, GKP 0268-70; MCZ 131510; D.
warreni, ASFS V42502; Dipsosaurus dor-
salis, USNM 29182; Enyalioides prae-
stabilis, USNM 7796; Enyaliosaurus
clarki, USNM 21452; Laemanctus ser-
ratus, USNM 82178; Leiocephalus aper-
tosulcus, MCZ 3404; L. barahonensis,
REE 1821; L. carinatus, FSM 14373,
18929, 23082, 21739, 21907, 39763; REE
1469, 1816, 1805-6, 1808; USNM 81709;
L. cubensis, MCZ 7281a; L. cuneus, FSM
8226, 8263-71, 8444, 8468-70; L. green-
wayi, REE 1814; L. inaguae, ASFS
10330, 10337, 10346; FSM 11525; L. loxo-
grammus, MCZ 38136; L. lunatus, REE
1815; L. macropus, REE 1819; USNM
59167; L. melanochloris, REE 1802; L.
personatus, REE 1803, 1811; L. pratensis,
ASFS V9844, V9851; L. psammodrom-
mus, FSM 33737-8, 34804-6, 25124-6; L.
punctatus, ASFS V8732, V8736, V8747;
L. raviceps, MCZ 11376; L. schreibersi,
REE 1810; FSM 12273, 12276; L. stic-
togaster, REE 1810; L. vinculum, REE
1812; Liolaemus chilensis, KU 161700;
L. fitzingeri, USNM 36933; L. kingi,
USNM 36918; L. lineomaculatus, USNM
36896; L. multiformis, KU 133796; L.
nigriceps, USNM 1642516; L. nitidus,
KU 161770; L. tenuis, USNM 5518; L.
weigmanni, USNM 70477; Mabuya ma-
bouya KU 113519; Morunasaurus annu-

PLEISTOCENE PUERTO RICAN HERPETOFAUNAS 69

laris, USNM 204240; Ophryoessoides iri-
descens, USNM 200912; Plica plica,
AMNH 61314; P. umbra, USNM 204266;
Phrynosoma cornutum, USNM _ uncat.,
Sauresia agasepsoides, ASFS V40784; S.
sepsoides, ASFS V40855; Sceloporus cy-
anogenys, USNM 47715; S. undulatus,
USNM 739, 29194; Sphaerodactylus mac-
rolepis, KU 49001-4; GKP 299; Stenocer-
cus varius, USNM 201321; Tropidurus
albemartensis, AMNH 77624; T. torqua-
tus, USNM 148772; Uracentron flaviceps,

USNM 201390; Uranoscodon supercili-
osa, USNM 202682; AMNH 61304; Uro-
saurus ornatus, USNM_ 16055; Wet-
morena haitiana, FSM 12314; ASFS
V20917.

Snakes

Alsophis portoricensis, RT 4829; KU
045669; Arrhyton exiguum, GKP 0274;
MCZ 5430; Epicrates cenchris, GKP 110;
E. inornatus, FSM 14434; MCZ 4684;
Typhlops richardi, GKP 0275.

70 MISCELLANEOUS PUBLICATION MUSEUM OF NATURAL HISTORY

AppEeNDIx II: Previousty REcoRDED FossiL OCCURRENCES OF AMPHIBIANS AND
REPTILES IN THE WeEsT INpIES (EXCLUSIVE OF ARCHAEOLOGICAL STUDIES). ALL ARE
LATE PLEISTOCENE TO SUB-RECENT. (* = Extinct species. )

Barbados

Testudinidae

(?) Geochelone sp.* Ray (1964)

Iguana iguana Ys

Barbuda

Leptodactylidae

Eleutherodactylus barbudensis* Auffenberg (1959), Lynch (1966)

(= Hyla barbudensis Auffenberg )
Gekkonidae

Thecodactylus rapicauda Auffenberg (1959), Etheridge (1964a)
Iguanidae

Anolis bimaculatus leachii Etheridge (1964a)

Anolis sp. (medium ) is

Anolis sp. (small) wi

Leiocephalus cuneus* a

Genus and species indet. My

(cf. Cyclura, Conolophis)

Teiidae

Ameiva griswoldi Auffenberg (1959)
Colubridae

Pseudoboa (cf. P. cloelia) ue

Cayman Islands

Leptodactylidae

Eleutherodactylus planirostris Morgan (1977)
Hylidae

Osteopilus septentrionalis Me
Gekkonidae

Aristelliger praesignis Mi
Iguanidae

Anolis conspersus a

Anolis sagrei Me

Cyclura nubila 4

Leiocephalus carinatus ue
Anguidae

Celestus crusculus He
Typhlopidae

Typhlops biminiensis i

Typhlops caymanensis AG
Boidae

Tropidophis caymanensis Ke
Colubridae

Alsophis cantherigerus Us
Crocodylidae

Crocodylus cf. acutus uy

PLEISTOCENE PUERTO RICAN HERPETOFAUNAS

Hylidae

Osteopilus septentrionalis
Testudinidae

Geochelone cubensis*
Gekkonidae

Tarentola americana
Iguanidae

Anolis lucius

Anolis equestris

Leiocephalus sp.
Colubridae

Alsophis cantherigerus

Testudinidae
Geochelone sp.*

Gekkonidae
Aristelliger lar
Iguanidae
Anolis chlorocyanus
Anolis cybotes
Anolis ricordi
Leiocephalus apertosulcus*
Leiocephalus personatus
Teiidae
Ameiva chrysolaema
Ameiva taeniura
Anguidae
Celestus costatus
Celestus stenurus

Gekkonidae
Aristelliger lar (= A. titan)
Aristelliger praesignis
Iguanidae
Leiocephalus jamaicensis*

Testudinidae
Geochelone monensis*

Testudinidae
Geochelone sp.*

Cuba

Koopman and Ruibal (1955)
Williams (1950), Auffenberg (1974)

Koopman and Ruibal (1955)

4}

4

Curacao

Hooijer (1963)

Dominican Republic

Etheridge (1965)

Jamaica

Hecht (1951)

Etheridge (1966a)
Mona Island
Williams (1952), Auffenberg (1974)
Navassa

Auffenberg (1967, 1974)

71

72 MISCELLANEOUS PUBLICATION MUSEUM OF NATURAL HISTORY

New Providence

Gekkonidae
Tarentola americana Etheridge (1966c)
Iguanidae
Anolis carolinensis
Anolis distichus ee
Anolis sagrei a
Cyclura sp.
Leiocephalus carinatus
Teiidae
Ameiva auberi
Testudinidae
Geochelone sp.* Auffenberg (1967)

NA
yt
Wy

Wt

Puerto Rico

Iguanidae
Cyclura pinguis (= C. portoricensis ) Barbour (1919)

St. Thomas
Iguanidae
Cyclura pinguis (=C. mattea) Miller (1918)
Sombrero Island
Testudinidae
Geochelone sombrerensis* Auffenberg (1967, 1974)

52.
53.

55.
57.

59.

61.
62.
65.

66.
67.
68.
69.

70.

RECENT MISCELLANEOUS PUBLICATIONS
UNIVERSITY OF KANSAS MUSEUM OF NATURAL HISTORY

Reproductive cycles in lizards and snakes. By Henry S. Fitch. Pp. 1-247, 16 fig-
ures in text. June 19, 1970. Paper bound.

Evolutionary relationships, osteology, and zoogeography of leptodactyloid frogs.
By John D. Lynch. Pp. 1-238, 131 figures in text. June 30, 1971. Paper bound.

Middle American lizards of the genus Ameiva (Teiidae) with emphasis on geo-
graphic variation. By Arthur C. Echternacht. Pp. 1-86, 28 figures in text. De-
cember 14, 1971. Paper bound.

A systematic review of the Teiid lizards, genus Bachia, with remarks on Heter-
odactylus and Anotosaura. By James R. Dixon. Pp. 1-47, 15 figures in text.
February 2, 1973. Paper bound.

Systematics and evolution of the Andean lizard genus Pholidobolus (Sauria:
Teiidae). By Richard R. Montanucci. Pp. 1-52, 8 figures in text. May 14, 1973.
Paper bound.

Reproductive strategies in a tropical anuran community. By Martha L. Crump.
Pp. 1-68, 13 figures in text. November 15, 1974. Paper bound.

A demographic study of the ringneck snake (Diadophis punctatus) in Kansas. By
Henry S. Fitch. Pp. 1-53, 19 figures in text. April 3, 1975. Paper bound.

The biology of an equatorial herpetofauna in Amazonian Ecuador. By William
E. Duellman. Pp. 1-352, 198 figures in text. August 30, 1978. Paper bound.

Leptodactylid frogs of the genus Eleutherodactylus from the Andes of southern
Ecuador. By John D. Lynch. Pp. 1-62, 23 figures in text. February 28, 1979.
Paper bound.

An ecogeographic analysis of the herpetofauna of the Yucatan Peninsula. By
Julian C. Lee. Pp. 1-75, 27 plates, 22 figures in text. February 29, 1980. Paper
bound.

Internal oral features of larvae from eight anuran families: Functional, systematic,
evolutionary and ecological considerations. By Richard Wassersug. Pp. 1-146,
37 figures in text. June 24, 1980. Paper bound.

The Eleutherodactylus of the Amazonian slopes of the Ecuadorian Andes (Anura:
Leptodactylidae). By John D. Lynch and William E. Duellman. Pp. 1-86, 8 fig-
ures in text. August 29, 1980. Paper bound.

Sexual size differences in reptiles. By Henry S. Fitch. Pp. 1-72, 9 figures in text.
February 27, 1981. Paper bound.

m
aa)

UO _—aEr
Anise

SNI

LSN
IES

LSNI

: = = z 4 : g

— a + a _ [ad ato

4 < c = es =

= re =i es = a S

= a oO a — a . —_

2 e 2 lies z 2 5 g

83 iyuvug ieee BRARI ES _ SMITHSONIAN _INSTITUTION ZNOILALILSNI NVINOSHLIWS Ssazluvug
ue os - on z

=) = = LA > > sie =)

= 3) ore WY 9) E 0 Le

- = PB Yi = iB ts -

Z = = 7h) se

INSTITUTION ee NVINOSHLIWS zt! BRARI Am SIO.N cl

L = = = at we = <

> = » a ge Sh Kh! Jy 2 =] Zz

>) aoe $ fg re) 25 GY ‘Ny oO SE Oo

2) ” SA + w 7p) Y 22) 122) (Ty >

2 Gy = a mae 2 é

A ea eles =: * ie oe :

N s * a a

aod 1yvVYsiIT_ LIBRARI ES SMITHSONIAN INSTITUTION NOILOLILSNI_ NVINOSHLIWS Sil yVUE

LJ = (rey , 4 = uw 2 Reais a

= =. of fy > = = :

c = L377 ZY = < =I <

¢ S oe S w = aw

B = a j = oO = fon)

= Oo = fe: (e) ~ (o) —

J = a) z =i z ot)

_INSTITUTION NOILNLILSNI_NVINOSHLINS S3IYVYgIT LIBRARIES SMITHSONIAN INSTITUTIC
Zz nS S - z ms

zs oO as = Oo ig =

"] = o = o = w

0 = a We 5 70 = _@

> = > \ NWS . & > = " >

E = W"s = e. a

a) m VAN é m m

” = 2) a css wn cs a

iS IYVUgIq ul I B RARI ES SMITHSONIAN INSTITUTION NOILNLILSNI _ NVINOSHLIWS |

: y ae = au ES = wy, = =

; Zz ‘rs S \ 3 GH =z =

Se \ tuys.

CBOs S\ b 2 Ne a WI, 2 b

E NY2 E AY 2.7” = 2

> = : BN. = = NN > = >

“INSTITUTION NOILALILSNI_ NVINOSHLINS S3IYVYGIT LIBRARIES SMITHSONIAN INSTITUTIO

: : 2 2 2 Z

z = & = 2 Lo

« = a S os Sj

M [aay = _ oO. =

) = ° = fe) = oO

; a = a) 2 = 2

Saiu’VvYsIT LIBRARIES INSTITUTION NOILNLILSNI NVINOSHLINS SJIYVus

4 LE = = - oe Zz

ow 2 oy = oo =

2 5 2 5 7 5

> Rs = = s

= a = a = B

r an z a Zz ay =

INSTITUTION NOILALILSNI S3JIMUVYUGIT LIBRARIES SMITHSONIAN _

CS 2 z Ed Lee z

é < pes e = =<

j = z 4 yal Jy 2 WS =) z

) <= 2) x= Wg D9 WKS oO %

7 g = 2 Gy = SN 2 =

3 > = > = > =

? Zs ” os 7) a C3 ”

S3IYVYSIT LIBRARIES

4
g!

SMITHSONIAN INSTITUTION NOILALILSNI NVINOSHLINS S3IuVvU
) S Ww S no S Ree 2)
: o ies o p= Oo Wr ¥
= = =e «4 = ao KG FS
= < = < = WS =
2 S ee = oe S SN ee
; rs} a 3 oy 5 ie
3 eZ a s 2 a) z= SY) hel
INSTITUTION | NOILALILSNI_NVINOSHIIWS ~S31YV¥S1T LIBRAR| ES_ SMITHSONIAN” INSTITUTIO
~ D
} acy ° Slt Gy oo tm 5 39'NG ° YY, D2 Kaan
> Kas, 2 Ka? XK. F LMR = Gar f GF = Kam

NOILALI4

-IBRARIES SMITHSONIAN

INSTITUTION

NOILALILSNI NVINOSHLINS S31uvyudIt

SMITHSONIAN

SMITHSONIAN

RARIES

NVI NOSHLINS

8iTLlis

SMITHSONIAN

NVINOSHLINS S31yuvYy

NVINOSHLIWS

NOILNLILSNI

SMITHSONIAN

INSTITUTION

SMITHSONIAN

SMITHSONIAN

*

eH (ES

NVINOSHLIWS
~

agi1_ ci

a = x = =
4 a <x 5 <
ra Ss oc = a
mo = oO - _ 0.8)
ce oO = (e) -
St z —) 2 -
INSTITUTION NOILNLILSNI NVINOSHLINS S3IYUVYUGIT LIBRARIES
LE = iS on z &
Oo ° o ° ow
x 5 a = a
> (= > = >
~ rE = = 2
Ww)
Oo z o Zz D
LIBRARIES INSTITUTION NOILNLILSNI
Zz = Oe a SRS <
a ay fy = AS =] z WO 5S
SGI ZEN 8 ZN 3
z “ay fe YS z E Zan
= Ww « pee wo .. =
—!NSTITUTION NOILNLILSNI_NVINOSHLIWS S3I1YVYGIT LIBRARIES
@ ul ae ee 2
= z 2 WN F = |
= < 2 ES = =
: : 2h Ve :
(e) —_ (e) = Oo :
z 4 Fa ne ay Fe
S3IYVYIT LIBRARIES SMITHSONIAN INSTITUTION NOILNLILSNI
S E S ys z |
5 20 = YY © = |
=) a 5 $Y, > 5
= cas Le L4L'2 E
re 2 e, YT fs = =
Zz m aie Zz
INSTITUTION NOILNLILSNI_ NVINOSHLIWS $3 IuVvugit wt IBRARIES
z Ht = A yy < = Z . :
a SS = iy, = = 5
8 Nat 9: z 8 NA |
a Se ANG oO Vf lye. bs oO a rN ¢
Ee ASS es a ee z = :
= Ww =: >=" Ss : 5
w oe = = = wo wat =
oat evualy LIBRARIES SMITHSONIAN INSTITUTION NOILNLILSNI_
> = ” -
é & bs a ul
a = w pea a =
bce e s e =
oO = a ca a =
~ fe) = ro) = c
va | =a a) 2 bh a
INSTITUTION NOILALILSNI NVINOSHLINS S31NYVUSIT. LIBRARIES
is z = ie z T z2
“w 2 w S w S
ay = 2 Fr Pe] E
> = >: = > :
2 = 2 = = r
a ee i 2 i Z
S s3ziuvudit SMITHSONIAN INSTITUTION NOILMLILSN!
wn Zz 7) = - 27) 2
= sot = < = aa
=e tippy Zz = a Se was
2 Cy = = ie z “YY =
> ¢ = a = > g =
z (2) . FZ ” E Fe Uv
_ INSTITUTION NOILALILSNI_NVINOSHLIWS $3 IYVYEIT_ LIBRARIES
= Ww 2 Rec Ww = u
g : GNX F g :
S ee S SQ = S :
) 2 re vp tanen re) =
z 2 z= = —j 2 pe
S3IYVYGIT LIBRARIES SMITHSONIAN INSTITUTION NOILQLILSNI
S e Ss Z = aN S woTITUY =
os Da N, Y Uy, 39 NO Y XN Y,
E 2 EM = GF, - GEM = GAM °