4 1

HARVARD UNIVERSITY

LIBRARY

OF THE

Museum of Comparative Zoology

,, ^he Library
""^eu^ Of Comparative Zoology
Harvard University

UNIVERSITY OF KANSAS
MUSEUM OF NATURAL HISTORY

MISCELLANEOUS

PUBLICATION

No. 56

Systematics of the Chiropteran
Famfly Mormoopidae

By

James Dale Smith

INIVERSITY OF KANSAS
LAWRENCE 1972

March 10, 1972

II

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 PubHcations, Museum of
Natural History, began with number 1 in 1946. Longer research papers are pub-
Hshed in that series. Monographs of the Museum of Natural History were initiated
in 1970.

Institutional libraries interested in exchanging publications may obtain the Occa-
sional Papers and Miscellaneotis PubHcations by addressing the Exchange Librarian,
University of Kansas Library, Lawrence, Kansas 66044. Individuals may purchase
separate nimibers of all series. Prices may be obtained upon request addressed to
PubHcations Secretary, Museum of Natural History, University of Kansas, Law-
rence, Kansas 66044.

University of Kansas
Museum of Natural History

Miscellaneous Publication No. 56
March 10, 1972

Systematics of the Chiropteran
Family Mormoopidae

By

James Dale Smith

Department of Biology
California State College
Fullerton, California 92631

University of Kansas

Lawrence

1972

University of Kansas Publications, Museum of Natural History

Editor: William E. Duellman

Miscellaneous Publication No. 56

pp. 1-132; 40 figures

Published March 10, 1972

Museum of Natural History

University of Kansas

Lawrence, Kansas 66044

U.S.A.

Printed by

University of Kansas Printing Service

Lawrence, Kansas

CONTENTS

INTRODUCTION 4

Materials and Acknowledgments 4

Methods 5

Historical Review , _ 10

ANALYSIS OF CHARACTERS 15

PHYLOGENETIC RELATIONSHIPS OF THE MORMOOPIDAE 38

SYSTEMATIC ACCOUNTS 41

Family Mormoopidae Saussure 41

Generic Analysis 43

Patterns of Non-Geographic Variation 48

Key to the Genera and Subgenera 55

Genus Pteronotus Gray 55

Subgenus Phyllodia Gray - 57

Pteronotus parneUii (Gray) 57

Subgenus Chilomjcteris Gray 78

Pteronotus macleayii (Gray) -- 79

Pteronotus fuUginosus ( Gray ) 83

Pteronotus personatus (Wagner) — 87

Subgenus Pteronotus Gray 94

Pteronotus davyi Gray "o

Pteronotus suapurensis ( ]. A. Allen) 104

Genus Mormoops Leach 10'

Mormoops bkiinviUii Leach 108

Mormoops me<Ji,aIophyUa Peters - — — -- 113

ZOOGEOGRAPHY AND SPECIATION OF MORMOOPIDS 120

SUMMARY - ^-^

RESUMEN - — - -- 1^

LITERATURE CITED -- - 1^7

INTRODUCTION

The mustached bats, naked-backed
bats, and ghost-faced bats of the New
World family Mormoopidae occur from
southern Texas through Mexico and Cen-
tral America into South America to south-
ern Brasil. Members of this family also
are found on the Greater Antilles, some
of the islands of the Lesser Antilles,
Trinidad, and in the Dutch West Indies.
The ecological associations of the in-
cluded species, although not yet clearly
defined, seem to be rather broad; the
bats range from humid tropical to semi-
arid and arid subtropical environments.

Since the discovery of mormoopid
bats in the early Nineteenth Century,
their taxonomic treatment has been
varied and, at times, controversial. Prior
to this study, mormoopids, all of which
are insectivorous, were generally con-
sidered as a distinct subfamily (Chilo-
nycterinae) of the family Phyllostomati-
dae. However, mormoopids lack a cu-
taneous nose-leaf like that possessed to
some extent by most phyllostomatids;
furthermore, the latter are principally
frugivorous. In recent years, mormo-
opids have been grouped in three genera
— Chilonycteris, Pteronotus, and Mormo-
ops — by some authorities, or in two —
Pteronotus and Mormoops — by others.

Mormoopid bats were first reviewed
critically by Rehn (1902, 1904a, and
1904b ) , whose studies were based on 188
specimens. Rehn recognized three gen-
era, eight species, and 14 geographic
races. I have had at my disposal more
than 5000 specimens; herein I recognize
two genera and eight species. Two spe-
cies are monotypic; six are polytypic and
are composed of 21 subspecies, three of
which are described as new.

The aims of this study were to collect
and correlate as much information as pos-
sible concerning the relationships of the
Mormoopidae, and to bring into focus
some of the problems facing mammalo-
gists with interest in the Neotropical bat
fauna. Specifically, the study involved:
a systematic revision, primarily con-

cerned with the familial and generic
levels; an interpretation of the phylo-
genetic relationships and evolution of
the recognized taxa; a study of varia-
tion, both geographic and non-geo-
graphic, within species; and a prelim-
inary appraisal of zoogeographic features
that have influenced the dispersal and
speciation of the group.

MATERIALS AND
ACKNOWLEDGMENTS

Most of the 5173 specimens examined
were conventionally prepared study
skins and skulls. Each specimen usually
was accompanied by a label bearing in-
formation concerning date and place of
capture, sex and frequently data re-
lating to reproductive condition, four
standard external measurements, and
name of collector. Additional material
examined included complete skeletons,
skulls unaccompanied by skins or re-
moved from specimens preserved in alco-
hol, and entire individuals in alcohol.

I am deeply indebted to the follow-
ing institutions and persons who put this
material at my disposal (abbreviations
preceding the names of institutions are
used throughout the text to identify the
source of specimens ) :

AMNH American Museum of Natural History
( Karl F. Koopman )

ANSP Academy of Natural Sciences of Phila-
delphia (R. R. Grant, Jr.)

ASC Albert Schwartz (private collection)

BMNH British Museum (Natural History) (J.
E. Hill)

CAS California Academy of Science ( Rob-
ert T. Orr)

ENCB Escuela Nacional de Ciencias Biologi-
cas, Mexico D. F. (Ticul Alvarez)

FMNH Field Museum of Natural History
( Joseph C. Moore and Philip Hersh-
kovitz )

IB Instituto de Biologia, Universidad Na-

cional Autonoma de Mexico, Mexico
( Bernardo Villa-R. )

IZT Instituto de Zoologia Tropical, Facul-

tad de Ciencias, Universidad Central
de Venezuela, Caracas (Juhani Ojasti
and Edgardo Mondolfi)

4

SMITH: SYSTEMATICS OF MORMOOPID BATS

KU The Museum of Natural History, The

University of Kansas (J. Knox Jones,

Jr.)

LACM Los Angeles County Museum ( Donald
Patten and Andrew Starrett)

MCZ Museiun of Comparative Zoology, Har-
vard University ( Barbara Lawrence )

MLS Museo de Historia Natural, La Salle,
Caracas, Venezuela ( Omar Linares )

MPEG Museum Paraense Emilio Goeldi, Be-
lem, Brasil (Fernando C. Novaes)

MSU The Museum, Michigan State Univer-
sity (Rollin H. Baker)

MVZ Museum of Vertebrate Zoology, Uni-
versity of California, Berkeley (Oliver
P. Pearson, W. Z. Lidicker, Jr., and
Seth B. Benson)

RMNH Rijksmusemn van Naturlijke Historic,
Leiden, Netherlands (A. M. Husson)

ROM Royal Ontario Museum (R. L. Peter-
son and J. R. Tamsitt)

TCWC Texas A and M University, Texas
Cooperative Wildlife Research Col-
lection (W. B. Davis and Dilford C.
Carter )

UMMZ University of Michigan, Museum of
Zoology (E. T. Hooper and W. H.
Burt )

USNM United States National Museum, in-
cluding the Biological Surveys Collec-
tion (C. O. Handley, Jr., and R. H.
Manville )

I thank the following agencies for
grants-in-aid of travel to the major North
American mammalogical collections : the
Committee on Systematics and Evolu-
tionary Biology (National Science Foun-
dation Program GB-4446X) at The Uni-
versity of Kansas (an 11-month trainee-
ship was also granted by this committee ) ;
the Committee on Grants-in-aid of Re-
search, the Society of the Sigma Xi; the
Watkins Museum of Natural History
Fund, The Museum of Natural History,
The University of Kansas; and the Re-
search Awards Committee, Kansas Acad-
emy of Science. Furthermore, I thank
the Universidad de Oriente, Cumana,
Venezuela, for a research grant to defray
the expenses of field investigation and a
vehicle during my 18-month stay in
Venezuela. Field work that resulted in
the collection of some of the material re-
ported herein was made possible by The
Kansas University Endowment Associa-
tion, and by several contracts (DA-49-
193-MD-2212, DA-49-193-MD-2788, DA-

MD-49-193-62-G54, and DA-MD-49-193-
63-G94) from the U.S. Army Medical
Research and Development Command.

I am especially grateful to J. Knox
Jones, Jr., who guided this research and
spent many hours critically reviewing the
manuscript. I am also indebted to Karl
F. Koopman and William M. Bass, III,
who have given a great deal of their time
in reviewing the manuscript, and to Terry
A. Vaughan for his valuable counsel re-
garding functional and anatomical as-
pects of chiropteran flight.

Valuable statistical advice and aid in
computer processing were provided by
F. James Rohlf and John Kishpaugh, De-
partment of Entomology, The University
of Kansas, and Dennis Powers, Royal
Ontario Museum, who wrote the com-
puter program (UNIVAR) for obtaining
standard statistics, univariate analysis of
variance, and sum of squares simultane-
ous testing procedure, used in this study.
I also appreciate assistance by J. E. Hill,
British Museum (Natural History), who
provided critical information regarding
holotypes housed in that institution.

Lastly, I take great pleasure in ac-
knowledging: my parents, Mr. and Mrs.
Frank Dale Smith, who encouraged me
and made possible the early part of my
formal education; my wife, Nancy, who
aided me with her patience and typed
the manuscript; many of my fellow grad-
uate students who offered useful sugges-
tions; and Mrs. Ruth Fauhl and Miss
Gayle Hassler, The University of Kansas
Library, for their assistance in obtaining
various important references.

METHODS

Cranial measurements were taken
with dial calipers, calibrated in twen-
tieths of a millimeter, and were recorded
to the nearest tenth of a millimeter.
Standard external measurements were
copied from specimen labels, but were
not used in analyses because of their high
degree of variabiHty, as indicated by high
coefficients of variation, probably reflect-
ing the variation in measuring techniques

MISCELLANEOUS PUBLICATION MUSEUM OF NATURAL HISTORY

Figure 1. Diagrammatic dorsal, ventral, and
lateral views of skull of Pteronotus parnellii
showing measurements used in this study. See
text for fiuther explanation of measurement
terminology and definition. Line above dorsal
view is 10 mm in length.

employed by different collectors (Van
Gelder, 1959:239). The length of the
forearm (measured with dial calipers on
dried skins and selected specimens pre-
served in alcohol) was taken from the
posteriormost projection of the elbow
(olecranon process) to the anteriormost
projecting part of the wrist joint with the
wing in a flexed position. Only adult
specimens (phalangeal epiphyses com-
pletely fused ) were utilized in this study.
Skulls not accompanied by skins were
considered to be adult if all teeth were
fully erupted and the cranium was well
ossified.

Cranial measurements employed in
this study were selected after preliminary
consideration of their variance within and
between samples. Most are standard
measurements customarily used by chi-
roptologists, but some arc of my own in-
vention owing to the unusual shape of
skulls in some mormoopids. All cranial
measurements employed are defined be-
low; letters in parentheses refer to figure

1. All measurements given in the text
are in millimeters unless otherwise noted.

Condijlohasal leugtli (A-A'). — Least dis-
tance from posteriormost projection of exoccip-
ital condyles to the anteriormost projection of
the premaxillae.

Zygomatic breadth (B-B'). — Greatest width
across zygomatic arches at right angle to
longitudinal axis of cranium.

Breadth of hraincase (C-C). — Greatest
width across the hraincase, at right angle to
the longitudinal axis of cranium, at point above
auditory meatus. This measurement was taken
somewhat more posteriorly on Mormoops than
on Pteronotus, owing to the different shape
of the skull.

Mastoid breadth (D-D')- — Greatest width
across mastoidal flanges, at right angle to
longitudinal axis of cranium in Pteronotus;
greatest width across mastoid processes in
Mormoops, because flanges are not developed
in this group.

Zijgorostral length (E-E'). — This measure-
ment, not easily obtained from skulls of some
groups of bats, proved to be a useful dimen-
sion of the frontal portion of the skull. In the
mormoopid bats, the glenoid fossae are raised
above the floor of the basicraniimi (viewed
from below), and are set slightly anterior to
and extend above the auditory bullae; this
allows for the pointed tip of the calipers to be
placed behind the post-glenoid process. The
measurement is the least distance between this
point and the anteriomiost projection of the
maxillae.

Interorhital breadth ( F-F' ) . — Least width
across the interorliital constriction, at right
angle to the longitudinal axis of the craniimi.

Rostral breadth (G-G'). — Greatest widdi
across rostrum, at right angle to longitudinal
axis of cranium; this measurement is approxi-
mately above the first molars in Pteronotus or
second premolars in Mormoops.

Alveolar length of maxillary toothroiv (H-
H'). — Least distance from posterior lip of
alveolus of M3 to the anterior lip of alveolus
of canine.

Alveolar length of mandibular toothroto
(I-T). — Least distance from posterior lip of
aKeolus of m3 to anterior lip of alveolus of
inner incisors.

Breadth of post-palatal extension (J-J'). —
Least width across posterior extension of hard
palate.

Dej)th of brainca.se (K-K'). — Due to the
position of the glenoid fossae (discussed imder
zygorostral length ) this measurement may be
taken directly from the skull withoiit the u.se
of a standardized glass slide, and despite the
fact that auditory bullae may be missing or
damaged; the calipers must have a relatively
thin knife-like blade. To take this dimension,
one blade is laid across the glenoid fossae and

SMITH: SYSTEMATICS OF MORMOOPID BATS

the other blade is brought into contact at the
desired position — in this study the point of
contact was the juncture of tlie midline and
frontal-parietal sutures.

Other quantitative characters used in
the generic analysis were 10 cranial an-
gles, CAB, CAE, CAD, ACE, ADE, DEC,
IFG, IFH, FIJ, FGJ (Fig. 2). Twenty
qualitative, multistate characters, coded
non-additively (Sokal and Sneath, 1963:
77), also were employed in this analysis
and are as follows: attachment of wing
membranes to body; direct or indirect
attachment of wing membrane and
uropatagium to ankle; ligamentous at-
tachment of wing membrane and uro-
patagium free or tightly bound to tibia;
shape of ear; notch on lateral edge of
ear present or absent; tooth-like serra-
tions on antero-medial edge of ear pres-
ent or absent; extent of fusion of ears on
top of muzzle; rostral tubercle present
or absent; extent of lip ornamentation;
lateral spikes on labio-nasal plate (Fig.
11) present or absent; shape of tragus
(Fig. 10); length of dorsal pelage; dis-
tinct color bands present or absent on
hairs of back; flange on anterior edge of
proximal end of the humerus present or
absent; second lower premolar long and
narrow or round and peg-like; number
of lobes on outer lower incisors; degree
of concavity of nasal root; heel on upper
inner incisors present or absent; degree
of bifurcation of upper inner incisors;
structure of upper canine.

Preliminary testing of males and fe-
males of three species, Pteronotus per-
sonatus, P. davyi, and Mormoops megalo-
phylla, revealed significant secondary
sexual variation. Additional study showed
that not all characters were thus in-
fluenced, nor did the same characters of
different species necessarily show sig-
nificant secondary sexual variation (see
discussion beyond). At any rate, during
the initial examination and measurement
of specimens and the subsequent analyses
of geographic variation, males and fe-
males were segregated.

Illustrations of skulls and skeletal ele-
ments were made by tracings from pho-

FiGURE 2. Diagrammatic representation on
skull of Pteronotus parnellii of cranial angles
used in the generic analysis section of this
study. See text for discussion. Line below
lateral view is 10 mm in length.

togiaphs of known scale or with a Wild
Heerbrugg Stereomicroscope fitted with
a drawing tube. The cranial angles, men-
tioned above, also were determined with
the aid of this scope. This was accom-
plished by recording the position of spe-
cific anatomical landmarks of the cra-
nium on paper and then measuring the
angles by means of a 360-degree pro-
tractor. This proved to be a useful tech-
nique, but was somewhat time consum-
ing. In addition to measuring cranial an-
gles, cusp positions on molar teeth were
examined.

Molt was studied by mapping molt
patterns of a number of individuals taken
at various times of the year. Data con-
cerning reproductive activity were re-
corded when present on specimen labels.
In those cases where no information was
given for a known female, an attempt was
made to determine at least whether or
not she had been lactating at the time of
capture, as judged by the presence of
enlarged mammae. Molt data and re-
productive information were treated to-
gether in order to determine whether the
two were correlated. Notations as to
color of pelage are for the most part in
the vernacular. However, in some cases
it was useful to use the color terms of
Ridgway ( 1912); these are capitalized in
the text.

In the systematic accounts the genera
and subgenera are arranged in phylo-
genetic order. Within subgenera, the
type species is considered first. Likewise,

8

MISCELLANEOUS PUBLICATION MUSEUM OF NATURAL HISTORY

in the species accounts, the nominate sub-
species is considered first, followed by
the remaining subspecies arranged from
north to south by geographic unit in cases
where there are more than two recog-
nized races. The familial and generic
synonymies consist of the citation to the
original proposal of the recognized name,
followed chronologically by other syno-
nyms. Subspecific synonymies consist of
the name as originally proposed, followed
by the first usage of the name combina-
tion here employed, if different from the
original name, and any pure synonyms.
Wherever necessary, other name combi-
nations are discussed in the remarks sec-
tion of the accounts of the individual
subspecies.

In the lists of specimens examined,
countries are listed from north to south,
with the Caribbean islands following
mainland localities; states, counties, prov-
inces, departments, or parishes are ar-
ranged alphabetically; localities are ar-
ranged from north to south within politi-
cal subdivisions. If two localities lie at
the same latitude, the westernmost is
listed first. Localities in italic type were
not plotted on distribution maps because
undue crowding of symbols would have
resulted. Additional marginal records
of note are cited in some accounts.

Statistical analijses. — Statistical anal-
yses used in this study were selected for
their utility and ease of interpretation.
Discussions and logic of the testing pro-
cedures used will be found in Simpson
et al (1960), Steel and Torrie (1960),
and Sokal and Rolhf ( 1969). All compu-
tations were performed on a GE 635 com-
putor made available to me by the Uni-
versity of Kansas Computation Center
through funds allotted to the Department
of Zoology.

In the past, some workers have been
satisfied simply to calculate means and
ranges of samples for the interpretation
of variation in mensural data or to use
"Dice-squares" or "Dice-grams," which
incorporate other valuable parameters,
such as standard deviation and standard

error of the mean. The Dice-square first
was used by Dice and Leraas ( 1936 ) .
These devices have been widely em-
ployed, but they have a number of dis-
advantages. As pointed out by Sokal and
Rinkel (1963:473), in the case where a
large number of group-means are to be
compared (such as in a study of geo-
graphic variation), trends are difficult or
impossible to follow by means of the
Dice-square. When the variation is linear
in character or when there are only a
few means to be considered, the Dice-
square may be applied satisfactorily.

Perhaps the most serious drawback to
the Dice-square, is that ^tests for pairs
of group-means have been erroneously
applied to them under the assumption
that ordinary levels of significance apply.
Sokal and Rinkel (1963:473) pointed out
that in such instances the investigator
does not take advantage of a pooled
variance of all samples and thereby ob-
tains a greater number of degrees of
freedom that, in turn, raise the level of
significance of the test being applied.
Also in this regard, such comparisons do
not comply with the random sampling
assumptions of the f-test in that they are
usually conducted a posteriori rather than
a priori. Furthermore, Duncan (1955:
25-26) showed that the protection levels
of the #-test decrease drastically with
each additional comparison beyond two
means. Therefore, a two-mean f-test
may have a protection level of 95.0 per
cent, whereas three- and four-mean tests
have protection levels of approximately
90.2 and 79.7 per cent, respectively.
These levels are so low as to invalidate
the use of multiple Mests as satisfactory
procedures at the five per cent fiducial
level.

For this study, standard statistics
(mean, range, standard deviation, stand-
ard error of the mean, variance, and
coefficient of variation) were computed
for all samples where geographic varia-
tion was of interest, or for both sexes in
those instances where secondary sexual
variation was studied. Where several
group-means were to be compared, a

SMITH: SYSTEMATICS OF MORMOOPID BATS

single classification analysis of variance
(univariate ANOVA) was used to test
for significant variation within the entire
sample of group-means. In this analysis,
an F value is computed and checked in
an F-table under the appropriate degrees
of freedom and the 0.05 level of sig-
nificance. If the observed F value was
greater than the critical F value (from
the table), significance difterences ex-
isted between group-means of the sam-
ple, thus warranting further testing. In
those cases where only two group-means
were tested (for example, comparisons
between males and females for secondary
sexual variation), no further testing was
necessary.

In most cases, significant F values
were observed when a number of group-
means were compared in an analysis of
geographic variation, and it was neces-
sary to continue testing for significant
and non-significant difterences among
these sample means. A wide variety of
"multiple comparisons tests" for a pos-
teriori comparisons of analysis of variance
exists; descriptions of these tests are
found in Sokal and Rohlf (1969). Sokal
and Rinkel ( 1963 ) and Sokal and
Thomas ( 1965 ) used and discussed one
of these, the Student-Newman-Keul mul-
tiple comparisons test, in analysis of geo-
graphic variation. The multiple com-
parisons test I used is Gabriel's ( 1964 )
Sum of Squares Simultaneous Testing
Procedure (SS-STP). The application of
the SS-STP to studies of geographic vari-
ation is discussed in detail by Gabriel
and Sokal (1969). Essentially, the sam-
ple means are ranked in order of de-
creasing magnitude and beginning with
the first mean, sums of squares are cal-
culated by sequentially adding subse-
quent sample means until a maximal non-
significant "subset" of means is deter-
mined. This process is repeated begin-
ning with the second sample mean and
proceeding until all non-significant sub-
sets are determined. The determination
of maximal non-significant subsets is ac-
complished by using a single variance —
the pooled, within group or error mean

square — obtained from the initial anal-
ysis of variance. An addition to the ad-
vantage of subset determination by a
single variance is that in the SS-STP anal-
ysis the probability of making a type 1
error (rejection of a null hypothesis that
is, in fact, true) is determined by an ex-
periment-wise error rate, which in turn
reflects the probability level of the se-
lected "critical" F value (0.05 level, in
this study ) .

Gabriel and Sokal (1969) suggested
that only samples from contiguous loca-
tions should be considered in calculating
maximal subsets of sample means; con-
tiguity being judged by geographic close-
ness. Such a system reduces the number
of comparisons and perhaps gives a more
precise picture of the geographic varia-
tion. While comparisons of locality
means from, say, Sinaloa, Mexico, with
those of Tingo Maria, Peru (certainly
non-contiguous localities), may not be
particularly informative, many system-
atically valid and interesting comparisons
can be made between localities that are
not contiguous. Overall locality compari-
sons in this study yielded a certain
amount of information concerning trends
and effects in insular and peninsular re-
gions.

As yet, no satisfactory method of il-
lustrating the results of the SS-STP has
been developed. Usually, localities rep-
resenting the various sample means are
plotted on a map, and each is identified
in some way with its respective mean,
which appears in an array of means lo-
cated to one side of the map. In this
array, means are arranged in order of de-
creasing magnitude from largest to small-
est and maximal non-significant subsets
of means are delineated by a line that
connects all means within a single non-
significant subset (Fig. 19). Such an
illustration allows the reader to see which
groups of sample means are "clustered"
and significantly different from which
other groups. However, as the number
of sample means ( localities ) and widely
overlapping subsets increases, the clarity
of the illustration tends to decrease. Of

10

MISCELLANEOUS PUBLICATION MUSEUM OF NATURAL HISTORY

course, the easiest situation to interpret
is the occurrence of non-overlapping sub-
sets. In addition, there are problems in-
volved in the correlation and comprehen-
sion of several (in this study 12) of these
SS-STP results simultaneously. Because
a particular locality may exhibit different
subset relationships in different charac-
ters, it would be desirable to be able to
determine the overall relationships be-
tween localities based on all characters
examined.

Gabriel ( 1967 ) developed a simul-
taneous testing procedure for a multi-
variate analysis of variance ( MANOVA),
and a program (MULDIS) is available
in the Numerical Taxonomy System from
the Computation Center at the Univer-
sity of Kansas. However, neither of these
procedures was used because of prob-
lems relating to missing data. Instead,
I employed another package program
( CLNST- Version 2) also available in the
Numerical Taxonomy System at Kansas.
Among other things, this program com-
putes a matrix of Pearson's product-mo-
ment correlations and a matrix of tax-
onomic distance coefficients from stand-
ardized character values. Additionally,
cluster analyses using UPGMA (un-
weighted pair group method using arith-
metic averages) are performed on the
correlation and distance matrices and a
phenogram is generated for each. These
phenograms are next compared with
their respective matrices, and a coef-
ficient of cophenetic correlation (Sokal
and Sneath, 1963) is computed for each
comparison. These matrices and their
respective phenograms may be used to
interpret the overall similarities between
OTUs (standardized locality means in
the case of geographic variation). I em-
ployed the distance matrix and distance
phenogram rather than the correlation
matrix and phenogram, because the co-
phenetic correlations wotc higher for the
former. Computation and discussion of
the distance coefficient can be found in
Sokal (1961), Sokal and Sneath (1963),
and Proctor (1966).

This same set of programs also was

used in the analyzing the generic clas-
sification of mormoopid bats. In addi-
tion, a matrix of correlations among char-
acters and a matrix of the first three
principal components based on correla-
tion among characters were computed.
Projections of OTUs onto these principal
components also were computed and
three-dimensional models were drawn
with a Benson-Lehner incremental plot-
ter from a tape prepared from the pro-
jection and distance matrices (Rohlf,
1967,1968).

HISTORICAL REVIEW

The taxonomic history of the family
Mormoopidae began in 1821 when Wil-
liam Elford Leach described Aello and
Mormoops. It is appropriate to point
out that in Leach's time the classifica-
tion of bats, as well as that of mammals
in general, was in a state of uncertainty
and disagreement. Nearly every promi-
nent taxonomist from the post-Linnaean
period of the Eighteenth Century to the
mid-Nineteenth Century presented, at
one time or another, his own concep-
tion of the natural arrangement of mam-
mals. For the most part, no two classifi-
cations were alike or even comparable.
In the following discussion, the original
orthography is preserved in appropriate
places for historical reasons and ease of
discussion.

Early in the formative period of
chiropteran classification, the mormo-
opid bats were variously allied with the
Noctilionidae and Phyllostomatidae,
both families having somewhat broader
limits than as currently defined. Be-
cause the mormoopids seem to be more
closely related to these two families
than to any other New World assem-
blage, I have attempted in the follow-
ing discussion to maintain the Noctili-
onidae and Phyllostomatidae in proper
perspective with the Mormoopidae.

Gray (1821) divided the bats into
two orders ( Fructivorae and Insecti-
vorae) in the class Cheiroptera. He
divided the Insectivorae into two fami-

SMITH: SYSTEMATICS OF MORMOOPID BATS

11

lies (Noctilionidae and Vespertilionidae)
and further subdivided the former into
three "Races" based primarily on char-
acteristics of the nose — nose simple,
nose warty, or nose appendaged. The
presence or absence of appendages on
the nose also seems to have influenced
Leach. In the family Vespertilionidae,
under the heading RhinophijUis niiUis
(bats without foliaceous nose append-
ages), Leach (1821a: 69) named the
genus AelJo. Within the same family,
but under the heading RhinophijUis in-
structae (bats with foliaceous nose ap-
pendages), Leach (1821b: 74) estab-
lished the genus Mormoops. Neither of
these genera appeared within the con-
text of a general classification until Gray
(1825:338) mentioned the genus Mor-
moops in an arrangement that contained
all bats in one family — Vespertilionidae.
Tliis family was placed in the order
Primates and was further divided into
five tribes, two of which possessed nose-
leaves and were designated Rhinolo-
phina and Phyllostomata. The genus
Mormoops was assigned to the Rhino-
lophitm. No mention was made of the
genus Aello.

Using Gray's tribes, in more or less
the same context, Bonaparte (1831:15)
arranged Mormops (invalid emendation
of Mormoops Leach by Cuvier, 1829)
with the PhyUostomina and Aello with
the Noctilionina. It is interesting to note
that Bonaparte placed all bats known to
him in the family Vespertiliones.

Gray (1838), in a revision of the
genera of bats (Vespertilionidae), fol-
lowed Bonaparte's arrangement of Aello
and Mormoops and incorporated Spix's
( 1823 ) terms Anistiophori ( nose ap-
pendage absent) and Istiophori (nose
appendage present) as subheadings
within the family. Also, in this paper,
Gray described a new mormoopid genus
and species, Pterom)tus Davyi, which he
placed after Aello in the tribe Noctilio-
nina ( Anistiophori ) .

Gray (1839) reported on a small
collection of mammals made by W. S.

MacLeay in Cuba and described an
additional mormoopid genus, Chilomjc-
teris, based on the speci(\s C. Macleayii.
Also present in this small collection was
the second known example of Mormoops
Leach. Until that time, taxonomists had
been misled by Leach's characterization
of Mormoops as having "a single nose-
leaf confluent with the ears." For this
reason, Mormoops was consistently re-
ferred to the family Phyllostomatidae
(or some similar grouping). With the
additional specimen in hand. Gray
(1839:3) concluded that Mormoops did
not have a nose-leaf, and that it was
closely allied to the Noctilionina. In re-
vising his arrangement of 1838, Gray
(1839:3) wrote: "This [Mormoops]
and the new Cuban genus Chilomjcteris,
which agrees with it in many particu-
lars, will together form a new section in
this group [Noctilionina] characterized
by the expansion of the lower lip and
the convex and rounded form of the
skull." This appears to be the first time
that Chilomjcteris and Mormoops were
associated in a distinct group, but no
reference was made to Aello or Ptero-
notus.

Meanwhile, Wagner (1839) pro-
posed a classification in which he asso-
ciated dermopterans (suborder Dermop-
tera) with bats (suborder Chiroptera)
in the order Volitantia. Bats were
divided into three families ( Frugivora,
Istiophora, and Gymnorhina). The lat-
ter two families were further partitioned
into tribes for which Wagner used the
term "Sippen." The tribes Desmodina
and Phyllostomata were contained
within the Istiophora. The Gymnorhina
was divided into three tribes (Brachy-
ura, Gymnura, and Vespertilionina )
based primarily on the nature of the
tail. The genera Mormops [sic], Chilo-
mjcteris, and Noctilio (pp. 445-449),
were assigned to the Brachyura.

An arrangement presented by Lesson
(1842:12-36), illustrates the extreme dif-
ferences of opinion among taxonomists
of that period. Five families of bats.

12

MISCELLANEOUS PUBLICATION MUSEUM OF NATURAL HISTORY

presumably those of Gray (1838) or
Bonaparte (1831), were assigned to the
tril)e Chiroptera, whieli along with the
Bimana and Quadrumana comprised the
order Mastomonodelphie in the subclass
Mammiferes Normanx. The only mor-
moopid genera treated were Aello and
Mormoops, arranged in the Noctilioneae
and Phyllostomineae, respectively.

No major changes were made in
Gray's classification between 1840 and
1860. In this period, however, many new
species were described, particularly
from the Americas. Relevant to this dis-
cussion are the descriptions of the fol-
lowing: the genus Lohostoma and the
species L. cinnamomeum and L. quad-
ridens by Gundlach (1840:356-358);
three new species of Chilonycteris
(gymnonofus, personafa, and ridjii^inoso)
from the Mato Grosso region of Brasil
by Wagner (1843:367); Chilonycteris
fuJiginosa from Haiti by Gray ( 1843a:
20), which according to the describer
resembled C. macleayi; and a new genus
and species Phyllodio porneUii from Ja-
maica, which Gray (1843b: 50) allied
with Mormoops and Chilonycteris.

Peters (1856) re-evaluated Leach's
description of Mormoops hlainmllii, con-
cluded that Leach was correct concern-
ing the presence of a nose-leaf, and
placed Mormoops in the family Phyl-
lostomata. Later, Peters (1864:381-382)
described Mormoops megalophylla from
Mexico and referred Lohostoma cinna-
momeum Gundlach to synonymy under
Mormops hlainviUii Peters, 1856 (not
Leach, 1821). Redescription of taxa
was a common practice in this period,
and M. hlainviUii Peters appears fre-
quently in the literature of the last half
of the Nineteenth Century (primarily
in papers by Peters). Although Peters
gave a lengthy characterization of
hlainviUii, he clearly stated (1856:411)
this his material was no different from
the bat described originally by Leach
in 1821.

M. H. de Saussure (1860:286-293)
allocated the genera Mormops and

Chilonycteris to "Sous-tribu" Mormop-
sins in the "Tribu" Noctilioniens of the
family Vespertilionides. This arrange-
ment appears to be the first in which the
mormoopids were allied \\ithin a dis-
tinct family-group. Also within the
framework of this same tribe, Saussure
erected two other subtribes, NoctUionins
and Emhallonurins. The phyllostoma-
tids were treated together as the only
other apparent "family" in his classifica-
tion.

Koch (1862-63) envisioned the
Mammalia as comprising fovu* major
groups or "Reiche." One of these, the
Primates, he divided into four orders
( "Ordnungen" ) — Bimana, Quadrumana,
Hemipitheci ( including Dermoptera ) ,
and Chiroptera. In a section concerning
families and genera of Chiroptera, he
(p. 298) erected two suborders — Car-
pophagen and Entomophagen; the lat-
ter containing the families Gymnorrhina
and Istiophora. In the Brachyura, a
subdivision of the family Gymnorrhina,
Koch (pp. 303-304) placed Mormops,
Chilonycteris, Phyllodia, Aello, and
Pteronotus along with several other
genera not currently associated with
mormoopids. Nowhere in his classifica-
tion did Koch ( 1861 ) use the term
"Mormopida" which does appear later
(p. 358) in a section concerned with
types of European bats and has been
quoted as the first family-group name
for the mormoopids. The relevance of
this name will be discussed at the end
of this review.

At this point in the history of chirop-
teran classification, both Peters and
Gray recognized the distinctness of
mormoopid bats. Peters (1865:257)
placed Mormops Leach, Chilonycteris
Gray (including Lohostoma Gundlach),
and Pteronotus Gray into a subfamily
group, which he termed Mormopes, in
the family Phyllostomata. In a synopsis
of the vespertilionid and noctilionid
genera, his last major contribution to
chiropteran classification. Gray (1866a)
placed Mormops in a tribe termed Mor-

SMITH: SYSTEMATICS OF MORMOOPID BATS

13

mopsina. The genera Phyllodia, Chilo-
mjcteris, and Pteronotus were allocated
to another, related tribe, Phyllodiana;
both tribes appear in the family Noc-
tilionidae.

In all of the classifications np to this
time, Aello Leach was overlooked or
mentioned only in passing, probably
due to a misplaced holotype. Gray
(1866b) located and re-examined this
crucial specimen. Although he did not
go so far as to synonymize Aello with
Mormoops, Gray did note that Aello
cuvieri agreed with Lohostoma cinna-
rnometim Gundlach, which he supposed
was only a variety of Chilomjcteris
rnacleayii. Peters (1866) also re-exam-
ined the holotype of Aello cuvieri and
found it to agree completely with Mor-
moops hlainvillii. He disregarded this
fact on the assumption that some mix-
up had occurred and the specimen
labeled as the holotype of Aello cuvieri
was not in fact the holotype of that
taxon.

Fitzinger (1870) assigned mormo-
opid bats, except Pteronotus, to the group
Noctiliones in the family Vespertilioni-
dae. For some reason he placed Ptero-
notus in the group Molossi of this same
family. Gill ( 1872 ) , in his arrangement
of the mammalian families, listed the
family-group name Mormopidae, but
gave no indication of included genera.

A milestone in bat classification was
Dobson's ( 1875 ) conspectus, which was
the first attempt to interpret phyloge-
netic relationships within the Chirop-
tera. Employing various anatomical
characters, Dobson (1875:345-346) di-
vided the order into two suborders,
Megachiroptera and Microchiroptera.
He postulated two distinct lines of
descent ("alliances") within the sub-
order Microchiroptera; the Vespertilio-
nidae and Emballonuridae were placed
near the bases of these two lines. Only
the emballonurine line, containing the
families Emballonuridae and Phyllo-
stomatidae, is of importance to this dis-
cussion; the family Noctilionidae was

eliminated, and the genus NoctiUo was
placed in the family Emballonuridae.
Dobson divided the Phyllostomidae into
two sulifainilies ( Lobostominae and
Phyllostominae), each of which was
subdivided into groups of allied genera.
The genera Mormoops, Chilomjcteris,
and Pteronotus were arranged in the
group Mormopes of the subfamily Lobo-
stominae. The remaining subfamily was
divided into four groups of genera —
Vampyri, Glossophagae, Stenodermata,
Desmodontes. As regards the relation-
ships within the emballonurine line,
Dobson thought the Emballonuridae
and Phyllostomidae were linked through
the Noctiliones (Noctilio) and Mormopes
(Chilomjcteris). With few modifica-
tions, this classification was the same
used by Dobson (1878), in which only
two genera, CJiilonycteris (including
Pteronotus) and Mormoops (including
Aello), were recognized in the subfam-
ily Lobostominae (group Mormopes).

Near the turn of the century, im-
portant classifications were proposed by
Flower and Lydekker (1891), Winge
(1892), and Trouessart (1897). The
arrangement presented by Flower and
Lydekker is only slightly different from
Dobson's (1875, 1878), but is of nomen-
clatural importance because it is the
first apparent usage of the subfamilial
name Chilonycterinae for the genera
CJiilonycteris and Mormoops. Winge's
(1892) treatment of the Chiroptera is
notable because he removed the genus
Noctilio from the family Emballonuridae
and associated it with the mormoopids
in his second "main division" of the
family Phyllostomatidae that he termed
Mormopini. He felt that the Mormopini
were somewhat more specialized than
bats of his other division (Phyllostoma-
tini). Trouessart ( 1897 ) placed the
genera CJiilonycteris, Pteronotus, and
Mormoops in a subfamily ( Mormopi-
nae) of the family Emballonuridae.
The genus Noctilio was assigned to one
of the sections (Noctilioneae) of the
subfamily Emballonurinae.

14

MISCELLANEOUS PUBLICATION MUSEUM OF NATURAL HISTORY

Mormoopid genera were reviewed
taxonomically bv Rehn (1902, 1904a,
1904b). Rehn (1902) studied 49 speci-
mens of Mormoops and asserted that
Mormoops, Chilomjcteris, and Dermo-
notus ( proposed as a replacement name
by Gill, 1901, for the supposedly pre-
occupied Pteronotus) comprised a well-
defined subfamily within the Phtjllo-
stomatidae. To this subfamily, he (p.
162) applied the name Mormoopinae
(see also Rehn, 1901). He rejected
Lobostominae Dobson, 1875, on the
grounds that Lobostoma Gundlach,
1840, "is a synonym of Mormoops and
Cliilonycteris." Under the current code
of nomenclature, the name Lobostomi-
nae is rejected for other reasons as ex-
plained below. Rehn (1904a) reviewed
Chilonycteris, based on an examination
of 103 specimens, and (1904b) Dermo-
notus (= Pteronotus), based on 36 in-
dividuals.

The taxonomic arrangement of Mil-
ler (1907) is the basis for the current
classification of the order as presented
by Simpson (1945), Hall and Kelson
( 1959 ) , and Koopman and Cockrum
( 1967 ) . Miller's classification was by no
means novel in its conception, having
been drawn in part from the earlier
works of Gray, Dobson, Winge, and
others. The order Chiroptera was ar-
ranged into two suborders (as defined
by Dobson, 1875), 17 famihes, and 19
subfamilies based on the comparative
anatomy of the wing, shoulder girdle,
sternum and associated ribs, and the
number and structure of the teeth. In
this classification, Miller placed the
genera Chilomjcteris, Pteronotus, and
Mormoops in the subfamily Chilonyc-
terinae ( Flower and Lydekker, 1891 ) of
the family Phyllostomidae, which he
further divided into six other subfam-
ilies— Phyllostominae, Glossophaginae,
Hemiderminae (^ Carolliinae), Sturni-
rinae, Stenoderminae, and Phyllonycteri-
nae. Simpson ( 1945 ) retained Miller's
classification, modifying it by grouping
the families into superfamilial categories

based on the "families" of Winge
(1892).

Recently, Dalquest and Werner
( 1954 ) accorded full familial rank to
the mormoopid genera and used the
name Chilonycteridae. Their conclu-
sions (to be considered in somewhat
greater detail below) were based on the
histological aspects of facial glands of
North American bats. This study seems
to have been overlooked by many mam-
malian systematists.

According to The International Code
of Zoological Nomenclature, 1958, Arti-
cle 29(d): "A family-group taxon
formed by the union of two or more
taxa of that group takes the oldest valid
family-group name among those of its
components, with change of termination
if required." From the foregoing his-
torical review, it is clear that Mormoops
was used consistently during the last
half of the Nineteenth and early part of
the Twentieth centuries as the basis for
various family-group names, but the
name Chilonycterinae was first used by
Flower and Lydekker (1891). In fact,
family-group names based on the genera
PhijUodia (Phyllodiana Gray, 1866a)
and Lobostoma ( Lobostominae Dobson,
1875) predate Chilonycterinae. There-
fore, the family-group name for the bats
here treated should be based on the
generic name Mormoops Leach, by pri-
ority, rather than on PhijUodia, Chilo-
nycteris, or Lobostoma.

The selection of authorship for the
earliest family-group based on Mor-
moops requires further consideration.
Palmer (1904:754) listed Mormopida
Koch (1862-63) as the first apparent
usage of Mormoops as the basis of a
family-group name. Both Miller (1907:
118) and Simpson (1945:57) also cited
this usage in their familial synonymies.
Miller ( 1907 ) took his citation directly
from Palmer ( 1904 ) and Simpson
( 1945 ) seems to have done the same.
The name Mormopida of Koch appears
once on page 358 in the following con-
text: "Eine davon (Synotus barbastel-

SMITH: SYSTEMATICS OF MORMOOPID BATS

15

lus) ivird zwar von anderen Forschern
ivegen der diirch die stellung der Nasen-
locher hedingten Physiognomie zu den
Trutzern (Mormopida) gestellt; da loir
aher das Verhaltnisz des Schivanzes zur
Periscelis hei der Eintheilung der Grup-
pen festhalten mussen, so kann die Gat-
tung Sijnotus nur hei den Vespertilionen
untergehrocht icerden." This is the only
mention of the term "Mormopida" in
Koch's entire paper nor does it appear
in his taxonomic arrangement, discussed
above. Inasmuch as this usage was
vague and without indication of in-
cluded kinds, the name "Mormopida" of
Koch must be designated a nomen
dubiiim.

As noted previously, Saussure (1860:
2<S6), clearly allocated the genera Mor-
moops and Chilonycteris to a family-
group (Sous-tribu) Mormopsins. Saus-
sure (1860) seems to be the first author-
ity to so group these bats and, there-
fore, must be recognized as the author

of the family-group name. In accord-
ance with Article 10(e) of the Code, the
name Mormopsins must be fully latin-
ized to Mormoopidae — Mormo (ghost),
ops (face) =r Mormoops, nominative
singular; Mormoop/os, genitive suffix
plus -idae == Mormoopidae). This
familial name is firmly based on priority.
Furthermore, Rehn (1902, 1904a, and
1904b), the most recent and major re-
viser of the entire group, used the name
Mormoopinae. Finally, use of this
name in preference to Chilonycteridae
is in the interest of nomenclatural sta-
bility.

As earlier noted, I have raised the
group to full familial status. My rea-
sons for doing so are presented in the
following section. Henceforth, I shall
refer to these bats as the Mormoopidae
or mormoopid bats, but may, from time
to time, use the terms "chilonycterine"
or "Chilonycterinae" in directly quoted
passages.

ANALYSIS OF CHARACTERS

The genera Chilonycteris (including
Phyllodia), Pteronotus, and Mormoops
were placed together in the subfamily
Chilonycterinae of the family Phyllosto-
midae by Miller (1907), who based his
classification on a detailed study of the
skeleton, particularly the wing, shoulder
girdle, and structure of the tooth cusps.
Although each family was characterized
in detail, many characters mentioned
were shared by several families. For
example, by my count, the greater tu-
berosity or trochiter was described as
large in size and articulating, in some
fashion, with the scapula in 11 different
families, seven of which occur in the
New World. This "secondary articula-
tion" is absent in the Mormoopidae. The
distal end of the humerus was described
for phyllostomatids in rather general
terms (pp. 116-117): "epitrochlea mod-
erately developed, usually with spinous
process, capitulum distinctly out of line
with the shaft." More or less the same

wording was used to characterize the
distal end of the humerus of at least six
other New World families. Likewise,
the presence of three bony phalanges in
the third manal digit in phyllostomatids
is shared with three other New World
families.

Reduced to the essential common
denominator, phyllostomatids can be
characterized as those New World bats
that possess a cutaneous nasal append-
age, commonly referred to as the nose-
leaf. Although generally well-devel-
oped, the nose-leaf may be greatly re-
duced as in the subfamilies Phyllonyc-
terinae (including BrachyphyUa as sug-
gested by Taboada and Pine, 1969) and
Desmodontinae (included in this fam-
ily following the suggestion of Machado-
Allison, 1967; and Forman et al, 1968)
and the stenodermine genus Sphaero-
nycteris. In another stenodermine genus
Centurio, the nose-leaf is curiously con-
torted and nearly obscured by the many

16

MISCELLANEOUS PUBLICATION MUSEUM OF NATURAL HISTORY

intricate ridges and folds. The tiny
wart-like bumps above and between the
nostrils of mormoopids (Fig. 10) have
been interpreted as a rudimentary nose-
leaf (notably by Leach and Peters).
Phyllostomatids also are unique among
New World Chiroptera in their adap-
tive radiation. From the outset of this
study, it was apparent that the only
possible reason to consider mormoopids
as members of the family Phyllostomati-
dae was to assume the presence of a
rudimentary nose-leaf.

One of my major objectives has
been to test the arrangement of
the mormoopid genera as a family
distinct from the Phyllostomatidae.
In doing so, I tried to avoid use of
classical characters employed by Miller
( 1907 ) and others and sought new char-
acteristics on which a classification
could be constructed. Like my prede-
cessors, I turned to the skeleton and soft
anatomy in search of such characters.
In order to understand the latitude of
variation of a particular character, I
examined representatives, both general-
ized and specialized, of every New
World family and several selected Old
World groups. Unfortunately, there is
a paucity of complete skeletal material
in the mammalian collections of major
North American institutions, and the
examination of more than one repre-
sentative of a taxon was not always
possible. In arranging mormoopids as a
distinct family, I relied heavily on the
following characters : 1 ) structure of
the distal end of the humerus; 2) struc-
ture of the proximal end of the humerus
and its relationship with the scapula;
3) structure of the proximal end of the
femur; and 4) structure and shape of
the tragus. When possible, I have indi-
cated the functional significance of the
characteristics employed in this study.
However, our knowledge of bats is still
relatively limited and the functional
significance of characters studied was
not always apparent. A further handi-
cap is the virtual absence of a fossil

record; the evolutionary stage of a given
character only could be surmised.

In the transition from a terrestrial
mode of life to a volant habit, many
aspects of the chiropteran anatomy have
undergone marked adaptive changes.
The most noticeable of these modifica-
tions are changes in the forelimbs and
pectoral girdle. As might be expected,
these changes involve more than elonga-
tion of the forearm and phalangeal ele-
ments to form a wing. Whereas the
forelimb of quadrupedal, terrestrial
mammals functions in the support of
the body as well as in locomotion, the
principal function of the forelimb in
bats is the latter. In addition, the opera-
tional plane of the forelimbs is vertical
to the longitudinal axis in terrestrial
mammals, but in bats this plane has
been changed to a lateral position, due
to a 90-degree dorsal rotation of the
forelimb. Muscles that formerly func-
tioned to suspend the body between the
forelimbs have been modified to in-
crease the efficiency of the wing beat.
The three major points of articulation
in the wing ( scapula-humerus, humerus-
radius, and radius-carpus) also have
undergone marked changes. Processes
on the wing bones that normally are
lateral or medial to the body in terres-
trial mammals become dorsal and ven-
tral, respectively, in bats.

Distal End of Humerus

The structure of the distal end of the
humerus provides several interesting
features that seem to be of taxonomic
value. Miller (1907:13-17) discussed
the modifications of the humerus with
reference to its structure in the Mega-
chiroptera and Microchiroptera. He was
primarily concerned with the size of
the trochiter and trochin (see Fig. 6 for
definition of terms) and their position
relative to the head of the humerus.
Distally, he noted the size and position
of the capitulum, the development of
the external ( lateral epicondyle ) and in-
ternal condyles (medial epicondyle),

SMITH: SYSTEMATICS OF MORMOOPID BATS

17

and the presence or absence of a spi-
nous process. Lawrence (1943) noted
tlie value of the humerus as a generic
character in several vespertilionids.
Some paleontologists, Revilliod ( 1917a,
1917b, 1920, and 1922) and Savage
(1951), have used the humerus as an
indicator of familial alliances.

Before considering the distal end of
the humerus in detail, it is useful to
consider the elbow joint of bats and the
general modifications that have taken
place. The humerus and radius are the
two major elements that form this joint.
In bats the ulna, which forms the largest
part of the forearm segment in terres-
trial mammals, is greatly reduced in size
and is represented by little more than
the olecranon process. This process, im-
portant in bats as well as in terrestrial
mammals, is the point of insertion for
the tendons of the M. triceps hmchii, a
major extensor of the forearm.

The most conspicuous part of the
distal end of the humerus is the spool-
shaped articular surface. In terrestrial
mammals this surface can be divided
into two parts — the capitulum, which
normally articulates with the proximal
portion of the radius, and the trochlea,
which forms the articular surface for
the ulna. Inasmuch as the ulna is vir-
tually absent in bats, the radius has
expanded medially (that is, ventrally)
to articulate with the trochlea in addi-
tion to the capitulum (Fig. 5). The
ulna, although reduced, still maintains
a small articular surface that fits into a
depression on the postero-medial edge
of the triangular articular surface of the
radius and is confluent with that surface
(Fig. 5, E-H).

The result of these modifications has
been to strengthen the elbow joint. The
motions of pronation and supination,
advantageous to some mammals, are
completely lost by the Chiroptera and
the hingelike elbow joint is restricted to
extension and flexion. Similar modifica-
tions that have produced the same sorts
of results can be seen in the forelimbs

of some ungulates, as noted by Gregory
(1949).

In his diagnoses of each family,
Miller (1907) used the term "spinous
process." According to these diagnoses,
every New World family had a spinous
process of one sort or another. Con-
siderable variation in structure is ex-
hibited by New World genera with re-
spect to the region medial to the distal
articular facet, and for this reason I
have found it useful to redefine the
spinous process. First, in a number of
bats there is a process that extends me-
dially (ventrally) from the medial epi-
condyle and that could be called the
ventral process, but to preserve conven-
tional anatomical nomenclature I will
refer to it as the medial process ( Fig. 3).
The length of this process may be equal
to the width of the distal articular sur-
face when well developed or may be
short and compressed against the medial
epicondyle when only weakly developed.
In some bats there is an additional struc-
ture that extends distally from the tip of
the medial process. Hereafter, I will re-
fer to this structure as the distal spinous
process (Fig. 3).

Capitulum and trochlea. — As men-
tioned above. Miller (1907) commented
on the position of the capitulum and
trochlea relative to the longitudinal axis
of the humerus. In figures 3 and 4 it is
evident that these articular surfaces
may be displaced laterally (that is, dor-
sally) from the longitudinal axis or may
be more or less bisected by this axis.
On this basis alone, the Vespertilionidae
(Fig. 3, R-S), Molossidae (Fig. 3, T-U),
and the Emballonuridae (Fig. 3, A-C)
may be separated from all other New
World families (but the emballonurids
do not resemble either the vespertili-
onids or molossids in other characters
of the distal end of the humerus). The
remaining families have an offset capit-
ulum and trochlea, although in the
genus Noctilio (Fig. 3D) and the mor-
moopid genera (Fig. 4) the capitulum
and trochlea lie in a somewhat inter-

18

MISCELLANEOUS PUBLICATION MUSEUM OF NATURAL HISTORY

Figure 3. Anterior view of the distal end of the left humerus of representative New World bats as
follows: A, Rhynchontjcteris naso; B, Balantiopteryx plicata; C, Diclidunts virgo; D, Noctilio lahialis;
E, Microntjcteris sijlvestris; F, Lonchorhina aurita; G, Glossophaga soricina; H, Anoura geoffroyi; I,
Lonchophylla robusta; J, Carollia perspicillata; K, Stitrnira lilium; L, Artiheus toltecus; M, Cen-
turio senex; N, Desmodus rotundus; O, Natalus stramineus; P, TliyioiHem tricolor; Q, Furipterus hor-
rens; R, Myotis grisescens; S, Lasiurus borealis; T, Tadarida hrasiliensis; U, Molosstis molossus. The
labels on A and C refer to the following structures mentioned in text: 1, lateral surface of capitu-

SMITH: SYSTEMATICS OF MORMOOPID BATS

19

Figure 4. Anterior view of the distal end of the left humems of five mormoopids: A, Pteronohis
parnellii; B, P. personatus; C, P. davyi; D, P. suapurensis; E, Moormoops megalophylla. See text for

discussion. The line above C is 4 mm in length.

mediate position. The lateral displace-
ment of these distal articular facets
seems to be correlated with the large
size of the muscle mass of the forearm
that originates from the medial process.

As I have mentioned, the distal
articular surface is spool-shaped and in
conjunction with the expanded proximal
surface of the radius effectively restricts
the elbow to a hinge-joint. During
flight however, a great deal of air pres-
sure must be sustained by this joint.
Therefore, lateral ( dorso-ventral ) move-
ment at this point must be minimized
if the wing is to remain rigid and not
buckle under these pressures.

The distal articular surfaces are
variously shaped (Fig. 3). Miller
( 1907 ) also noted these differences, but
failed to note their functional impor-
tance. It has already been stated that
bats with heavy forearm muscle masses
usually have rather well-developed me-
dial processes. In these bats (phyllosto-
matids and emballonurids, for example)
the distal articular surfaces are not
markedly modified. The central part of
the spool, the central portion of the
capitulum, is spherical in phyllostoma-
tids or slightly oblong in some emballo-
nurids; the notable exception here is
Diclidurus (Fig. 3C). The lateral part

of the capitulum and the medial troch-
lea may be nearly equal in distal exten-
sion or the trochlea may extend some-
what farther distally than either part of
the capitulum. At first glance, it would
seem that this type of elbow joint (Fig.
5C) might be vulnerable to the exces-
sive pressures that develop in this re-
gion during flight. However, in these
bats the proximal forearm muscle mass
may act as a binding and prohibit any
appreciable lateral movement at this
joint.

With the development of faster,
more efficient modes of flight, the size
of the forearm muscle mass is reduced
and its functions replaced by automatic
devices (discussed beyond). When
this muscle mass is reduced, the elbow
joint must be reinforced to withstand
aerodynamic pressures. Accordingly,
the articular surfaces of the elbow re-
gion have undergone several types of
structural reorganization. The simplest
of these is seen in Pteronotus (Fig. 4,
A-D). The central part of the capitu-
lum is slightly larger and the corre-
sponding articular surface of the radius
is deepened to form a "tongue-and-
groove" arrangement. A reverse of this
construction is seen in Mormoops (Fig.
4E) and Diclidurus (Fig. 3C). The

lum; 2, trochlea; 3, medial process; 4, medial epicondyle; 5, central surface of capitulum; 6, radial
fossa; 7, distal spinous process. See text for discussion of distal end of the humerus. The line above

B is 4 mm in length.

20

MISCELLANEOUS PUBLICATION MUSEUM OF NATURAL HISTORY

Figure 5. Anterior view of elbow region of four bats: A, Pteronotus parnellii; B, Mormoops mega-
lophijlla; C, Artiheus toltecus; D, Molossus 7nolossus. Below each of these (E-H) the radius is fig-
ured to show the articular facets. The labels on D and H identify: 1, radius; 2, flexor fossa, into
which the tendons of the M. biceps brachii and M. hrachialis insert; 3, reduced ulna; 4, sesamoid
bone. See text for discussion. The line below E is 4 mm in lengtli.

central portion of the capitulum is re-
duced in size and the lateral part of the
capitulum and the trochlea remain more
or less unchanged. The portion of the
radius that embraces the central capit-
ulum is modified to fit into the scooped-
out distal articular surface. Another
modification of the "tongue-and-groove"
arrangement is seen in the vespertili-

onids and molossids (Fig. 3, R-U), in
which the rounded surface of the cen-
tral capitulum becomes angular or nar-
rowed and tilted diagonally. In con-
junction with these structural modifica-
tions of the distal articular surface of
the humerus, the groove in the proximal
articular surface of the radius is deep-
ened and shaped to correspond with the

SMITH: SYSTEMATICS OF MORMOOPID BATS

21

central capitulum. All of these modifi-
cations in the structure of the spool-
shaped articular surface of the distal
end of the humerus increase the stabil-
ity of the elbow joint by restricting
lateral movement.

Spinous and medial processes. — Al-
though features of the distal end of the
humerus are important in grouping vari-
ous taxa, other striking features are the
structures located medially (that is,
ventrally) to the capitulum and troch-
lea. These are the medial and distal
spinous processes. The presence of a
strong medial process is characteristic
of the Phyllostomatidae (Fig. 3, E-N)
in that this process is large and bulky
and associated with a heavy forearm
muscle mass. The medial process of the
emballonurids (Fig. 3, A-C) is rather
short and broad, and the extreme por-
tion is curled distally. For the most
part, the Noctilionidae (Fig. 3D) re-
sembles the emballonurids in this re-
spect. Curiously, in the genus Diclidti-
rus, this region of the distal end of the
humerus is greatly modified; the medial
process is shortened, slanted distally,
and gives rise to a short distal spinous
process. In contrast, the medial process
of vespertilionids ( Fig. 3, R-S ) is mark-
edly smaller and less pronounced; the
forearm muscle mass is also noticeably
lighter. The bats of the family Molos-
sidae (Fig. 3, T-U) have an extremely
short medial process, but the distal spi-
nous process is typical.

The Phyllostomatidae can be divided
into two distinct groups, based on the
shape of the medial process. One of
these, the Micronijcteris-\me, has a bi-
lobed medial process. The other, the
PhyUostomus-\ine, exhibits a rather sim-
ple medial process that lacks any sign
of bilobation. The MicronycterisAine in-
cludes the phyllostomatine genera Mi-
cronycteris, Macrotus (not figured),
and Lonchorhina, the Glossophaginae
(Fig. 3, G-I), and the Carollinae (Fig.
3J). The genera Lonchorhina, Macro-
tus, and Lonchophylla of the Micro-

nycteris-line have departed slightly from
the basic bilobed medial process by
enlargement of the distal lobe to vary-
ing degrees. This lobe is particularly
well-developed in Lonchorhina and
superficially resembles the distal spi-
nous process of moiTnoopid genera ( Fig.
4). I was concerned, at first, by this
distally projecting lobe and wondered
if it might be an intermediate step to-
ward the development of the distal
spinous process as exemplified in the
Mormoopidae. Terry A. Vaughan kind-
ly consented to dissect several individ-
uals of Lonchorhina with special ref-
erence to this question. After doing so
he wrote ( personal communication ) :
". . . the spinous process in the species
Lonchorhina aurita is not homologous
to that in the chilonycterines, at least if
the muscle attachments are used as an
indication of homology." In addition,
this elongated lobe in the phyllostoma-
tids does not appear to be associated
with any automatic devices as discussed
beyond.

Although skeletal material of the
phyllostomatine genera Tonatia, Mi-
mom, Chrotopterus, and Va7npijrum was
not available for examination, they are
here grouped into the Phyllostomus-line
along with Phyllostomus, Trachops, and
PhyllodernM. Other members of this
group include the Sturnirinae, Steno-
derminae, Phyllonycterinae, and prob-
ably the Desmodontinae.

Three small families, Natalidae,
Furipteridae, and Thyropteridae, resem-
ble the more generalized phyllostoma-
tids in that the distal articular surface is
offset from the shaft of the humerus and
the medial process is moderately devel-
oped. This process is most strongly de-
veloped with respect to these three
families in the genus Thyroptera (Fig.
3P), in which there are two small pits
for muscle attachment in the extreme
tip of the process. In both the genera
Natahis and Furipterus (Fig. 3, O and
Q, respectively), the medial process is
noticeably different in comparison with

22

MISCELLANEOUS PUBLICATION MUSEUM OF NATURAL HISTORY

Thyroptera. It is moderately short with
a prominent "spinous process" com-
pressed against the medial epicondyle
in the genus Furipterus, whereas in
Natalus the medial process is wider and
appears to be trilobed.

The Vespertilionidae and Molossidae
have extremely short medial processes
that are compressed against the medial
epicondyle. Viewed from the medial
side, the spool-shaped distal articulation
has a deep fossa into which the medial
process protrudes. Typically, vesper-
tilionids do not exhibit distal spinous
processes (see Mijotis, Fig. 3R). How-
ever, in at least one genus, Lasiurus
(Fig. 3S), there is a pronounced distal
spinous process that is especially promi-
nent in the large Lasiurus cinereus; the
spinous process seems to be associated
with sustained flight and is of interest
inasmuch as these bats are well-known
for their long migratory flights.

One of the most striking difi^erences
between the Vespertilionidae and Molos-
sidae with regard to the distal end of
the humerus is the presence of a long,
well-developed distal spinous process in
the latter family. Among the molossid
genera examined in this study, Tadaricla
(Fig. 3T) exhibited the shortest distal
spinous process and Molossus moJossus
( Fig. 3U ) had the longest.

The medial process of the Monno-
opidae is of moderate length in the
species of the genus Pteronotus (Fig.
4, A-D) and markedly shortened in
Mormoops. In all mormoopid genera,
the distal spinous process is well de-
veloped and projects well beyond the
medial epicondyle and distal articular
surface. As regards the distal end of
the humerus, the Mormoopidae is as
distinct from the Phyllostomatidae as
the latter is from any other family.

Functional significance of the distal
end of the humerus. — The distal end of
the humerus, in conjunction with the
radius and to a certain extent the ves-
tigial ulna, forms the articular joint on
which the forearm pivots. In addition,

muscles that control the movement of
the manus originate from this region.
During the downstroke (or power
stroke), this region must be rigid in
order to withstand the air pressures
from below and obtain optimal lift from
energy expended by the pectoral and
associated muscles. During the up-
stroke, this region, along with the car-
pus, is flexed to allow for recovery of
the wing in preparation for the sub-
sequent power stroke.

In considering the functional signifi-
cance of the distal end of the humerus,
it is useful to review some of the mech-
anisms and also the evolution of chirop-
teran flight. Owing to the paucity of a
fossil record for bats (the scanty fossil
remains reveal little about the actual
transition to a volant habit), the student
of chiropteran functional anatomy must
speculate on the evolution of flight
based on observations of living repre-
sentatives. The oldest known fossil
(Icaronijcteris index Jepsen, from the
early Eocene of Wyoming) is clearly a
full-fledged bat, as also are the two Old
World Eocene fossils, Archaeonycteris
Revilliod and Palaeochiropteryx Revil-
hod.

The early precursors of the Chirop-
tera probably were arboreal insectivores
capable of short, gliding "fluttery"
flights in pursuit of small flying insects.
The bony structure of the forelimbs of
these early ancestors may not have been
noticeably different from those of other
arboreal, non-gliding kinds; this is to
say, the ulna may have been a moder-
ately well-developed bone and the distal
articular surfaces of the humerus still
possessed both radial and ulnar facets.
The medial process may have been tri-
angular and supported a rather massive
forearm muscle mass. A comparable
situation in which morphological ap-
pearance does not necessarily reflect
differences in behavior exists today be-
tween tree squirrels of the genus Sciurus
and the gliding squirrels of the genus
Glaucomys. The early precursors of

SMITH: SYSTEMATICS OF MORMOOPID BATS

23

bats may have been nocturnal or, as
seems more likely to me, crepuscular,
becoming nocturnal as appropriate de-
vices for acoustic orientation were de-
veloped.

The wing was formed by an elonga-
tion of the forearm and manal elements
and, in the course of its formation,
probably was "Happed" with varying
degrees of success — from clumsy flut-
tering to sustained flight with a rela-
tively stable wing beat. Flexing and
extending of the wing in these early
bats probably was accomplished through
strict muscle control. As mentioned
above, the forearm muscle mass was
probably relatively heavy in the early
bats and the medial process well de-
veloped to support this musculature.
Revilliod (1917a: 173, fig. 9) illustrated
the distal end of the humerus of Falaeo-
chiropteryx spiegeli in which the distal
articular surface is almost completely
set to one side of the longitudinal axis
of the shaft and the medial process is
massive and must have supported a
heavy forearm musculature. Inasmuch
as the medial process was evidently
large in the early bats and is also large
in those living groups that possess a
generalized mode of flight, I believe
that the presence of such a well-de-
veloped medial process is an indicator
of primitiveness. I therefore regard the
Phyllostomatidae as a generalized group
of bats with respect to the distal end
of the humerus.

As bats continued to adapt to the
aerial niche, there evolved more efficient
modes of flight. In some, efficiency im-
proved through slight modifications of
pre-existing structures. One of the more
important was the development of the
distal spinous process which, in conjunc-
tion with the M. fexor carpi ulnaris,
became involved in the automatic con-
trol of the flexion of the distal parts of
the wing. In a generalized bat such as
Macrotiis (Phyllostomatidae), this mus-
cle is large and has a heavy, fleshy belly
(Vaughan, 1959:91), and the distal seg-

ment of the wing is flexed solely by
muscular effort. In Macrotus as well
as in several other phyllostomatids
( Vaughan, personal communication ) ,
the origin of the M. flexor carpi ulnaris
is from the proximal end of the ulna,
the medial tip of the medial process, or
from a combination of both. As Vaughan
(1959:91) stated, the origin of this
muscle is not affected by movements of
the radius.

In contrast, Vaughan (1959:91)
noted that the elastic belly of the M.
flexor carpi ulnaris was relatively small
in Eumops perotis and Myotis velifer,
and that in E. perotis the elasticity of
the muscle is reduced by a thick fascial
sheath that envelops the belly portion
and extends from the large tendons of
origin and insertion. In M. velifer,
there is a tendonous core that passes
down the center of the belly from origin
to insertion. In both of these rather
specialized bats, the M. flexor carpi
ulnaris originates from the tip of the
distal spinous process and inserts on
the pisiform, which in turn is firmly at-
tached to the fifth metacarpal. The re-
duced elasticity of this muscle to a
cable-like strand of connective tissue, as
well as its attachments, allow it to flex
the fifth digit and thereby the entire
manus when the forearm (radius) is
flexed by the M. biceps brachii. The
spinous process acts as a proximal ex-
tension of the radius away from the
center of rotation in the elbow joint. As
the radius is flexed, the tip of the distal
spinous process and the origin of the M.
flexor carpi ulnaris travel through a
wide arc, thereby flexing the fifth digit
with little muscular effort. Similar auto-
matic devices for the extension of the
manus are discussed bv Vaughan (1959:
81).

Aside from the energy conserved by
such automatic devices, the heavy
muscles that ultimately control the dis-
tal portion of the wing are located
proximally, nearer the center of gravity,
and the potential weight cf the wing is

24

MISCELLANEOUS PUBLICATION MUSEUM OF NATURAL HISTORY

reduced. In addition, because the con-
trol is automatic, a certain amount of
precision may be introduced into the
wing beat. All of these factors contrib-
ute to the increased overall efficiency of
flight.

In the Mormoopidae, the develop-
ment of the distal spinous process and
the structure of the M. flexor carpi ul-
naris and their involvement in an auto-
matic Hexing device seem to be directly
comparable to those discussed above for
vespertilionids and molossids. The
functional moiphology of the forelimb
of monnoopids was discussed by
Vaughan and Bateman (1970). Their
conclusions agree with the above, and
they concluded that monnoopids have
a much more specialized forelimb than
phyllostomatids.

In summary, the construction and
function of the distal end of the humer-
us appears to set the mormoopids apart
from the Phyllostomatidae and other
groups of bats. These differences likely
relate to a somewhat more specialized
mode of ffight in mormoopids as com-
pared to the rather generalized flight of
phyllostomatids.

Proximal End of Humerus

In classifying bats, Miller (1907)
evidently was greatly influenced by the
structure of the proximal end of the
humerus. In fact, the size of the greater
tuberosity (trochiter) relative to the
lesser tuberosity (trochin), and whether
or not the trochiter impinged on the
scapula, were major characters used in
his classification. Aside from the head
of the humerus, the greater and lesser
tuberosities are the most prominent fea-
tures of the proximal end of that bone.
As mentioned above, Miller noted that
the greater tuberosity was frequently
larger than the lesser, and in most cases
projected above (proximal) the head
( Fig. 6 ) . However, in several groups of
bats — emballonurids (Fig. 6A) for
example — the greater tuberosity barely
rises above the head; in other families —

Molossidae (Fig. 61) for example — the
greater tuberosity is extremely long and
projects well above the head of the
humerus. In those bats with long
greater tuberosities, Miller also noted a
peculiar articulation between this tuber-
osity and the scapula, anterior and
medial to the glenoid fossa. He called
this articulation a "secondary" or
"double" articulation. According to
Miller (p. 13), this unique articulation
allowed for a stronger joint, while at
the same time limiting motion to a
single plane. However, Vaughan (1959:
54) pointed out that such an articula-
tion would limit the movement of the
humerus, not to a vertical plane as
Miller evidently thought, but to a hori-
zontal ( antero-posterior ) plane that, in
Vaughan's estimation, could not produce
the wing beat.

Vaughan (1959:72) interpreted the
elongated greater tuberosity and scapu-
lar contact as a "locking" device to stop
the upstroke and increase the efficiency
of the power stroke. As the humerus
reaches the peak of the upstroke, the
greater tuberosity strikes the scapula
and thereby transfers the force of the
stroke to that bone. The scapula tends
to rock upwards, lifting the lateral bor-
der as it absorbs this transfer of force.
The coordinated contraction of the pos-
terior division of the M. serratus anterior
effectively stops the upstroke and starts
the power stroke in a whiplash fashion.
The power stroke, thus started, is car-
ried through by the powerful M. pec-
toralis and M. subscapularis.

One of the major differences pointed
out by Miller (1907:116-117) between
mormoopids and phyllostomatids was
the absence of the secondary articula-
tion in the former. The greater tuberos-
ity of phyllostomatids exceeds (prox-
imally) the rounded head of the hu-
merus (Fig. 6, C-D). Both the M.
supraspinatus (elevates, abducts, and
rotates Immerus) and M. infraspinatus
(flexes, abducts, and rotates humerus)
insert on the tip of this process. In

SMITH: SYSTEMATICS OF MORMOOPID BATS

25

Figure 6. Medial, dorsal, and proximal (above) views of the proximal end of the humerus of repre-
sentative New World bats: A, Balantiopteryx plicata; B, Noctilio lahialis; C, Micromjctcris stjivestris;
D, Desmodus rotundus; E, Natalus stramineus; F, Furiptenis horrens; G, Thtjroptcra tricolor; H,
Mtjotis grisescens; I, Molossus molossits. The labels on A and C refer to the following structures
mentioned in text: 1, greater tuberosity (trochiter); 2, head; 3, pectoral ridge; 4, lesser tuberosity
(trochin); 5, supraglenoid fossa. See text for discussion of the proximal end of the humerus. The

line between H and I is 4 mm in length.

26

MISCELLANEOUS PUBLICATION MUSEUM OF NATURAL HISTORY

accordance with the scheme described
by Vaughan (1959:72) and outHned
above, the greater tuberosity seems to
make contact with the scapula during
the terminal part of the upstroke. The
locking function of this system is not so
highly developed as in the vespertilio-
nids or molossids and a greater degree
of freedom is apparent. The rounded
head of the humerus also may indicate
a greater capability for rotation of the
humerus in the phyllostomatids than in
some of the more specialized families.

The lesser tuberosity of phyllosto-
matids projects ventrad and is con-
nected to the shaft by a heavy ridge that
eventually forms the medial ridge. The
lesser tuberosity does not extend proxi-
mal to the head of the humerus in
typical phyllostomatids, but in Desmo-
dus ( Fig. 6D ) this process nearly equals
the greater tuberosity in proximal ex-
tension. This modification may be re-
lated to the adducting quality of the M.
suhscapiilaris and the agility of this
vampire in terrestrial locomotion. Other
desmodontines, DiphyUa and Diaemus
(not figured), more closely resemble
typical phyllostomatids in this respect.
The two thick, bipinnate parts of the
M. siibscapularis, one of the powerful
flight muscles, insert on the proximal
surface of the lesser tuberosity. Other
important muscles (M. htissimus dor si,
M. teres major, and parts of the M.
triceps, for example) are associated with
more distal portions of the medial ridge
that forms the lesser tuberosity.

The head of the humerus in the
Phyllostomatidae is continuous with the
greater tuberosity (Fig. 6, C-D), and
there is a well-defined depression (su-
praglenoid fossa) located anterior to
the head and between the two tuberosi-
ties. The supraglenoid tuberosity of the
scapula fits int(j this fossa (Vaughan,
1959:54), and contributes to the solid
locking of the humerus and scapula at
the terminus of the upstroke.

The Emballonuridae and Noctilioni-
dae (Fig. 6, A and B, respectively)

closely resemble each other in structure
of the head of the humerus. In these
two families the greater tuberosity
barely extends beyond the head of the
humerus and the lesser tuberosity, al-
though widely separated from the head,
remains even with or slightly distal to
it. The head of the humerus is slightly
canted and elliptical in shape (viewed
medially). Viewed from the proximal
end (Fig. 6, A-B), the head tapers an-
teriorly between the tuberosities and
merges into the proximal root of the
pectoral ridge. A supraglenoid fossa, as
might be expected, is not present.

The structure of the proximal end
of the humerus in the Natalidae, Furi-
pteridae, and Thyropteridae (Fig. 6, E-
G) more or less resembles that found in
the Vespertilionidae. In all three fami-
lies, the greater tuberosity exceeds the
head of the humerus in proximal exten-
sion and the head is rather small and
rounded. In the Vespertilionidae (Fig.
6H), the head of the humerus is large,
round, and slightly tilted toward the
lesser tuberosity, which is relatively
short in its ventral extension. The
supraglenoid fossa is present in all of
these families. The medial ridge in
Furipterus (Fig. 6F) is prominent,
whereas this ridge gradually tapers into
the shaft of the humerus in representa-
tives of the other three.

The greatest modifications of the
proximal end of the humerus are seen
in the Molossidae ( Fig. 61 ) . The greater
tuberosity reaches its most extreme
proximal extension in this group and
the lesser tuberosity is reduced in size
as in the Vespertilionidae. The head of
the humerus is large and markedly
tilted toward the lesser tuberosity. The
supraglenoid fossa is deep and well de-
fined by the proximal ridges of the
greater and lesser tuberosities, which
merge anterior to the head and form the
proximal root of the pectoral ridge.

As noted by Miller (1907:119), the
greater tuberosity of the mormoopid
genera (Fig. 7) barely extends above

SMITH: SYSTEMATICS OF MORMOOPID BATS

27

Figure 7. Medial, dorsal, and proximal ( above ) views of the proximal end of die humerus of four

mormoopids: A, Pteronotus parnellii; B, P. personatits; C, P. suapurensis; D, Mormoops megalophijlla.

See text for discussion. The line in tha middle of the figure is 4 mm in length.

the head of the humerus and no "second-
ary articulation" exists. In this respect,
mormoopids resemble emballonurids
and, to a certain extent, noctilionids.
The lesser tuberosity is large and ex-
tends ventrad. In the species of Ptero-
notus, there is an elliptical pit that ex-
tends from the distal tip of the lesser
tuberosity toward the proximal root of
the pectoral ridge. The heavy anterior
head of the M. suhscapularis inserts into
this pit. In the genus Mormoops (Fig.
7D ) , the lesser tuberosity extends slight-
ly proximal and the pit for insertion of
the M. suhscapularis is not well defined.

The head of the humerus is ovoid in
Pteronotus (Fig. 7, A-C), and in P.
suapurensis (Fig. 7C) the head is
slightly inclined toward the greater
tuberosity. The supraglenoid fossa is
well developed in all mormoopid genera.
Adaptive significance of the proximal
end of the humerus. — The proximal end
of the humerus tends to be structurallv
more conservative than does the distal
end in that a wide array of modifica-
tions is apparent distally (Figs. 3 and
4), but fewer, more subtle difi^erences
are seen at the proximal end ( Figs. 6
and 7). The most extreme differentia-

28

MISCELLANEOUS PUBLICATION MUSEUM OF NATURAL HISTORY

tion is found in the Molossidae. The
proximal part of the humerus is involved
primarily in the up-and-down motion of
the wing, whereas the distal segment is
involved in manipulations of the distal
elements of the wing. These different
specific functions probably account for
the differential development observed at
each end of the humerus.

The proximal end of the humerus in
the generalized terrestrial mammal has
small tuberosities and the head is the
most prominent structure. As men-
tioned above, the working plane of this
humerus is horizontal, whereas that of
the Chiroptera is vertical to the body.
With the transition to a volant habit,
the tuberosities migrated proximally to
a position on either side of the head. In
the Megachiroptera, the head of the
humerus remains the single most promi-
nent feature of the proximal end, but
the tuberosities are present on either
side. The greater degree of mobility
allowed by the large, rounded head and
small tuberosities is well illustrated by
Kulzer ( 1968 ) . In this paper, which
deals with flight in Eidolon helvtim, one
individual is pictured at the peak of the
upstroke. In this position, the elbow
and wrist joints of each wing nearly
touch over the back. To my knowledge,
no microchiropteran, fossil or living, ex-
hibits such a primitive stage in the
structure of the humerus as seen in the
Megachiroptera. Revilliod (1917a: 172,
fig. 8 ) illustrated a fragment of the right
proximal end of the humerus of Palaeo-
chiropteryx tupaiodon with the greater
tuberosity intact. The head was ap-
parently rounded and both tuberosities
(lesser incompletely preserved) were
prominent, although the greater barely
extended above the head. The structure
of this humerus resembles to some ex-
tent that found in living pteropodids
and even phyllostomatids. However,
judging from the structure of the hu-
merus, PaJaeochiropteryx probably was
a more generalized bat than any living
species.

Among living Microchiroptera the
Emballonuridae, Noctilionidae, and
Mormoopidae could be classed as primi-
tive in that they lack a proximal exten-
sion of the greater tuberosity. The next
stage in the specialization of the proxi-
mal portion of the humerus is probably
best exemplified by the Phyllostomati-
dae, in which the greater tuberosity is
extended proximally and probably con-
tacts the scapula at some late phase in
the wing beat. Next, and only slightly
differentiated from this previous stage,
would be the condition found in the
Natalidae, Furipteridae, Thyropteridae,
and Vespertilionidae. Finally, the ulti-
mate stage of specialization of the prox-
imal end of the humerus is seen in the
Molossidae.

Miller (1907) failed to distinguish
among the differences in the proximal
end of the humerus; he simply referred
to them as "secondary articulations" or
"double articulations." Walton and
Walton (1968:17-18) also studied the
osteology of the wing elements in
phyllostomatids and seem to regard
Miller's use of the two terms "secon-
dary" and "double" as an indication of
two distinct articulations. Inasmuch as
Miller made no distinction, I do not
believe one was intended. I am con-
vinced, after examining bony structures
in many representatives of New World
families, that there is more involved in
the evolution of the shoulder joint than
simply the elongation of the greater
tuberosity and its relationship with the
scapula. The shape of the head of the
humerus and the depth of the supra-
glenoid fossa, for example, also must
play an important role.

With respect to the proximal end of
the humerus, mormoopids are more gen-
eralized than are phyllostomatids. Su-
perficial dissections of Ptcronotus sua-
purensis and Artibeus lituratus indicate
only minor differences in the primary
musculature of the shoulder region. As
described above, there are some special
provisions (oval pit in lesser tuberos-

SMITH: SYSTEMATICS OF MORMOOPID BATS

29

ity) for the insertion of the M. suh-
scapithris in the mormoopids (Fig. 7).
The insertion of this muscle in Artibeus
litiiratiis and Macrotus icaterhousii
(Vaughan, 1959:74) seems to corre-
spond with that of the Mormoopidae.

The head of the humerus is round
and centrally positioned in phyllostoma-
tids, but it is oblong and somewhat
tilted in the Mormoopidae, thereby in-
dicating some degree of specialization
in the latter. The supraglenoid fossa is
present in the phyllostomatids, although
relatively shallow. The modifications
seen in mormoopids probably result in
less freedom of rotation at the shoulder
joint than in phyllostomatids. Freedom
of movement in the latter may be re-
lated to a higher degree of maneuver-
ability, because these bats normally fly
in and around dense vegetation. As
concerns the Mormoopidae, modifica-
tions of the distal end of the humerus
result in increased precision in the wing
beat and consequently a swifter mode
of flight than found in phyllostomatids.
By way of comparison, molossid bats
illustrate the greatest degree of modifi-
cation in the head of the humerus in
that it is markedly oblique in shape, the
supraglenoid fossa is deep, and the
greater tuberosity is extremely elongate
(Fig. 61). It is my opinion that the long
greater tuberosity, deep supraglenoid
fossa, and oblique head of the humerus
in the Molossidae restrict motion of the
humerus to a slightly different plane
than in other bats— a plane that seems
to be somewhat more antero-posteriorly
oriented than the normal up-and-down
wing beat of the others.

In conclusion. Miller's term "sec-
ondary articulation" does not seem gen-
erally applicable to the humero-scapular
relationships. Mormoopids lack any
such "articulation" insofar as the greater
tuberosity is concerned. Phyllostomatids
and other groups with elongated greater
tuberosities probably benefit from some
sort of "weak" relationship with the
scapula. The term "articulation" is

poorly chosen in this context, and
Vaughan (personal communication)
prefers the term "locking device." More
research on the chiropteran shoulder
joint is indicated.

Proximal End of Femur

Vaughan (1959) considered the
anatomy of the hind limbs of Eumops,
Mijotis, and Macrotus and showed that
they play an important role in ffight,
while retaining much of their capability
to function in terrestrial locomotion. As
was the case with the forelimbs, the
hind limbs of bats have been relieved
of much of the weight-bearing respon-
sibility characteristic of terrestrial mam-
mals. The femur has rotated nearly 180
degrees dorsad and the whole limb has
taken on a saurian posture, with the pes
directed laterally and the knee joint
dorso-laterally. The head of the femur
in most terrestrial mammals projects
mediad and the greater trochanter ex-
tends proximally beyond the knob-like
articular process. The lesser trochanter
occupies a position well below (distal)
the level of either the head or the
greater trochanter. In the Chiroptera,
the head of the femur is the most promi-
nent feature of the proximal end of this
bone. It is rounded and directed slightly
mediad, and apparently (due to this
superior position) is capable of a great
deal of rotation and freedom of move-
ment.

The variation in structure of the
proximal end of the femur, like that of
the proximal end of the humerus, seems
to be conservative. Major features of
this portion of the femur are the size of
the greater and lesser trochanters and
the position of the head of the femur.
With the exception of the mormoopids
(Fig. 8, F-K), all bats examined in my
study have two well-developed trochan-
ters that are located on the lateral sur-
face of the femur and are about equal
in size. The head lies between the tro-
chanters, along the longitudinal axis of

30

MISCELLANEOUS PUBLICATION MUSEUM OF NATURAL HISTORY

Figure 8. Lateral \'iew of the proximal end of the femur of representative New World bats: A, Rhyn-
chonycteris naw; B, Saccoptenjx hilineata; C, Balantiopteryx plicata; D, Diclidunis virgo; E. Noctilio
labialis; F, Pteronotus parnellii; G, P. personatus; H, P. fuliginosus; I, P. davyi; J, P. suapurensis; K,
Mormoops megalophylla; L, Micronycteris sylvestris; M, Glossopliaga soricina; N, CawIIia perspicillata;
O, Sturnira lilitim; P, Artihens tolteciis; Q, Desmodiis rotundus; R, Natalus stramincus; S, Furipterus
horrens; T, Thyroptera tricolor; U, Myotis grisescens; V, Lasiurus borealis; W, Molossiis ynolossus.
The labels on A refer to the following stiaictures mentioned in text: 1, head; 2, lesser trochanter; 3,
greater trochanter. See text for discussion of the proximal end of the femur. The line above B and C

is 4 mm in length.

the shaft, and extends antero-mcdially
(Fig. 8).

The overall structure of the femur
in Desmodiis (Fig. 8Q) is noteworthy
because the bone is laterally compressed
and has numerous grooves and ridges
(as described by Miller, 1907:176).

The lesser and greater trochanters are
widely spread and almost equal the
head in size and proximal extension.
These adaptations apparently relate to
the agility of these bats in terrestrial
locomotion.

Other interesting modifications are

SMITH: SYSTEMATICS OF MORMOOPID BATS

31

seen in the genera Natalus and Fiiri-
pterus (Fig. 8, R and S) in which both
trochanters are somewhat reduced in
size. In the vespertiHonid genus Lasi-
unis (Fig. 8V), the head is spherical
and the greater trochanter is shghtly
smaller than the lesser trochanter. The
molossid bats (Fig. 8W) exhibit a char-
acteristic hook-like projection on the
greater trochanter.

The functional significance of the
marked differences in the size and shape
of the two trochanters observed be-
tween mormoopids and other bats is not
readily apparent. Muscle groups that
are associated with the trochanters play
important roles in bracing the hind limb
against the lateral pull of the wing mem-
brane during its cycle and may become
involved in the forward stride in ter-
restrial locomotion. Other muscles that
originate from the region of the tro-
chanters also control the posture of the
hind limb in flight.

Tragus

The ears of the Microchiroptera are
complex and usually appear as cone-like
horns, directed anteriorly. The ears of
the Megachiroptera, on the other hand,
resemble those of terrestrial mammals.
In addition, the interior of the pinna
and the auricular cavity of microchirop-
teran ears may be furnished with an
elaborate array of folds and flaps, the
most conspicuous of which is the tragus.
Although greatly reduced in size in the
molossid bats, the tragus is present in
all New World families. It is absent in
the Old World family Rhinolophidae
( including Hipposideridae ) .

The tragus is attached just anterior
to the external auditory meatus and
seems to be situated directly in the path
of sound waves that enter the auricular
cavity. Griffin (1958:129) noted that
the tragus probably does not interfere
appreciably with the passage of sound
waves to the auditory meatus. Although
there is no evidence to support or refute
this assumption, it seems highly prob-

able to me that the size, shape, and
pubescence of the tragus alter, in some
way, the character of these sound waves
as they travel through the external ear.

Dobson (1878), Flower and Ly-
dekker (1891), and Miller (1907), among
others, commented briefly on the struc-
ture of the tragus as it pertained to their
taxonomic arrangements. Numerous re-
cent authors, for example Davis (1965:
230) and Peterson (1968:6), have illus-
trated and described the tragus with
reference to differences between species,
and I have found it to be a highly re-
liable character at the familial, generic,
and specific levels of classification. Dif-
ferences at the specific level are mani-
fest in proportion and size, whereas
generic and familial affinities are usually
reflected in variations in shape.

The tragus of emballonurids is typ-
ically a rather small flap of skin that
projects laterally into the auricular
space. In most cases, it is broad and
squared as in Saccopteryx (Fig. 9B),
but may be erect and bluntly spatulate
as in Rhijnchonycteris and Peropteryx
(Fig. 9, A and C). The lateral edge
(that which projects into the auricular
cavity) may be simple, notched, or ser-
rated, and may be naked or clothed with
short bristle-like hairs.

In the Noctilionidae (Fig. 9D) the
tragus is a slender, pinnately-lobed
structure with one finger-like projection
on the cranial edge (directed toward
the side of the head) and three or four
on the lateral edge. Short bristle-like
hairs cover most of tragus, which has a
small secondary lobe near the base.

Typically, the tragus of phyllostoma-
tids is divided into a main section,
usually lanceolate in shape (Fig. 9, F-
H), and a secondary oblique lobe near
the base on the lateral edge that projects
ventrad and posteriorly. Generally, the
main support for the tragus is a ridge,
which may be heavy to weak and ex-
tends along the longitudinal axis of the
main section on the cranial side. This
ridge causes the central portion of the

32

MISCELLANEOUS PUBLICATION MUSEUM OF NATURAL HISTORY

Fic;uRE 9. Tragus of some representative New World bats: A,
Rlitjnchontjcteris naso; B, Saccopteryx hilineata; C, Peroptenjx
macrotis; D, Noctilio labialis; E, Lonchorhina aurita; F, PhtjUostoinus
hastatus; G, GlossopJiaga longirostris; H, Lionijcteris spurrelli; I, Uro-
denna hilohatiini; J, Artiheus lituratus; K, CaroUia perspicillata; L,
Sttirnim liliiim; M, Dcsmochts lotiindus; N, Ametrida centurio; O,
Phijllonijcteris poetji; P, Natahis straminetts; Q, Eptesicus fuscus. See
text for discussion of tragus. Line below A is 5 mm in length.

SMITH: SYSTEMATICS OF MORMOOPID BATS

33

Figure 10. The tragus of mormoopids: A, Pteronotus parnellii; B, P.

personahis; C, P. macleayii; D, P. fuliginosiis; E, P. daviji; F, P.

suapurensis; G, Monnoops megalophijlla; H, M. blainvillii. See text

for discussion of tragus. Line below G is 5 mm in length.

tragus to be concave, either deep or
shallow, depending on the nature of the
ridge. The extremely elongated tragus
of Lonchorhina aurita (Fig. 9E) is
strongly convex, rather than concave,
and the supporting ridge is absent. The
lateral edge of the phyllostomatid tragus
may be simple as in the Glossophaginae
(Fig. 9, G and H) or notched, serrated,
or both (see Uroderma, Artibeus, Carol-
lia, and Sturnira in Fig. 9, I-L). Pubes-
cence is also variable from completely
nude as in Uroderma and Ametrida
(Fig. 9, I and N) to exceedingly hairy
as in Desmodus (Fig. 9M).

The tragus of members of the Nata-
lidae (Fig. 9P) is curled into a curious
cone-like structure, from which many
long hairs project. Disc-winged bats
(family Thyropteridae, not illustrated)
have a tragus that is essentially similar
to that of natalids, except that it lacks
the long hairs. From the description
and illustration given by Husson (1962:
200-201), the tragus of Furipterm hor-

rens (family Furipteridae ) also seems
to resemble that of natalids.

The least variation in shape was ob-
served in the Vespertilionidae. In this
group the tragus is usually spatulate ( as
in Eptesicus fuscus. Fig. 9Q), but may
be acuminate. Variation observed in
this group relates primarily to size. In
almost every case, the lateral edge of
the tragus is simple or only slightly
crenulate, and pubescence, if present,
usually is restricted to the basal portion
of the structure.

As mentioned above, the tragus of
molossids (not illustrated) is greatly
reduced in size. In most cases it is little
more than an inconspicuous finger-like
projection as typified by the genus
Molossus, but may be relatively large
and squared as in the genus Tadarida.

Generally speaking, the tragus of
the above-mentioned New World fami-
lies is little more than a simple flap of
skin that projects into the auricular
cavity. This is also true for those Old

34

MISCELLANEOUS PUBLICATION MUSEUM OF NATURAL HISTORY

World families that possess a tragus,
with the exception of the Myzopodidae,
in which the structure is mushroom-
shaped (Walker et al, 1964:333).

By comparison, the tragus of the
Mormoopidae is considerably different.
In this group, there is a rather interest-
ing sequential transition from the sim-
plest, in Pteronotus parnellii (Fig. lOA),
to the most complex, in Mormoops
megaJophijUa and M. blainviUii (Fig.
10, G and H). This transition is from a
seemingly simple lanceolate tragus to
one with a secondary fold of skin that
lies at a right angle to the main longi-
tudinal axis of the structure. This sec-
ondary fold is barely more than a
pocket-like structure in the cranial edge
of the tragus in Pteronotus parnellii
(Fig. lOA). In Pteronotus macleaijii,
P. fuliginosus, P. daviji, and P. suapur-
ensis (Fig. 10, C-F), the fold is well
developed. In Pteronotus personatus,
the fold is not so markedly developed,
but appears as a cranially oriented shelf
(Fig. lOB). Finally, in the genus Mor-
moops, the fold is greatly increased in
size and forms the major part of the
tragus. Below the secondary fold, the
cranial edge of the tragus is swollen
and forms a heavy fleshy ridge. The
pubescence is variable in this group,
but long filamentous hairs are common
on the blunt end of the tragus around
the edge of the secondary fold, and
along the cranial edge below this fold.

Correlative Information from
Other Sources

From the foregoing accounts, it is
apparent that the family Mormoopidae
stand.s apart in a number of morpho-
logical characteristics from the Phyllo-
stomatidae, with which it has been
grouped (as a subfamily) in the past.
Several noteworthy examples of other
differences have appeared in the litera-
ture, and these are summarized here as
additional evidence of the distinctness
of the Mormoopidae in comparison with

phyllostomatids and other New World
chiropteran families.

Echolocation. — Various aspects of
the acoustic orientation in bats have
occupied the interest of mammalogists
for some time and a wealth of informa-
tion has accumulated. Most investiga-
tions in this regard have been descrip-
tive and have been based on Holarctic
bats, principally of the family Vesper-
tilionidae. More recently, echolocation
of Neotropical bats has been studied,
and some attempts now are being made
to understand better the functional and
behavioral aspects of echolocation
(Grummon and Novick, 1963; Novick,
1963b, 1965; and Novick and Vaisnys,
1964). Several Neotropical species of
the Phyllostomatidae (here including
the mormoopid genera) and Desmoclus
were first examined by Griffin and No-
vick ( 1955 ) . More recently, with the
advent of better high-frequency sound-
detecting equipment. Griffin (1958),
Griffin and Grinnell (1958), Grinnell
and Griffin (1958), and Novick (1958,
1962, 1963a) have contributed greatly
to our knowledge of echolocating sys-
tems in Neotropical bats.

The high-frequency sound emissions
of several mormoopid species — Pterono-
tus parnellii, personatus, and suapuren-
sis — were recorded and analyzed by
Griffin and Novick (1955). It is note-
worthy that these authors found that
mormoopids emitted high-frequency
sounds that more closely resembled
those emitted by emballonurids than
those of "other" phyllostomatids. The
pulses were elicited orally by mormo-
opids rather than nasally as in phyllo-
stomatids, were long and relatively in-
tense, and lacked the progressive fre-
quency modulation observed in vesper-
tilionids, molossids, and noctilionids.
The range of the predominate frequency
was observed between 20 and 100 kilo-
cycles. By comparison, phyllostomatids
emitted shorter pulses of intermediate
intensity with a rather simple frequency
spectrum (predominant frequency 12 to

SMITH: SYSTEMATICS OF MORMOOPID BATS

35

50 kilocycles ) and harmonics of variable
amplitudes. These same results' were
again obtained by Novick (1958).

Novick ( 1963a ) re-examined the
acoustic orientation pulses of several
phyllostomatids ( including desmodon-
tines) and the mormoopid species
Pteronotus parnellii, per.wnatus, and
davyi. His findings were generally con-
sistent with those obtained by Griffin
and Novick ( 1955 ) , who had used some-
what less sophisticated equipment. In
this paper, Novick (1963a:47) noted
that parneUii was strikingly different
from the other mormoopid genera
studied, as well as phyllostomatids in
general. The particular features of the
pulses that seemed most significant to
him were their unusually long duration
and slight degree of frequency modula-
tion. On the other hand, personatus and
davyi elicited similar pulses of shorter
duration and with gradual frequency
modulation, dropping from beginning to
end. All three species seemed to have
two harmonics and personatus frequent-
ly produced a third overtone, which
Novick (1963a: 48) thought might be
significant.

The members of the family Mormo-
opidae appear to differ markedly from
phyllostomatids judging from the re-
sults of studies concerning chiropteran
echolocation thus far conducted. Dr.
Novick (personal communication) stated
that he had doubted the relationship of
the "Chilonycterinae" to the Phyllosto-
matidae for some time based on their
behavior in locomotion, feeding in cap-
tivity, and roosting, as well as the
marked differences in frequency pat-
tern, intensity, and other features of
their sonar design. Information con-
cerning the acoustic orientation of Mor-
moops is not yet available.

Hair structure. — The scales of chi-
ropteran hair were examined micro-
scopically as early as 1875 by J. D.
Macdonald, the results being published
as a footnote in Dobson (1875:355-356).
More recently, Benedict (1957) exam-

ined numerous representative genera of
bats in this regard. She was concerned
not only with scalation, but also filament
length, distribution of pigmentation,
structure of the medulla, and differen-
tiation of the pelage into "overhair" and
"underhair." Benedict noted that "chilo-
nycterines" have scales arranged in a
simple annular fashion, whereas phyllo-
stomatids have scales arranged in an
alternating fashion, but she found no
other peculiarities separating the two
groups.

I have examined, in a cursory man-
ner, whole-mount slides of hair from
some representatives of New World
families. The mormoopids (except
Mormoops) have the coronal scales ar-
ranged in an annular fashion. Perhaps
a better description would state that
these scales approach the unequal has-
tate type as described by Benedict
(1957:290). Scales of Mormoops hair,
by comparison, are clearly unequal has-
tate and have the terminal edge mark-
edly serrated. Although no marked
differences in hair structure distinguish
mormoopids from phyllostomatids, some
interesting features do exist that merit
further investigation.

Studies of karyotypes. — Some useful
information concerning the phylogenetic
relationships among the New World
Chiroptera has been obtained recently
from studies of karyotypes (see espe-
cially Baker and Patton, 1967, and
Baker, 1967, 1970).

Baker ( 1967 ) examined many phyllo-
stomatid genera representing nearly
every subfamily except the Phyllonyc-
terinae and Desmondontinae; later, For-
man et al. ( 1968 ) found that karyotype
data generally supported the inclusion
of the vampires as a subfamily of the
Phyllostomatidae. In this paper. Baker
examined the karyotypes of Pteronotus
parneUii, personatus, and davyi and
found that all three species were indis-
tinguishable. The karyotype of Mormo-
ops megaJophyUa was first reported bv
Baker and Hsu (1970).

36

MISCELLANEOUS PUBLICATION MUSEUM OF NATURAL HISTORY

The mormoopids examined have a
rather high number of metacentric pairs
(11) as compared with the phyllosto-
matids examined (one to eight, mean
six). Mormoopids also have one pair of
submetacentrics (none to seven in phyl-
lostomatids ) , no subtelocentric pairs
(none to six in phyllostomatids), and
six acrocentric pairs (the majority of
phyllostomatids examined had none).
The diploid number in the four mor-
moopid species examined was 38 with a
fundamental number (total number of
chromosome arms) of 60 (62 in Mor-
moops megalophylla). In phyllostoma-
tids, the diploid number averaged 29
( 16-46 ) and the fundamental number
averaged 52 (24-68).

From this information, Forman et al.
( 1968 ) concluded that the "chilonyc-
terines" were unique karyotypically
among the Phyllostomatidae and that
they were sufficiently distinct to merit
at least subfamilial status.

Histological aspects of facial glands.
— ^The histological aspects of the glands
of the facial regions of North American
bats have been reported on by Werner
et al. (1950), Dalquest et al. (1952),
Werner and Dalquest (1952), and Dal-
quest and Werner (1954). The last
paper is mostly a summary of previous
work and the systematic implications of
this research. Numerous genera, repre-
senting all New World families except
the Noctilionidae, Furipteridae, and
Thyropteridae, were examined. Dal-
quest and Werner (1954) found that
each of the families considered could be
adequately characterized and identified
solely on the basis of the histology of
the facial region. Their results conform
closely to the existing taxonomic ar-
rangement of bats, excepting that they
recognized the mormoopids as a sep-
arate family which they referred to as
the "Chilonycteridae." Their reasons for
this action, however, were not made
exactly clear.

The mormoopid genera examined by
Dalquest and Werner (1954) were

Pteronottis parnellii, P. davyi, and Mor-
moops megalophylla, all of which were
found to be histologically similar with
regard to the facial region, except that
Mormoops megalophylla was observed
to have submaxillary glands composed
of serous cells rather than of the mucous
type commonly found in other bats.
Molossids also were found to have se-
rous cells in the submaxillary glands ( as
well as in the sublingual glands). None
of the other New World families ex-
hibited serous sublingual glands.

These authors also found that the
New World families could be grouped
by the presence or absence of the sudor-
iferous glands in the facial region.
These glands were present in the Em-
ballonuridae, Vespertilionidae, and Na-
talidae. Sudoriferous glands were ab-
sent from the facial area in the "Chilo-
nycteridae," Phyllostomatidae, Desmo-
dontidae, and Molossidae. The presence
or absence of these glands was thought
to be conservative and consistent among
New World groups. Families that pos-
sessed such glands were presumed to be
primitive and more closely related to
one another than to groups that did not
have these glands. Consequently, Dal-
quest and Werner suggested a phylo-
genetic arrangement that reflected these
two groupings.

Selected ectoparasites. — Although
several groups of ectoparasites are found
on bats, only a few groups parasitize
bats exclusively. Three of these so-
called "obligate" parasite groups are the
Spinturnicidae ( Acarina ) , Nycteribiidae
(Diptera), and Streblidae (Diptera).
In the past, mammalogists have been re-
luctant to accept phylogenetic informa-
tion based on host-parasite relationships,
particularly those involving ectopara-
sites. Part of this reluctance evidently
stems from the fact that few careful col-
lections of ectoparasites have been
made. In the Chiroptera, ecological fac-
tors are of paramount concern, espe-
cially when two or more host species
come into close physical proximity at

SMITH: SYSTEMATICS OF MORMOOPID BATS

37

feeding and roosting sites. This, coupled
with the fact that some of the ectopara-
sites of bats pass stages of tlieir Hfe
history away from the host, has led to
caution concerning assumptions of host
specificity and phylogenetic implications
based on taxonomic concordance of
parasites and their hosts. Host speci-
ficity is discussed in some detail by
Wenzel and Tipton (1967).

At least one group of chiropteran
ectoparasites seems to be relatively host-
specific — the wing mites of the family
Spinturnicidae (Acarina, Mesostigmata).
The life history of these mites has been
modified by the reduction of nymphal
stages and the development of ovovivi-
parity (Baer, 1952, and Rudnick, 1960).
These adaptations make this group a
reasonable one to turn to for some in-
dication of phylogeny within the
Chiroptera.

According to Machado- Allison (1967:
224) mites of the family Spinturnicidae
are intimately involved with the Chi-
ropteran family Phyllostomatidae, being
found on all except the endemic Antil-
lean subfamily Phyllonycterinae. Earlier,
Machado- Allison (1967:225) suggested
the inclusion of the vampire bats in the
Phyllostomatidae based on the occur-
rence of the spinturnicid species Peri-
glischurus herreroi on Desmodus ro-
tundus. Other species of this genus have
been reported from Noctilio leporinus
( Noctilionidae ) and Natalus stramineus
(Natalidae) by Furman (1966). Re-
cently, Machado- Allison (1965b) re-
viewed the species of the spinturnicid
genus Periglischurus Kolenati. The
spinturnicids that he encountered on
bats of the species Pteronotus parnellii
were so distinct morphologically that he
( 1965a ) named them as a new genus
and species, Cameronieta thomasi.
Machado-Allison (personal communica-
tion) also thinks that the species Peri-
glischurus strandtmanni, known from
Mormoops megalophylla, is referable to
his genus Cameronieta.

In his review of the spinturnicid

mites of Panama, Furman (1966:166)
placed the genus Cameronieta Machado-
Allison in synonymy under Periglischu-
rus, based on a close resemblance be-
tween C. thomasi and P. elongatus Fur-
man, 1966 (known from Pteronotus
parnellii and P. suapurensis) . Furman
(1966) regarded the species thomasi
and elongatus as distinct, pending fur-
ther investigation to demonstrate his
hypothesis for the existence of hetero-
morphic females; he considered both
species close relatives of P. strandt-
manni. The resolution of the Cameron-
ieta-Periglischurus problem remains in
the hands of the acarologists. The point
to be made is that apparently the spin-
turnicid mites found on mormoopid bats
are, as a group, distinct from those
found on phyllostomatids, implying dis-
tinctiveness in the bats as well.

The batflies of the family Streblidae
are not as host specific as are spinturni-
cids, but they also suggest some inter-
esting relationships. Wenzel et al.
( 1966 ) , reporting on the streblids from
Panama, indicated (p. 652) that repre-
sentatives of the strebhd subfamily
Nycterophiliinae (Trichobius and Nijc-
terophilia) occur primarily on the
"Chilonycterinae" and Natalidae, but
also are known from Plataliiw (Glosso-
phaginae) and Macrotus ( Phyllostomi-
nae). Wenzel et al. (1966) continued
by stating: "The host distribution of
these groups of Trichobius and Ntjctero-
philia support the view of Dalquest and
Werner (1954) and de la Torre (1961)
that the Chilonycterinae should be re-
garded as a separate family of Vesper-
tilionoidea rather than as a subfamily of
the Phyllostomatidae."

Brain size. — Recently, Findley
( 1969), in a paper concerned with brain
size of various bats, recorded some
quantitative and qualitative differences
between mormoopids and phyllostoma-
tids. Relative brain size was found to
be smaller in the former. Additionally,
he noted that mormoopids are similar

38

MISCELLANEOUS PUBLICATION MUSEUM OF NATURAL HISTORY

to molossids and lasiurines in that the
cerebral hemispheres are relatively
small, the cerebelli relatively large, and
that the midbrain is visil)le between
these two areas. The cerebral hemi-
spheres apparently are large in many
phyllostomatids. As noted by Findley,
brain size alone probably indicates little
concerning the biology and relationships
of bats; enlargement of particular cere-
bral regions, however, may have taxo-
nomic significance correlated with other
morphological features.

Immunologic relations. — In a recent
paper Gerber and Leone ( 1971 ) re-
ported on immunologic comparisons of
the sera of certain phyllostomatid bats.
On the basis of turbidimetric tests, they
found Pteronotus suapurensis to differ
appreciably from other species of phyllo-
stomatids (Glossophaga soricina, Carol-
lia perspicillata, Artibeus jamaicensis,
and Desmoclus rotundus). This distinct-
ness appeared to be almost as great as
that demonstrated by representatives of
the Vespertilionidae and Molossidae.

PHYLOGENETIC RELATIONSHIPS OF THE MORMOOPIDAE

The student of chiropteran evolution
is faced with one of the largest and most
diverse groups of mammals and yet one
of the poorest known in terms of past
history. Bat fossils are relatively com-
mon in some fissure faunas of Europe,
but have not been reported in detail.
Some of these remains (all representing
fully evolved bats) have been recovered
from deposits as far back as early
Eocene. In the New World, only a few
supposedly ancient chiropteran fossils
were known until recently (the genera
Picrodus and Zamjcteris from the North
American Paleocene). Simpson (1945:
17(S, 180) regarded these as "possible"
phyllostomatid bats, but assigned them
to Insectivora, incertae sedis. McKenna
(1967:608) placed these two genera in
the Omomyoidea, incertae sedis.

Jepson (1966, 1970) described Icaro-
nycteris index from the early Eocene
(Green River formation) of Wyoming,
and Savage (1951) described Notonyc-
teris magdalenensis from the late Mio-
cene of the Neotropical La Venta Fauna
of Colombia. Although Notonycteris is
not as well preserved as Icaronycteris,
it is clearly an early representative of
the Neotropical family Phyllostomati-
dae. The majority of the remaining
fossil bats in the New World are of
much later age, and most are members
of Recent (or closely related) species.

As mentioned earlier, there is a rea-

sonably good array of fossil chiropteran
material from the early Cenozoic era of
the Old World. Notable among these
are Archaeonycteris Revilliod and Pa-
laeochiropteryx Revilliod from the lig-
nite beds (Messel, early Eocene) near
Darmstadt, Austria. These are the earli-
est known bats in the Old World assem-
blage. Members of the Rhinolophidae
(Pseudorhinolophus) are found as early
as middle Eocene, and emballonurids
(Vespertiliaviis), vespertilionids (Nyc-
terohius), and molossids (fragment of
proximal end of humerus, genus inde-
terminable) occur in the Quercy fauna
( Eocene-Oligocene border ) . These
early bats are not noticeably "primitive"
and resemble modern species in many
features. Detailed accounts of these Old
World fossil bats were given by Revil-
liod (1917a, 1917b, 1920, 1922).

Schlosser (1911:72) described Pro-
vampyrus orientalis based on a right
hinnerus from the Fayum beds (early
Oligocene) of Egypt. He assigned this
fossil bat to the family Phyllostomatidae
because this humerus seemed to resem-
ble that of "Vampyrus auritus" and
"Stenodernia undatum." Schlosser (1911:
pi. 13) illustrated this bone, which he
reported to be 49 mm in length. The
proximal end clearly is not that of a
phyllostomatid, the greater tuberosity
being long and extending well beyond
the rounded head. Little can be deter-

SMITH: SYSTEMATICS OF MORMOOPID BATS

39

mined from the distal end as to struc-
ture of the medial process, because it is
obscured. In my opinion, this species
is best referred tentatively to the Mega-
dermatidae.

Because of the meager fossil record,
mammalogists who have studied the
evolution of bats have relied on resem-
blances among representatives of living
groups. The task of constructing a
workable phylogeny is further compli-
cated by the fact that little is known as
to how the majority of the features used
to distinguish groups of bats relate to
chiropteran evolution. Much of our
current conception of chiropteran phy-
logeny dates back to Dobson (1875),
Winge (1892), and Miller (1907). The
arrangements of these workers primarily
reflect the number and kinds of teeth in
the dental arcade and the configuration
of the cusps on the molar teeth.

The relationships of the Mormoopi-
dae are not readily apparent; mormo-
opids having been shuffled between the
Noctilionidae and Phyllostomatidae.
Dobson (1875:353) thought that Noc-
tilio connected the Emballonuridae with
the Phyllostomatidae through the "Lobo-
stominae" (= Mormoopidae ) . In order
to clarify the relationships within the
Chiroptera, Simpson (1945) grouped
Miller's (1907) families into superfa-
milial units based on Winge's (1892)
families. As a result, the Noctilionidae
was grouped along with the Rhinopo-
matidae and Emballonuridae in the
superfamily Emballonuroidea. The mor-
moopids were carried as a subfamily of
the Phyllostomatidae into the Phyllo-
stomatoidea along with the vampire
bats. All of the remaining New World
families were placed in the Vesper-
tilionoidea.

Judging by the proximal and distal
portions of the humerus, the proximal
end of the femur, and the tragus, the
mormoopids are quite distinct from the
Phyllostomatidae. Using these same cri-
teria, the Mormoopidae bears little re-
semblance to the Noctilionidae or Em-

ballonuridae. Yet in some other respects
the Mormoopidae, Noctilionidae, and
Phyllostomatidae resemble one another
more than they resemble any other New
World group. The premaxillary bones
are complete and solidly fused to the
maxillaiy and palatine bones; the three
families are probably the only New
World bats in which an os penis is
characteristically absent, and the upper
incisors usually are strong, heavy teeth,
although these are secondarily modified
or lost in some phyllostomatids. Unlike
the New World emballonurids, there is
a marked sella turcica (cup-like depres-
sion in the floor of the braincase that
holds the hypophysis) behind which
there is a prominent clinoid plate (both
are minimally developed, but present,
in the Noctilionidae). In the Mormo-
opidae and Phyllostomatidae, the pos-
terior clinoid processes (viewed through
the foramen magnum) usually are long
and have small terminal knobs. The
foramina ovale are similarly situated in
the three families; the foramen may be
hidden beneath an anteriorly projecting
ridge of the posterior sphenoid ( alisphe-
noid and basisphenoid complex) or
tightly squeezed against the auditory
bulla in other groups of bats. For the
most part, the basicranial region of these
bats is relatively simple, lacking any
indication of the extremely deep, some-
times partitioned, basisphenoidal pits
found in emballonurids, the function of
which remains unknown.

In all three families, there is a curi-
ous array of wart-like bumps and ridges
on the central portion of the lower lip
and chin regions. Labial ornamentation
is rather simple in Noctilio, consisting
of a central tubercular pad on the lower
lip surrounded by a secondary trans-
verse ridge that communicates with the
lower lip via two connecting ridges. In
the Mormoopidae, this region is much
more complicated. The central tubercu-
lar pad of Pteronotus is more or less
unchanged, but the lower lip is ex-
panded and plate-like with many con-

40

MISCELLANEOUS PUBLICATION MUSEUM OF NATURAL HISTORY

centric rows of small bumps that sm*-
roimd the central tubercle. This region
in Mormoops is extremely complicated
in that the central tubercle is enlarged
with a flat, shield-like plate, covered
with many wart-like bumps, that extend
anteriorly from the tubercle. Beneath
this plate, the notched lower lip unfurls;
as in Noctilio, there is a secondary trans-
verse ridge below the lower lip. The lip
pad of phyllostomatids is usually a tri-
angular pad that may be simple or may
possess several concentric rows of wart-
like bumps. These structures are absent
in all other New World bats.

Despite the above-mentioned simi-
larities between the mormoopids, noc-
tilionids, and phyllostomatids, there are
a number of other features to be con-
sidered. Both Noctilionidae and Mor-
moopidae, for example, have character-
istically long calcars that support the
posterior half of the uropatagium and
that extend considerably farther caudad
than do the feet. The posterior portion
of the uropatagium is usually flexed for-
ward beneath the proximal part; this
characteristic posture is never observed
in phyllostomatids. Unlike phyllosto-
matids, several of the distal caudal verte-
brae protrude as a "free tail" from the
dorsal surface of the uropatagium. (In
the Phyllostomatidae the "free" segment
of the tail, if present, usually involves
only the last or last two caudal verte-
brae.) In the above respects, the Noc-
tilionidae and Mormoopidae resemble
emballonurids.

The wing membrane attaches rela-
tively high on the side of the body in
both mormoopids and noctilionids.
Some members of the Phyllostomatinae,
probably the most primitive or general-
ized subfamily of the Phyllostomatidae,
also share this high wing attachment.
Generally, the morphology of the phal-
lus of mormoopids and noctilionids
is similar, in contrast to that of phyllo-
stomatids. In the latter family, the glans
is usually a prominent and knob-like
structure, whereas in members of the

other two families the glans is fan-
shaped with many intricate longitudinal
folds and grooves. Also, in these two
the clitoris of the female has an acute,
and not infrequently long, pendant tip.

As regards the dentition of the three
families, it is difficult to make an ac-
curate evaluation. The dental arcade of
most phyllostomatids has been modified
and reduced. The characteristic dilamb-
dodont molar is retained, more or less
unchanged, in the subfamily Phyllo-
stomatinae (extremely modified in other
phyllostomatids ) , Monnoopidae, and
Noctilionidae.

Judgements as to the phylogeny and
evolution of major groups are therefore
largely intuitive. It seems to have been
generally agreed that the majority of
the Microchiroptera originated in the
Old World; New World groups presum-
ably evolved from Old World migrants.
The chiropteran fauna of the New
World comprises several endemic (and
possibly autochthonous) groups, among
which are the Mormoopidae, Noctilioni-
dae, and Phyllostomatidae. In addition,
the New World and the Old World
share several cosmopolitan families —
Emballonuridae, Vespertilionidae, and
Molossidae. The fact that vespertilio-
nids and molossids have their greatest
diversity in the Old World and the fact
that emballonurids are represented there
by a fossil genus of Eocene or Oligocene
age seems to suggest that the center of
origin and dispersal of these families
also was there. This assumption, coupled
with the generally accepted idea that
microchiropteran bats evolved from an
emballonuroid-like ancestor, suggests
that mormoopids, noctilionids, and
phyllostomatids were derived from an
Old World, emballonurid migrant(s).

The fossil record of bats, although
scanty, does yield several key points re-
garding the progression of chiropteran
evolution. By Eocene times, there was
a world-wide group of fully evolved
bats, represented in the New World by
Icaronycteris index and in the Old

SMITH: SYSTEMATICS OF MORMOOPID BATS

41

World by Archaeomjcteris and Palaeo-
chiropteryx. These early bats were
indeed primitive, not only in num-
ber and kinds of teeth and dilambdo-
dont molars, but also in many features
of the posteranial skeleton. The primi-
tive nature of the humerus of Archaeo-
nycteris and Palaeochiropteryx is illus-
trated by the broad medial process,
simple distal articular surface, short
(not extended proximally) greater tu-
berosity, and rounded head. ( These fea-
tures cannot be readily observed in the
illustrations of /. index — see Jepsen
1966:1334.) In addition, the pelvic
girdle, femur, and long tail contribute
to the primitive appearance of these
ancient bats. Do they represent a world-
wide grade in the evolution of chirop-
terans or merely several vmsuccessful
lines from the earliest ancestors of
chiropterans?

In consideration of the phylogeny of
New World bats, especially mormo-
opids, noctilionids, and phyllostomatids,
the above propositions are of paramount
importance. The discovery of Notonyc-
teris indicates that phyllostomatids were
well established in an evolutionary sense
by the Miocene. It is also apparent that
the oldest emballonurids from the Quer-
cy fauna ( Eocene-Oligocene of the Old
World) were similar to modern species
of that group. The time involved in the
differentiation of the phyllostomatids

from the emballonurid-like ancestor (if
this were the case) seems to shift un-
duly the whole evolutionary sequence
of bats as well as placental mammals to
a much earlier, yet vindocumented age.
It seems reasonable to suppose that a
"paleochiropteran" grade existed in
both the Old World and the New World
during middle to late Paleocene times.
The importance of this group is that it
could provide a plausible ancestor, both
on the basis of age and generalized
structure, for the Noctilionidae, Mormo-
opidae, and Phyllostomatidae, assuming
that these three families are intimately
related.

For the reasons outlined above, I
think that the Mormoopidae, Noctilioni-
dae, and Phyllostomatidae represent
three distinct families that probably had
a common ancestral line. It seems pre-
mature at this time to place all three in
the superfamily Phyllostomatoidea, al-
though further investigation may show
this to be the proper arrangement. Fur-
thermore, they may represent a truly
autochthonous lineage that evolved
from a paleochiropteran grade in the
New World, although there is no clear
evidence for or against this hypothesis.
In addition, the Mormoopidae seems to
be more closely related to noctilionids
than to phyllostomatids, and in many
respects the Noctilionidae appears to be
the least specialized of the three families.

SYSTEiMATIC ACCOUNTS

Family MoRMOOProAE Saussure

Mormoopsins Saussure, 1860, Rev. Mag. Zool.,

ser. 2, 12:286, July.
Mormopes Peters, 1865, Monatsber. k. preuss.

Akad. Wiss., Berlin, p. 257.
Mormopsina Gray, 1866, Ann. Mag. Nat. Hist.,

ser. 3, 17:93, February.
Phtjllodiana Gray, 1866, Ann. Mag. Nat. Hist.,

ser. 3, 17:93, February.
Lobostominac Dobson, 1875, Ann. Mag. Nat.

Hist., ser. 4, 16:350, November.
Chilonijcteninae Flower and Lydekker, 1892,

An introducton to the study of Mammals

living and extinct, p. 672.
Mormopinae Trouessart, 1897, Catalogus mam-

malium tam viventium quam fossilium,
1:150.
Monnoopinac Relin, 1902, Proc. Acad. Nat.
Sci. Philadelpliia, 54:162, June 11.

Type genus. — Mormoops blainvillii
Leach, 1821.

Geographic distribution. — New
World, from southern Arizona and
southern Texas south through Mexico
(including Baja California and the
Yucatan Peninsula), British Honduras,
Central America, and South America
(including Trinidad) to Mato Grosso,
Brasil; generally restricted to tropical

42

MISCELLANEOUS PUBLICATION MUSEUM OF NATURAL HISTORY

'"^^i^

Figure 11. LalMo-nasal region of six mor-
nioopids: A, Pteronotus parnellii; B, P. per-
sonatus; C, P. davyi; D, P. macleayii; E, P.
fuliginosus; F, Mormoops megalophtjUa. See
text for discussion.

habitats below 3000 meters. Endemic
species or races also are found on the
Greater Antilles (Cuba, Jamaica, His-
paniola, and Puerto Rico) and in the
Dutch West Indies. One species, Ptero-
notus davyi, is known from the Lesser
Antillean islands of Grenada, Dominica,
Marie Galante (off Guadeloupe), and
Martinique.

Description. — Mustached bats,
naked-backed bats, and ghost-faced bats
of the family Mormoopidae range in
size from the small Pteronotus fuli^no-
sus (forearm length 35-40) to the large
P. parnellii (forearm length 54-65). Un-
like the phyllostomatids, with which
they have been grouped in the past, the
mormoopids do not possess a nose-leaf.
Instead, the lips have been expanded
and ornamented with various Haps and
folds that form a "funnel" into the oral
cavity when the mouth is opened. Short
bristle-like hairs surround this "funnel"
and may act to direct airflow toward the
scoop-like mouth in the same manner as
in some insectivorous birds. The nostrils
have been incorporated into the ex-
panded upper lip; above and between
them are various bumps and ridges that,
along with the region immediately
lateral to the nostrils, form a labio-nasal
plate. The shape of this plate is useful
in distinguishing taxa of this family
(Fig. 11 and discussion beyond).

The rostrum is relatively long in the
genus Pteronotus, but short and up-

turned in Mormoops. The pinnae of the
ears are funnel-shaped, with a distal
portion that may be long and lanceolate
(Pteronotus), or extremely short and
rounded (Mormoops). The medial edge
of the distal portion of the pinna may
be simple, but is serrated in several
species. The lower edge of the pinna is
always confluent with a ridge that ex-
tends along the lower lip. Dorsally the
ears may be weakly connected by two
low and inconspicuous ridges that join
on the rostrum and form a prominent
rostral tubercle as in the subgenera
Phyllodia and Chilomjcteris; both ridges
are extremely inconspicuous in the sub-
genus Pteronotus and the rostral tuber-
cle is reduced or absent. In the genus
Mormoops, the ears are united by two
prominent bands that meet on top of
the rostrum and form a high, V-shaped
tubercle. The tragus is complex (Fig.
10), and the eyes are small and incon-
spicuous as compared with the large
and prominent eyes of phyllostomatids.

The wing membrane is variously at-
tached to the side of the body. In the
subgenus Pteronotus, the wing mem-
branes meet and fuse on the dorsal mid-
line, giving these bats a naked-backed
appearance, but the attachment is some-
what lower on the side in the other
mormoopids. The wing membrane and
extensive uropatagium attach to the leg
by way of a weak to strong ligament,
which may be tightly bound to, or en-
tirely free from, the tibia. The foot is
normal and the calcar is long and
usually Hexed forward against the tibia.
A tail is always present and approxi-
mately half of the distal portion pro-
trudes dorsally from the uropatagium.

The pelage of the mormoopids is
short, fine, and densely distributed over
the body. Bats of this family were pre-
viously thought to possess two distinct
color phases, but as will be shown later,
these phases appear to be seasonal. It
is noteworthy that the naked-backed
bats possess a thick, long pelage beneath
the dorsally fused wing membranes.

SMITH: SYSTEMATICS OF MORMOOPID BATS

43

The rostrum and braincase are well
defined regions of the mormoopid cra-
nium; they are more or less in the same
plane in the genus Pteronotus, but the
rostrum is markedly up-turned, resulting
in an abruptly rising forehead in Mor-
moops (Fig. 37). The premaxillaries
are complete and sohdly fused. Al-
though some teeth are noticeably re-
duced in size in certain species, the
dental formula for all mormoopids is
i 2/2, c 1/1, p 2/3, m 3/3 = 34. The
upper molars have a virtually unmodi-
fied dilambdodont cusp pattern and the
last upper premolar is only slightly
molariform. The upper incisors may be
heavy, nearly peg-like, and possess a
distinct lingual heel, or they may be thin
and blade-like (without lingual heel);
usually the inner pair are noticeably
l)ifurcate.

Post-cranial features of these bats
have been discussed. The proximal end
of the humerus is characterized by an
elliptical head that is somewhat canted
toward the greater tuberosity. Both tu-
berosities equal the head of the humerus
in proximal extension and there appears
to be no special relationship between
the humerus and the scapula. However,
the supraglenoid fossa is well developed
and, along with the supraglenoid proc-
ess, forms an effective locking mechan-
ism. Distally the capitulum and troch-
lea are well developed in the genus
Pteronotus- in Mormoops the central
portion of the capitulum is reduced in
size, probably to reinforce the hinge-
like elbow joint. A medial process is
present and has a well-developed spi-
nous process that is involved in the
automatic control of the distal segment
of the wing. The olecranon portion of
the ulna is strongly developed in Mor-
moops, but markedly reduced in Ptero-
notus. The greater and lesser trochan-
ters of the proximal end of the femur
are greatly reduced and located close
to the longitudinal axis; the head of the
femur is spherical and directed craniad.

Generic Analysis

The species of mormoopid bats have
been fairly well defined. However, in
the past, as many as four genera, PhxjUo-
dia, Chilonycteris, Pteronotus, and Mor-
moops, have been recognized (Gray,
lS66a:93); also as few as two have been
recognized — Chilonycteris and Mormo-
ops (Dobson, 1878:447), and Pteronotus
and Mormoops (Burt and Stirton, 1961:
25). Miller (1907:119-121) assigned
mormoopids to three genera, Chilonyc-
teris, Pteronotus, and Mormoops. For
the most part, these classifications seem
to have been primarily founded on two
characteristics— the shape of the crani-
um (more or less normal in Pteronotus
and Chilonycteris and curiously flexed
in Mormoops) and the fusion of the
wing membrane on the back (fused in
Pteronotus and not fused in Chilonyc-
teris). Burt and Stirton (1961:25), in
relegating Chilonycteris to synonymy
under Pteronotus, considered the dorsal
fusion of the wing membrane to be of
no more than specific value.

In constructing a classification of the
Mormoopidae, I have attempted to de-
limit the "natural groups" within the
family by using as many characteristics
as possible. Close examination of crani-
al and external features of these bats
seemed to indicate the existence of at
least three, and possibly four, taxonomic
units within the family. While some of
these features could be expressed purely
by linear dimension, others were more
subtle, being proportional or qualitative
in nature.

In order to quantify and correlate
the observed variation in the cranial and
external characters, I took advantage of
the multivariate comparisons and corre-
lations offered by techniques of numeri-
cal taxonomy. All computations were
carried out using the NT-SYS system of
computer programs (developed by F.
James Rohlf with assistance of John
Kishpaugh and Ron Bartcher) on the
GE 635 computer at The University of
Kansas Computation Center. My first

44

MISCELLANEOUS PUBLICATION MUSEUM OF NATURAL HISTORY

2.160

,760

DISTANCE

360 0.960 0.560

0.160

r

A

fuliginosus

macleayii

personalus

davyi

blainvillii

suapurensis

megolophylla

parnellii

I 830

.530

.230

0.930 0.630 0.330

B

c

fuliginosus

macleayii

personatus

davyi

suapurensis

blainvillii

megalophylla

parnellii

.740

1.440

140

0.840 0.540 0.240

c

— fuliginosus

— macleayii

— personatus

— davyi

— suapurensis

— parnellii

— blainvillii

— megalophylla

Figure 12. Distance phenograms of mormoopid species based on an un-
weighted pair group method of chister analysis using arithmetic averages
(UPGMA, see Sokal and Sneath, 1963:182-185) computed from distance
matrices on standardized characters: A, length of forearm and 11 selected
cranial measurements; B, 10 selected cranial angles (Fig. 2) plus those param-
eters in A; and C, 20 multistate characters in addition to the dimensions used
in A and B. The cophenetic correlation coefficient for these three phenograms
are 0.906, 0.881, and 0.938, respectively. Scales of distances are located above

their respective phenogram.

Figure 13. Three-dimensional models of the projections of mormoopid bats
onto the first three principal components based on correlations among 12, 22,
and 42 characters; A, B, and C, respectively (see Rohlf, 1967 and 1968, for a
discussion of 3-D models). The ta.xa are identified as follows: 1, fuUgitiosiis;

SMITH: SYSTEMATICS OF MORMOOPID BATS

45

2, macleaijii; 3, pamellii; 4, personatiis; 5, davyi; 6, siiapurensis; 7, hlainvilUi;
8, megalophijUa. These models represent a somewhat less distorted summary
view of the phenetic relationships shown in the distance phenograms (Fig. 12,
A-C). Horizontal lines between "balls" illustrate the connections for the short-
est simply connected network in ?i-dimensional (character) space projected
onto each model (see Kruskal, 1956, and Cavalli-Sforza and Edwards, 1967).
The first three factors of the principal components analyses here illustrated
account for 98.12, 95.22, and 89.59 per cent of the variation in A, B, and C,

respectively. See text for discussion.

46

MISCELLANEOUS PUBLICATION MUSEUM OF NATURAL HISTORY

attempt to classify the mormoopids in-
corporated only 11 selected cranial
measurements and the length of the
forearm (see Methods). My purpose in
doing this was to determine what, if
anything, these parameters contributed
to the classification at the generic level.
Figures 12 and 13 show the results of
this analysis. As was expected, size
alone was the single most influential
aspect in this test. In figure 12A species
are arranged in the phenogram from the
smallest (ftiUginosus) at the top to the
largest (parneUii) on the bottom. The
coplienetic correlation for this pheno-
gram was 0.906. It is interesting to note
that the species of nearly equal size
(davyi and hlainvillii, suapurensis and
megaloplnjUo) were set off together.
The effect of size also is apparent in the
three-dimensional projection ( Fig. 13A ) ,
especially along the X axis, which ac-
counted for 82.14 per cent of the trace.
Only 13.28 per cent of the trace was
expressed on the Y axis and 2.71 per
cent on the Z axis. Therefore, linear di-
mension alone, as represented by the
cranial and external measurements em-
ployed in this study, was found to be of
minimal value in classifying the mormo-
opids at the generic level.

My next concern was to introduce
shape as a parameter into the analysis
along with the above-mentioned exter-
nal and cranial dimensions. A represen-
tation of shape was obtained by measur-
ing 10 selected cranial angles (see
Methods). The results of this analysis
( Figs. 12B and 13B ) were somewhat
more consistent with a priori estimates
of the resemblances. Again, parnellii
was clearly separated from all other
mormoopids — by a distance coefficient
of 1.800 (Fig. 12B). Both ghost-faced
bats, Mormoops megalophyUa and M.
hlainvillii, were associated and set
apart from other species by a distance
coefficient of 1.470. However, the large
naked-backed species, suapurensis, was
set oflF from the smaller naked-backed
species davyi by a distance coefficient

of 1.140, the latter being aligned more
closely with the species personatus.
These two species do indeed resemble
each other in size and, more or less, in
the shape of the skull. The cophenetic
correlation for this phenogram was
0.881, the influence of shape nearly
balancing the effect of size. The per
cent of trace along the X axis was 47.16,
39.41 on the Y axis, and 8.65 on the Z
axis. The total effect of shape was to
more or less balance the effect of size as
seen in the relatively wider, higher pro-
jection (Fig. 13B). Although shape in
conjunction with size gave a more rea-
sonable arrangement than size alone,
suapurensis and parnellii still seemed
"out of place."

The next phase of the analysis was
to quantify some of the more obvious
qualitative characters. Following Sokal
and Sneath (1963:77), I selected 20
characters from the cranial and external
anatomy and coded these as non-addi-
tive multistate characters (see Meth-
ods). These characters represent an
array of features that have, in the past,
been weighed heavily (dorsal fusion of
wing membrane), weighed lightly
(shape of tragus and attachment of
flight membranes to ankle), or not used
at all (number of lobes on i2 and basin-
ing of nasal root). In accordance with
procedures of numerical taxonomy,
these characters were given equal
weight as were the other 22 characters
mentioned above and incorporated
herein. The association of the two
ghost-faced species was maintained
with a distance coefficient of 1.680 from
the remaining taxa (Fig. 12C). Again,
parneUii is separated from the other
taxa, but this time with a smaller dis-
tance coefficient (1.620). Both naked-
backed species are allied and the two
small Antillean species macleayii and
fidiginosus retained their close associa-
tion throughout the analysis; the great-
est distance coefficient between these
two insular taxa, 0.400, was observed in
the first phase of the analysis (Fig.

SMITH: SYSTEM ATICS OF MORMOOPID BATS

47

12A). The species personatus still re-
mains associated with macleayii and
fiiJiginosiis on the one hand as well as
with daviji and suapurensis on the other;
distance coefficients were 0.930 and
0.960, respectively. The cophenetic cor-
relation for this phenogram was 0.938.
The segregation of these taxa is more
clearly illustrated in the projection ( Fig.
13C). The majority of the trace is seen
on the X and Y axis (47.59 and 32.71,
respectively) with 9.29 on the Z axis.
Although further analysis might affect
the relative positions of the taxa, it is
doubtful that drastic departures from
the general arrangement presented
above would occur. There remains the
task of deriving a suitable classification
from the above information.

At least three systematic interpreta-
tions of the results seem feasible. The
least satisfactory of these choices would
be to recognize four genera — Phyllodia
(including the species parneUii), Chilo-
nycteris (including fuliginosiis, maclea-
yii, and personatus), Pteronotus (includ-
ing davyi and suapurensis), and Mor-
moops (including hlainviUii and mega-
lophylla). Such an arrangement would
create a taxonomic imbalance, owing
primarily to the great degree of differ-
ence between Mormoops and the other
three genera. This classification (Al in
Fig. 14) would seem to require the
adoption of two subfamilies, Chilonyc-
terinae (including Phyllodia, Chilonyc-
teris, and Pteronotus) and Mormoopinae
(Mormoops). An advantage of this ar-
rangement would be that it emphasizes
the basic dichotomy within the family.
However, recognition of subfamilies
might tend to overemphasize the dis-
tinctness of Mormoops, and the alloca-
tion of the three apparent, but not un-
usually distinct, units (Phyllodia, Chilo-
nycteris, Pteronotus) to generic rank
would seem to oversplit the family at
that level.

Another possible choice would be to
recognize three genera (A2 of Fig. 14)
— Phyllodia and Mormoops (with the

I

Subfamily f

Genus f

Species

Subgenus L

Genus I

2 I

-\ r

J L

J L
J L

3

Genus

Subgenus
Species

^s/^

Figure 14. Possil)le systematic alternatives (A)
to the supposed phylogenetic arrangement (B)
of the mormoopid species: 1, parnellii; 2, fti-
liginosus; 3, macleayii; 4, personatus; 5, davyi;
6, suapurensis; 7, hlainviUii; and 8, megalo-
phylla. Each of three possihle interpretations
(discussed in text) can he deciphered hy reading
rows of hrackets from left to right in estimations
1 and 2 and right to left in 3. The outermost
bracket in each case signifies the familial level.
The horizontal and vertical gaps in the "phylo-
genetic tree" (B) relate to degrees of inferred
differences and similarities.

included taxa described above) and
Pteronotus (comprised of two subgenera
Chilonycteris and Pteronotus). There
probably would be no need for sub-
families in this arrangement and the
recognition of subgenera within the
genus Pteronotus would indicate the
apparent close relationship between the
two units included. However, the main
dichotomy in the family would not be
as conspicuous as in the previous classi-
fication, nor would it be readily appar-
ent that Phyllodia is more closely allied
with Pteronotus (as defined here) than
to Mormoops.

As should be apparent now, there
are two major problems to be consid-
ered in the formulation of a satisfactory

48

MISCELLANEOUS PUBLICATION MUSEUM OF NATURAL HISTORY

classification of the mormoopid bats.
The first of these is the extreme distinct-
ness of the bats of the genus Mormoops.
These are easily distinguished on the
basis of a number of characteristics
(tragus, structure of humerus, lip and
facial morphology, shape of cranium,
size and shape of various teeth, for
example) and there can be little doubt
of their generic status. On the other
hand, the bats that have been referred
to as PhyJIodio, Chilonycteris, and Ptero-
notus are not so distinct from each other
as they are as a group from Mormoops,
although Phyllodia does seem to be
somewhat more distinct from Pteronotus
and Chilonycteris than the latter two
are from each other. These points, it
seems to me, should be expressed, inso-
far as possible within the limitations of
our current system of classification, in
the interest of developing categories of
more or less equal rank.

I propose the following classification
(A3 in Fig. 14) as that which most
closely represents the relationships
within the Mormoopidae as now under-
stood. Two genera, Pteronotus and
Mormoops, are recognized and are as-
sumed to be of approximately equal
"taxonomic rank." This seems reason-
able in view of the overall nature of
variation in the family and the apparent
basal dichotomy. The genus Pteronotus
is the less specialized of the two. This
is apparent, I think, in the rather gen-
eralized structure of the distal end of
the humerus ( Fig. 4, A-D ) and general-
ized to moderately specialized tragus
( Fig. 10, A-F ) , and by the lesser degree
of development of lip and facial excres-
cences and the long, pointed ears that
are not connected (or only slightly so)
over the nose. The genus Mormoops,
on the other hand, appears to be much
more specialized in these features. The
crania of these two genera are markedly
different, especially because of the curi-
ous flexion in Mormoops, which may be
related to orientation of the mouth to an
extreme anterior position where it can

act more efficiently as a scoop for in-
sects during foraging. Inflation of the
braincase in the postparietal regions may
reflect major specializations of the pos-
terior portion of the brain that in turn
relate to more integrated control of cer-
tain neural activities.

I have chosen to divide the genus
Pteronotus into three subgeneric units —
Phyllodia, Chilonycteris, Pteronotus — ^to
indicate the existence of at least three
distinct, yet closely related, lines of de-
velopment. The subgenus Phyllodia
seems to be the most generalized and is
set off from Chilonycteris and Pterono-
tus as indicated by the slightly larger
distance coefficient 1.680 (Fig. 12C).
In addition to the above considerations,
this classification is the most conserva-
tive and, as such, is probably the best
suited at this stage of our knowledge of
chiropteran systematics.

Patterns of Non-geographic Variation

In order to assess the patterns of
geographic variation within individual
species of the Mormoopidae, it was
first necessary to examine several fea-
tures of the non- geographic variation —
namely secondary sexual variation, in-
dividual variation, and seasonal varia-
tion with reference to pelage and molt.
I have not been particularly concerned
in this study with variation with age be-
cause mormoopids appear to mature
rather rapidly, and once coalescence of
epiphyseal and cranial sutures has oc-
curred it is relatively difficult to distin-
guish young adults from older animals.
For the purposes of this study, speci-
mens that had completely fused pha-
lageal epiphyses and cranial sutures and
a fully erupted dental arcade were
judged as adults.

Secondary sexiud variation. — At the
outset of this study, I examined repre-
sentative populations of Pteronotus per-
sonatus, P. davyi, and Mormoops mega-
lophylla in order to determine whether
it would be necessary to segregate the
sexes during the initial data collection;

SMITH: SYSTEMATICS OF MORMOOPID BATS

49

significant secondary sexual variation
was apparent. Results of subsequent
tests on several relatively large samples
of Pteronotus parneUii, P. personatus, P.
doviji, and Mormoops megalophyJla re-
vealed some interesting aspects of mor-
moopid secondary sexual variation (Fig.
15). With several exceptions, probably
due to sample size, males were as large
as, or larger than, females of the same
species from the same geographic area.
The kind and degree of secondary sex-
ual variation differed locally; that is, the
sexes of a species were not always sig-
nificantly different in the populations
tested. In cases when samples were sig-
nificantly different, they did not neces-
sarily vary in the same characters from
one population to the next. For exam-
ple, males and females of P. parneUii
(Fig. 15) from Sinaloa, Mexico, differed
significantly in a single character, the
zygomatic breadth. Males and females
from Jalisco differed by two characters
— zygomatic breadth and condylobasal
length. In a population of parneUii
from Amazonas, Venezuela, the sexes
differed significantly by every character
except length of the forearm and
breadth of the interorbital constriction.
No single character was observed to
vary consistently throughout all samples
of parneUii tested, although condyloba-
sal length and depth of the cranium
were significantly different between
males and females from nearly every
locality tested.

In P. personatus, the breadth across
the mastoid flanges and depth of the
cranium were consistently different in
each sample tested ( Fig. 15 ) . However,
heterogenous secondary sexual variation
was also observed in this species. Of
the four species tested, Mormoops me^^a-
lophyUa exhibited the least amount of
secondary sexual variation within the
populations tested.

In both P. parneUu and personatus
distance coefficients between sexes be-
come progressively larger southwardly
in the range of the species. This general

trend is most striking in populations of
parnelUi. The one insular population
(from Jamaica) of this species also
exhibits considerable distinctness be-
tween males and females. On the other
hand, P. davyi seems to show a reverse
trend with the greatest difference be-
tween males and females observed in
the most northerly (Sonoran) popula-
tion (Fig. 15), although the Guatemalan
sample also shows a relatively high dis-
tance coefficient. Testing of more
southerly populations of personatus and
davyi was unsatisfactory owing to small
samples.

Generally, the males and females of
Mormoops rnegalophyUa, although dif-
fering in several characters, exhibited
consistently lower distance coefficients
than other mainland species analyzed.
No readily apparent geographic trend
in secondary sexual variation was de-
tected, and much of the variation ob-
served appears to be random.

The endemic Antillean species Ptero-
notus fuUginosus, P. macleayii, and Mor-
moops hlainviUii could not be satisfac-
torily tested due to insufficient sample
sizes. Zygomatic breadth, breadth of
the braincase, and breadth of the post-
palatal extension were slightly different
in the Cuban population of P. macleayii.
Males and females of P. fuUginosus
from Puerto Rico exhibited marked dif-
ferences in mastoid breadth, breadth of
rostrum, length of maxillary and man-
dibular toothrows, and depth of crani-
um. Other smaller differences were
recorded in length of the forearm and
zygomatic breadth. The small Mormo-
ops hlainviUii from Puerto Rico, like
the mainland M. megalophylla, showed
only slight secondary sexual differences.

Only a preliminary and highly specu-
lative interpretation of the meaning of
the observed secondary sexual variation
in mormoopids can be made. Little is
known about the ecology and interspe-
cific interaction of the majority of Neo-
tropical bats, to say nothing of inter-
action between sexes of one species.

50

MISCELLANEOUS PUBLICATION MUSEUM OF NATURAL HISTORY

ABGDEFGH

Pteronotus parnellii

+

+

0

®

©

0

0

0

0

0

0

0

©

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

Pteronotus personatus

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

©

0

0

0

©

Pteronotus davyi

Locolity

Sinoloa

Jolisco

Chiapos

Nicaragua

Panama

Venezuela

Amazonas

Jamaica

Jalisco
Morelos
Veracruz
Guatemala
El Salvador

Sonora

Jalisco

Oaxaca

Veracruz

Guatemala

Texas

Nayarit

Morelos

Guatemala
El Salvador
Ecuador
Guracao

Figure 15. Secondary sexual variation in 12 external and cranial dimensions (A-L) of four mor-
moopid l)ats. Dimensions are as follows: A, length of forearm; B, zygomatic breadth; C, breadth of
braincase; D, mastoid breadth; E, zygorostral length; F, interorbital breadth; G, rostral breadth; H,
length of maxillary toothrow; I, breadth of post-palatal extension; J, condylobasal length; K, depth
of braincase; L, length of mandibular toothrow. A plus indicates that the character was significant
at the 0.05 level of probability, whereas a plus and circle indicate significance at the 0.01 level. Co-
efficients of distance between the larger males and females appear in the far right-hand column for

each locality. See text for discussion.

-j-

0

0

0

Mormoops megalophylla

D

stance

0

.726

0

.925

0

.813

1

.002

0 823

0.624

1

284

1

054

0

785

0

654

1

024

i

255

1

305

0.983

0

529

0

464

0

358

0

730

0

429

0

308

0

833

0

445

0

395

0

208

0

456

SMITH: SYSTEM ATICS OF MORMOOPID BATS

51

9.0 _

8.0

7.0

6.0

5.0 _

4.0

3.0

2.0

-0 _

0.0

—

7

—

—

—

—

4
2

—

~

5

—

3
2

6^
1

—

—

5

7

7
1

1 7

2

4

1

7

54
6

8

2
4
5

7
1

3

3

2'

—

4

4

8

4

^?

4

1

8

2 5

8

7

3

6

5

5

5

_^

4

6 8

5

1 56

4

1 7

2 8

^7

6 1
2

6

3

125
8

3
8

4
5

8

8
6 7

3

3

2

6

6
3

7
6

2
3

1

1

3
1

1

7
1

1

1

1

1

3
1

1

4
1

A

C

D

F

G

H

I

J

K

L

9.0

_8.0

_7.0

6.0

_5.0

4.0

3.0

2.0

.0

0.0

Figure 16. Coefficients of variation (ordinate) of external and cranial dimensions used in this study.
These dimensions (abscissa) are as follows: A, length of forearm; B, zygomatic l)readth; C, breadth
of braincase; D, mastoid breadth; E, zygorostral length; F, interorbital breadth; G, rostral breadth;
H, length of ma.xillary toothrow; I, breadtli of post-palatal extension; J, condylobasal length; K, depth
of braincase; and L, length of mandibular toothrow. The plotted numljers refer to mormoopid taxa
as follows: 1, Pteionotus fuliginosus; 2, P. macleaijii; 3, parnellii; 4, P. personatus; 5, P. davyi; 6, P.
suapurensis; 7, Mormoops blainvillii; 8, M. megaloplujlla.

Perhaps males are utilizing the "niche"
in different ways than are females,
which could conceivably account for
differences in size. An interpretation of
the geographic trends in se.xual distinct-
ness is even more difficult. In the case
of Sonoran P. davyi, the less than op-
timal xeric habitat might cause direct
competition between sexes of equal size,
whereas in more favorable mesic habi-
tats there is less emphasis on competi-
tion between males and females. This

interpretation, however, does not seem
to fit the situation found in either P.
parneUii or P. persotiatus. Nevertheless,
the observed differences seem to repre-
sent real variation and certainly merit
further investigation.

Individual variation. — Most charac-
ters examined in this study revealed
relatively little individual variation in
all species; coefficients of variation for
characters usually were low (Fig. 16).
The breadth across the post-palatal ex-

52

MISCELLANEOUS PUBLICATION MUSEUM OF NATURAL HISTORY

tension was the most variable parameter
tested; coefficients of variation in this
measurement ranged as high as 5.0 to
approximately 9.0. Generally, individual
values were tightly clustered around the
mean, giving a bell-shaped, normally
distributed curve that in most cases was
tall and narrow with a relatively wide
base; the curve frequently was skewed.

Externally, a small amount of in-
dividual variation was observed. Sev-
eral species, Pteronotus personatus,
macleayii, and fuUginosiis, have two to
four tooth-like serrations on the proxi-
mal edge of the pinna. The number of
rows of wart-like bumps on the central
pad of the lower lip also varied intra-
specifically. No marked variation was
observed in the structure of the tragus.
The wing membrane of mormoopid bats
usually is black, whereas many other
Neotropical bats (especially phyllosto-
matids) commonly have a white (non-
pigmented) blotch extending from
about the middle of the second phalanx
of the third digit to the wing tip. Only
one individual of a mormoopid (USNM
370057, Pteronotus parnellii, from Ran-
cho Grande, Venezuela) possessed white
wing tips.

Phillips and Jones (1969:516-520)
reported no dental abnormalities in
specimens of Pteronotus fuliginosus,
macleayii, suapurensis, or Mormoops
hlainvillii. Several individuals of P. par-
nellii, personatus, davyi, and M. megalo-
phylla had lost one or more teeth in life,
and a few cases of developmental
anomalies were found. I examined one
specimen of P. fuliginosus (ROM
37128) from St. Clair Cave, Jamaica,
with an extra left upper incisor; the
anomalous tooth was slightly smaller
than 12 and was squeezed between it
and the canine. A disease-caused anom-
aly was observed in a specimen of Mor-
moops megalophylla ( UMMZ 109750)
from near San Cristobal, Chiapas, in
which the entire protocone and hypo-
conal shelf v/as eroded away on both
first upper molars. The erosion was al-

most perfectly symmetrical on both of
these teeth; only the W-shaped cusp
pattern remained. Carlton J. Phillips
( personal communication ) indicated
that this condition might have been
caused by dentinogenesis imperfecta,
which results in a breakdown and rapid
erosion of the dentine layer of the tooth.

Pelage and molt. — Past workers
thought two distinct "color phases" oc-
curred in mormoopid bats. Quite fre-
quently, reddish or orangish bats and
brownish bats were found in collections
near or at the same locality and taken
at the same time of year. In fact,
Thomas (1892a) based the name Ptero-
notus davyi fulvus on a brightly-colored
(fulvous) individual. Also, Wagner's
( 1843 ) specific name rubiginosa sug-
gests the reddish color of his Brazilian
specimen; J. A. Allen (1911) applied
the subspecific name fusca to fuscous
brown specimens of P. parnellii from
northwestern Venezuela. Allen ( 1894 )
also discussed color phases of P. davyi
and concluded that these were of no
particular value in judging subspecific
differences. At the outset of my study,
it seemed reasonable to suspect genetic
polymorphism in color among mormo-
opids. Peculiarities of the molt process
in this group, coupled with the paucity
of molting individuals in collections,
may explain why it did not occur to pre-
vious investigators that "color phases"
might be related to some biochemical
change due to aging of the hair or
bleaching from environmental sources
( ammonia concentration from guano, for
example), or a combination of both.

In the summers of 1966 and 1967 in
Nicaragua, several individuals of Ptero-
notus parnellii were captured that had
a brilliant fulvous band around the
outer margins of the haired portion of
the body, including a narrow ring
around the base of the ears, the anal-
genital region of one female, and on the
tip of the nose. The remaining areas of
the body had a dark brown, glossy pel-
age. Except for slight discoloration on

SMITH: SYSTEMATICS OF MORMOOPID BATS

53

the skin of the upper flanks, there was
no sign of active molt. Upon micro-
scopic examination of the pekige of
these specimens I found individual red-
dish hairs interspersed among the new
l)rown hairs. The density of reddish
hairs decreased near the midline, where-
as brownish hairs, although present,
were infrequent in the lateral reddish
band. Further investigation revealed
that the so-called "color phases" are en-
tirely related to changes in color asso-
ciated with the annual molt.

The progression of molt and changes
in color of hair are discussed below for
Pteronotus pornelUi. All other mormo-
opids follow this same general scheme,
with a few modifications that are men-
tioned in the discussion. Fresh pelage
in parnellii usually is dark Prout's Brown
to blackish, depending somewhat on
geography. As the pelage ages, it be-
comes gradually paler (Snuff Brown).
Exactly what causes this change is not
presently understood, although several
workers ( Constantine, 1958a, 1958b,
and Mitchell, 1964 ) have related bleach-
ing of hair to ammonia, which accumu-
lates in high concentrations in guano-
filled caves. At any rate, the change in
color may progress all the way to a
brilliant fulvous or orange. In the lat-
ter case, new molt is obvious.

Generally, molt begins on the nape
and throat with extensive loss of hairs.
As hair is replaced in these areas, sev-
eral centers (usually symmetrically lo-
cated) of discoloration of the skin, and
possibly new hair growth, can be de-
tected slightly anterior to (or between)
the ears, in the shoulder region over the
scapulae, and in a single middorsal
patch on the rump ( Fig. 17 ) . Ventrally,
there appears to be only a single center
— the throat region (Fig. 17) — from
which active hair replacement prog-
resses. These centers continue to ex-
pand in all directions (Fig. 17) until a
single continuous band of active hair
replacement encircles the body. Molt
seems to progress more rapidly on the

dorsum than on the venter and con-
tinues until new hair is apparent over
the entire surface of the body. In many
cases, significant molt progression takes
place before it is readily noticeable. At
stage 4 or 5 ( Fig. 17 ) , the old pelage
begins to slough off in patches or by
individual hairs. Active molt may termi-
nate and the skin return to normal col-
oration before portions of the old pelage
are sloughed at the peripheral edges of
the haired surfaces (as described above
for the Nicaraguan material). Eventu-
ally, all of the old fulvous hairs fall out
leaving the darker fresh pelage.

Apparently, the rate at which the
hairs change color varies individually
and may be a phenotypic expression of
some genetically caused differences in
biochemical composition of the hair.
Environmental factors such as ammonia
concentrations may amplify this color
change. Another factor that seems to be
important is the reproductive cycle,
which may delay molt in pregnant or
lactating females and thereby allow for
continued color change in old pelage.

There seems to be but one molt an-
nually in mormoopid bats; most popula-
tions of Pteronotus molt from May to
July, whereas molt in Moormoops raega-
lophijUa occurs from June to September.
Because of a paucity of material, I am
uncertain if mormoopids undergo a sec-
ond molt, but several molting indivi-
duals (usually females that were appar-
ently in terminal lactation) were col-
lected in October, November, Decem-
ber, and early January. Nevertheless,
most individuals of Pteronotus are ready
to molt near the onset of the rainy sea-
son, the time at which Pteronotus are
reproducing. Since both processes are
seemingly stressful physiological condi-
tions, it is not unreasonable to presume
that only one occurs at any given time.
Therefore, one might suspect that in
May and early June the males, freed
from their less strenuous reproductive
cycle, are beginning to molt. Males
and non-pregnant, non-lactating females

54

MISCELLANEOUS PUBLICATION MUSEUM OF NATURAL HISTORY

FiGtJRE 17. Generalized diagram of molt progresssion in mormoopid bats. This sequence (based on

Pteronotus parnellii ) and deviations from it are discussed in text. The top ( dorsal views ) and bottom

(ventral views) rows depict molt progression (stages 1-6) from left to right.

molt mostly within this period, and
both sexes have fresh pelage in July and
August. On the other hand, reproduc-
ing females, judged by the presence of
embryos or enlarged, discolored mam-
mae (a presumed indication of lacta-
tion) were found throughout April,
May, June, and early July with increas-
ingly redder pelage. Probably shortly
after termination of lactation, these fe-
males molt from the brilliant fulvous
pelage to the darker fresh pelage. Occa-
sionally, males with old fulvous pelage
and no apparent molt in progress were
found in August.

All of the bats of the genus Pterono-
tus follow the general molt pattern out-
lined above. In the subgenus Chilonyc-
teris, where the fresh pelage consists of
white or grayish hairs interspersed
among the darker hairs, the terminal
phases of molt (stages 4 and 5, Fig. 17)
may result in a beautifully-colored in-

dividual tinged with orange, gray, and
white. Progression of the molt in the
naked-backed P. daviji and P. suapuren-
sis is similar to that in P. parnellii, at
least that part of the molt that can be
seen readily; events that occur under
the fused wing membrane are not
known.

Constantine ( 1958a ) reported on
color variation and molt in Mormoops
rnegalophylki from Frio Cave, Texas.
The molt progression described by him
is similar to that of P. parnellii shown in
figure 17. However, there are several
noteworthy differences in the character
of the pelage of these two monnoopid
genera. Whereas the hairs of Pteronotus
are short, stiff, and usually monocolored
(bicolored in the Antillean species),
those of Mormoops are long, lax, and
have at least four distinct color bands.
The basal band occupies about one-fifth
of the filament length and is dark

SMITH: SYSTEMATICS OF MORMOOPID BATS

55

brownish (Prout's Brown). This basal
band changes color abruptly to a paler
whitish or pinkish buff band that is two
or two and one-half times as wide as the
basal band. The third color band is
about equal in length to the basal band,
is buffy brown, and not sharply deline-
ated from the preceding band. The
hair filament is tipped with a dark
brownish band. Constantine (1958a:
345) described a fifth color band ("zone
X") that he found irregularly between
bands two and three, but the color
gradation between these two bands is so
gradual that I was unable to detect this
fifth zone of coloration with any regu-
larity.

Although "color phases" do exist in
the genus Mormoops, they are not so
distinct nor so striking as those in
Pferonotus. As the pelage of Mormoops
megalophylla ages, the wide second
color band gradually becomes redder;
eventually the entire filament may be
affected. The presence of several other
color bands may contribute to the subtle
change in the color of the hair. In any
event, once the reddish tinge has been
incorporated into the middle paler
bands the overall appearance of the
pelage becomes reddish. This I suspect
is the explanation for the reddish colored
M. megalophylla from western Mexico,
named as a distinct subspecies by Davis
and Carter (1962:65). None of the ma-
terial examined, including the holotype
and type series of these bats (collected
on 26 July), was in what might be
called "fresh pelage." In fact, there are
three adult males (KU 94032, 94034,
and 94035) from near Santa Lucia, Sina-
loa, collected on 28 July, that are in the
process of molting from the "charac-
teristic" reddish or rufus pelage to a
darker, more grayish pelage. Nothing
is known concerning the molt sequence
of M. hlainviUii, but I presume it re-
sembles that described above for M.
megalophylla.

Key to the Genera and Subgenera

1. Rostral portion of cranium elevated,
producing abruptly rising forehead;
p2 not noticeably reduced in size
(as compared with pi); tragus com-
plex, with extensive secondary fold
(Fig. 10); ears funnel-shaped and
round, lacking a long lanceolate por-
tion, and united on the rostrum by a
pronounced band _. Genus Mormoops

Rostral portion of cranium not or
only slightly elevated, braincase and
rostrum more or less in same plane;
p2 reduced in size to a small, peg-
like, unicuspid tooth; tragus with
secondary fold small to moderately
large (Fig. 10); ears funnel-shaped,
with pronounced lanceolate portion,
noticeably separated but may be
connected by a low, inconspicuous
ridge Genus Pteronotus

2. Wing membrane fused on middorsal
line giving a naked-backed appear-
ance; rostral breadth always greater
than length of maxillary toothrow — .
Subgenus Pteronotus

Wing membrane not fused on mid-
dorsal line; rostral breadth equal to
or less than length of maxillary tooth-
row 3

3. Basioccipital narrowly constricted
between auditory bullae; basisphe-
noid with two narrow and deep fur-
rows; forearm usually more than 50
mm Subgenus Phyllodia

Basioccipital not constricted be-
tween auditory bullae; fiu-rows in
basisphenoid wide and shallow; fore-
arm usually less than 50 mm

Subgenus Chilonycteris

Genus Pteronotus Gray

MUSTACHED AND NaKED-BACKED BaTS

Pteronotus Gray, 1838, Mag. Zool. Bot., 2:500,

February.
Chilonycteris Gray, 1839, Ann. Nat. Hist., 4:4,

September (type, Chilonycteris macleayii

Gray). Valid as a subgenus.
Lohosioma Gundlach, 1840, Arch. Naturgesch.,

4:357 (type presumably Lobostoma quad-

56

MISCELLANEOUS PUBLICATION MUSEUM OF NATURAL HISTORY

Figure 18. Side xiew of rostral portion of crania of three mormoopids showing maxillary root of zygo-
matic arch and position of infraorliital foramen. A, Pteronotiis pameUii; B, Ptewnotus personatus;
and C, Mormoops inegalophijUa. Horizontal line below B is 5 mm in length.

ridens Gundlach = Chilontjcteris macleaijii
Gray).

Phyllodia Gray, 1843, Proc. Zool. Soc. London,
p. 50, October (type, Phyllodia panicUii
Gray). Valid as a subgenus.

Dermonotus Gill, 1901, Proc. Biol. Soc. Wash-
ington, 14:177, September 25. Replace-
ment name for Pteronoius Gray, 1838,
thought to be preoccupied by Ptcronotxis
Rafinesque, 1815, a nomen nudum (see
Miller, G. S., Jr., Proc. Biol. Soc. Washing-
ton, 18:223, 17 October 1905).

Type species. — Pteronotiis clavi/i
Gray, 1838.

Distribution. — Humid to semixeric
Neotropics (generally below 3000 me-
ters) from southern Sonora and Tamau-
lipas, Mexico (excluding Baja Califor-
nia and portions of the Mexican pla-
teau), southward through British Hon-
duras, Central and South America (in-
cluding Trinidad) to the Mato Grosso
region of Brasil (excluding Dutch West
Indies). Two endemic species, fuligino-
sus and macleaijii, and several subspe-
cies of parneUii are found in the Greater
Antilles. Pteronotus davyi is also found
on the leeward Antillean islands of Gre-
nada, Dominica, Marie Galante (off
Guadaloupe), and Martinique. For fur-
ther information regarding distribution,
see the following species accounts.

Description. — Bats of this genus are
highly variable in size. Pteronotus ftili-
ginosus torrei from Cuba is the smallest
representative of the family, and Ptero-

notus parneUii nihiginosus from South
America and Pteronotus suapurensis
from Central America are among the
largest members of the family. Exter-
nally, members of this genus are easily
distinguished from the genus Mormoops
by short, stiff pelage, long pointed pin-
nae of ears that may be weakly con-
nected dorsally by a low ridge on the
rostrum that is confluent with a small
V-shaped rostral tubercle, proximal por-
tion of the pinna funnel-shaped and
confluent at lower edge with lower lip,
plagiopatagium and uropatagium firmly
attached directly to ankle or to tibia by
a tightly bound ligament, tragus lanceo-
late with a small secondary fold on
cranial edge or spatulate with a marked
secondary fold (Fig. 10, A-F).

The cranium is "normal" in appear-
ance; that is, the rostrum and braincase
are, more or less, in the same plane.
The rostrum may be slightly tipped up-
ward in the subgenus Pteronotus, but
never is so extremely elevated as in
Mormoops. The zygomatic arches are
complete and well developed, and lie,
more or less, in the same horizontal
plane as the rostrum and braincase.
The lambdoidal ridges are strongly de-
veloped and form a prominent flange
from the posteriormost part of the low
sagittal crest to the tip of the short
mastoid process that overlays the tym-

SMITH: SYSTEMATICS OP^ MORMOOPID BATS

57

panic ])iilla posterior to the external
auditory meatus. Viewed from the
ventral side, the presphenoid, which lies
between the pterygoid wings and forms
the roof of the mesopterygoid fossa, is
elevated "above" the level of the basi-
sphenoid giving it a "step" appearance.
In all species except P. personatus, the
infraorbital foramen opens (on the side
of the rostrum at about its widest point )
at the end of a short, tubular canal that
is visible through the thin bone of the
maxillary. In P. personatus and in Mor-
moops this foramen opens abruptly at
the anterior root of the zygomatic arch,
the canal being formed only by a thin,
bridge-like spicule of bone ( Fig. 18 ) .

The molars are typically dilambdo-
dont, and the entire dental arcade is
compact, without striking diastemas.
The upper incisors are prominent and
fill the space between the canines. The
lower incisors are generally trifid, small,
and bear a lingual heel. In the sub-
genus Phyllodia, these teeth are large
and bulky. The second lower premolar
is markedly reduced to a small peg-like,
unicuspid tooth and is almost always
excluded (lingually), or nearly so, from
the toothrow.

Subgenus Phyllodia Gray

Type species. — Phyllodia parnellii
Gray, 1843.

Geographic distribution. — Greater
Antilles and on the mainland from
southern Sonora and Tamaulipas, Mex-
ico, southward along Pacific and Gulf
coastal lowlands, throughout Central
America, and into South America where
it occurs along the Caribbean coast (in-
cluding Trinidad) and south into the
Amazon Basin to the Mato Grosso re-
gion of Brasil. These bats have not
been reported from west of the Andes
in Colombia, Ecuador, or Peru, nor from
French Guiana or eastern Brasil.

Description. — Size medium to large;
wing membranes attached low on sides
of body and attached to ankles by way
of a short, tightly bound ligament; back

covered by a wide band of short stiff
hairs; labio-nasal plate simple; margin
above each nostril with several irregular,
wart-like tubercles; deep emargination
between nostrils; area lateral to nostrils
lacking spike-like projection (Fig. IIA);
tragus simple, lanceolate, secondary
fold small (Fig. lOA); antero-medial
edge of ear pinnae smooth, lanceolate
portion broad; ears connected by two
low, inconspicuous ridges that fuse on
top of the muzzle forming a prominent
rostral tubercle; profile of skull rela-
tively flat, rostrum not elevated, fore-
head sloping gradually onto rather long,
broad braincase; upper incisors robust
and peg-like, inner pair distinctly bi-
furcate and with broad, rounded heel;
lower incisors also heavily constructed,
inner pair trilobed, outer pair bilobed;
tympanic rings small in proportion to
those of other subgenera, covering ap-
proximately one-third of the auditory
bullae; basioccipital region narrowly
constricted between bullae and with two
narrow and deep longitudinal furrows
into which the M. longns capitis inserts.
Remarks. — This monotypic subgenus
is quite distinct and is not liable to be
confused with either of the other sub-
genera— Chilonycteris or Pteronotus.
Novick (1963a: 47) noted that the acous-
tic orienting pulses emitted by P. par-
nellii were strikingly different from
those of any other mormoopid or phyllo-
stomatid studied by him.

Pteronotus parnellii (Gray)
Parnell's Mustached Bat

( Synonymy under subspecies )

Distribution. — Same as for subgenus
(Fig. 19).

Description. — Sec account of sub-
genus.

Relation.ships. — In many respects,
Pteronotus parnellii is the most primi-
tive member of the family Mormoopi-
dae. The tragus is not so ornamented as
in other species of the family, which
tend toward complexity through empha-

58

MISCELLANEOUS PUBLICATION MUSEUM OF NATURAL HISTORY

59

21.50 ]

58

13.11

58

21.35 ]

59

13.06

54

21.25 ]

60

13.05

57

21.24 ]

54

13.01

41

21.12 ]

57

12.98

37

21.09 ]

41

12.97

60

20.99 ]

37

12.91

40

20.96 ]

55

12.90

43

20.92

35

12.87

35

20.88

43

12.85

55

20.87 ]

36

12.80

31

20.85

44

12.80

53

20.81

42

12.79

42

20.80

56

12.77

45

20.79

38

12.75

56

20.73

53

12.71

44

20.62

40

12.68

36

20.57

45

12.67

38

20.55

34

12.65

47

20.54

31

12.61

39

20.48

30

12.58

51

20.48

39

12.55

46

20.39

51

12.53

50

20.38

46

12.51

34

20.35

22

12.50

33

20.34

32

12.46

28

20.32

27

12.44

27

20.31

48

12.42

49

20.31

50

12.42

48

20.30

14

12.40

14

20.29

29

12.39

30

20.28

28

12.38

52

20.26

25

12.37

25

20.23

47

12.36

26

20.22

20

12.35

32

20.21

33

12.34

13

20.17

13

12.31

29

20.13

26

12.30

22

20.10

49

12.28

20

20.07

17

12.23

17

19.90

16

12.16

24

19.89

15

12.16

23

19.82

24

12.07

16

19.77

52

12.06

15

19.70

23

12.03

21

19.67

21

11.97

11

19.62

9

11.91

12

19.55

18

11.90

7

19.50

12

11.87

5

19.48

11

11.85

8

19.44

7

11.84

4

19.42

8

11.81

9

19.41

5

11.79

10

19.39

19

11.78

6

19.34

6

11.73

62

19.21

4

11.72

18

19.18

10

11.71

19

19.17

3

11.58

2

19.15

1

11.57

3

19.13

2

11.55

61

19.12

62

11.34

65

18.90

61

11.13

1

18.90

65

11.02

63

17.47

63

10.25

64

16.40

64

10.15

I

II
III
nil
iiii
iiii
mil
mill
iiiiiii
iiiiiii
IIIIIII
IIIIIII
IIIIIII
IIIIIII
IIIIIII
IIIIIII
IIIIIII
mini
IIIIIII
IIIIIII
mm
mm
mil
IIII
IIII
III
III
II
I
I
I

I

II

II

I

B

Figure 19. Results of two typical SS-STP analyses (A, condylobasal length and B, zygomatic breadth)
of geographic variation of Pteronotiis parnellii. Vertical line to the right of each array of means con-
nect maximally non-significant subsets at the 0.05 level; for example, locality 59 with a mean
condylo]:)asal length of 21.50 is not significantly different from localities 58-56 with means ranging
from 21.35 to 20.73, hut does differ significantly from all other localities. See text for key to localities.

SMITH: SYSTEM ATICS OF MORMOOPID BATS

59

9.000 7.000 5.000 3000 1.000

rl

■c

I -I

2

3

4

6
19
18

5
10

7

8

9
12
II
21 -^
61
62
65
13-
26
14
25
22
27
33
49

■ 28

- 32

- 47
• 29

- 48

- 17

■ 20

- 52

- 30

- 34

- 38

- 39

- 46

- 51

- 50

- 15-1

- 16

- 23
-24

- 31

- 40

- 44
-45
-43

- 35

- 42

- 55

- 56

- 53

- 36

- 37
-41
-54

- 60
-57
-58

- 59-"

- 63

- 64

A

}o

o

-D

>V

9.000 7.000 5.000 3.000 1.000

Figure 20. Phenogram of geographic variation
of Pteronotus parnellii computed from the dis-
tance matrix on standardized locality means and
clustered by the imweighted pair-group method
using arithmetic averages (UPGMA). The
cophenetic correlation for this phenogram is
0.863. Symbols to the right of the phenogram
relate to localities as plotted in figure 21. See
text for key to localities.

sis of a secondary fold (Fig. 10). Sec-
ondly, the rostrum is only slightly ele-
vated and maintains an "in line" posi-
tion with the braincase, whereas the
upturned snout in other mormoopid
bats seems to be an adaptation for more
efficient feeding and/ or echolocation.
The braincase is large and bulky and
may reflect a generalized brain with
less refined control centers than in other
species, and the extra long sonic pulses
(as reported by Novick, 1963a) emitted
by this species may indicate a more
generalized mode of echolocation. On
the basis of close resemblance P. par-
nellii seems to have its closest ties with
the subgeiius Chilonycteris and particu-
larly with the species P. personatus. Ap-
parently P. parnellii diverged from the
mainline of mormoopid evolution at a
rather early time.

Geographic variation. — As previously
understood, parnellii was divided into
eight geographic races — three on the
mainland and five on the Greater An-
tilles (one of which, P. p. gonavensis, is
known only from sub-Recent cave de-
posits ) . Prior to Koopman's ( 1955 ) re-
port, the insular and mainland popula-
tions were regarded as distinct species,
parnellii and rubiginosus, respectively.
Koopman's proposition that they were
conspecific has caused some controver-
sy, not particularly concerning the rela-
tionships of the bats, but with the name
to be used if the two groups are con-
specific. This problem, which concerns
date of publication and priority, has
been discussed by de la Torre (1955:
696), Felten (1956:72), Koopman and
Martin (1959:4-5), Husson (1962:75-
76), and recently by Jones (1966:448).
The International Code (Article 21b, ii,
p. 19) is explicit about interpreting the
date of publication when the exact date
of issue within a year is unknown, as in
the case of rubiginosus Wagner, 1843,
This work must be considered as having
been published on the last day of the
year, 31 December 1843, thereby giving
priority to Gray's name, parnellii, which

60

MISCELLANEOUS PUBLICATION MUSEUM OF NATURAL HISTORY

was published in October of that year.
Furthermore, de la Torre's argument
(1955:696) that ruI)i<i,inosiis should be
retained because it is "the better known
and more widely used name" has no
basis' because the name parnellii has
been consistently used in the primary
literature dealing with Antillean bats.

Although the above considerations
are important, the basic concern is
whether or not these two taxa are mem-
bers of the same biological species. Be-
cause it is doubtful that direct evidence
of intergradation between Antillean and
mainland populations ever will be avail-
able, conspecificity can be judged only
on the basis of resemblance. All of the
insular populations are smaller in crani-
al and external size than are those from
the mainland (Table 1), although P. p.
parneUii from Jamaica is only slightly
smaller than P. p. mexicanus from west-
ern Mexico. The pelage of the Antillean
races of parnellii appears to be "tri-
colored" dorsally, owing mostly to a
more pronounced whitish central zone
on the hair filaments in contrast to the
darker basal and terminal zones. These
zones of color on the hairs of the back
cannot be called color bands because
there is a gradual change from one to
the other, in contrast to an abrupt,
sharply defined change of color as seen
on hairs of the venter. Insular popula-
tions average slightly paler in color than
most mainland subspecies. The tragus
of the insular bats closely resembles
that of mainland specimens except that
the distal half is slightly broader. In
other respects, there is little difference
between insular and mainland bats and
at this time there seems to be nothing
to gain by splitting these at the specific
level.

In two typical results of the SS-STP
analyses ( Fig. 19 ) sample means of the
Antillean populations (localities 61-65)
are located at the small extreme of the
size variation. It is also apparent that
there are two groups within these insu-
lar populations. One of these includes

larger bats from Cuba, Jamaica, and
Puerto Rico (localities 61, 62, and 65,
respectively), and the other contains
smaller individuals from Hispaniola and
La Gonave (localities 63 and 64). These
two groups are separated by a distance
coefficient of 7.800 in the distance
phenogram (Fig. 20). In this pheno-
gram, the grouping together of locali-
ties 63 and 64, which correspond to P. p.
pusiUus and P. p. gonuvensis, respective-
ly, indicates their similarity and their
rather marked distinctness from other
populations of P. parnellii. However,
the distance coefficient of 1.800 between
locality 63 and 64 also indicates a nota-
ble degree of distinctness between pu-
siUus and gonavensis.

In the couplet encompassing locali-
ties 61, 62, and 65, Cuba and Jamaica
( 61 and 62 ) are more closely associated
than either is to Puerto Rico (65), sug-
gesting a close relationship between
specimens from Cuba and Jamaica. On
the other hand, the geographic position
of Puerto Rico and the distinctness of
the interposed populations of Hispaniola
and La Gonave suggest independent
variation of the Puerto Rican bats.
These features are discussed in greater
detail in the subspecies accounts that
follow.

The mainland populations of P. par-
nellii are currently assigned to three
geographic races: mexicanus Miller
(1902:401) from southern Sonora and
Tamaulipas south to the region of the
Isthmus of Tehuantepec in southern
Mexico; juscus J. A. Allen (1911:262)
from southern Mexico through Central
America to the Caribbean coast of
northern Colombia and Venezuela (in-
dividuals from Trinidad were assigned
to this race by Goodwin and Green-
hall, 1961:222); and nihiginosus Wag-
ner (1843:367) presumably occurring
throughout the Amazonian lowlands to
as far south as the Mato Grosso region
of Brasil. On the basis of data from
more than 1800 specimens of this spe-
cies, I conclude that the geographic

SMITH: SYSTEM ATICS OF MORMOOPID BATS

61

Figure 21. Map showing selected distance coefBcients (from the distance matrix) between localities
from which specimens of P. parnellii were analyzed in the study of geographic variation. Symbols —
circles, squares, triangles, and hexagons — correspond with particular couplets of the distance pheno-

gram (Fig. 20). See text for key to localities.

variation, as expressed by cranial and
external dimensions, is much more com-
plex than implied by these names. Al-
though this species is abundantly repre-
sented in museum collections, there still
remains a serious paucity of material
from several critical areas. Additional
specimens and ecological data are
needed from eastern Panama, northern
Colombia, and the northwestern portion
of the Amazon Basin along the eastern
margin of the Andes.

Relatively little is known about the
ecology of these bats. From my limited
experience in western and southern
Mexico, Nicaragua, and northeastern
Venezuela, I think that P. parnellii has
rather broad ecological tolerances. For
example, at Palenque, Chiapas, this

species was captured along trails cut
through well-stratified, tropical decidu-
ous forest. Near Santa Lucia, Sinaloa,
specimens were caught in patches of
deciduous forest that interdigitated with
the oak and pine-oak forests of higher
elevations. In western Nicaragua, P.
parneUii was netted in xeric scrub forest
on the Cosiqiiina Peninsula and on the
western portion of the Escudo Central
in park-like savanna. In addition these
bats are common in the humid Carib-
bean lowlands of eastern Nicaragua.
The same range of arid to humid habitat
also was noted in northeastern Vene-
zuela.

Greatly simplified, the single most
obvious trend in the geographic varia-
tion of the mainland representatives of

62

MISCELLANEOUS PUBLICATION MUSEUM OF NATURAL HISTORY

this species is an increase in external
and cranial size, along with a less
marked change in color from pale to
dark, at progressively more southern
latitudes (Fig. 19). Cranial and exter-
nal size as well as color seem to be
governed by yet poorly understood eco-
logical factors. Apparently, P. parnellii
as well as the entire genus Fteronotus,
is mesophilic in habitat preference. In
areas where there is a noticeable de-
parture from the preferred humid tropi-
cal environment there is a concordant
phenotypic expression of small size and
pale color. In figure 20, the uppermost
couplet — distinguished by a distance co-
efficient of 2.200— groups together popu-
lations of small-sized individuals of the
entire subspecies P. p. mexicanus, a
population from the northern part of
the Yucatan Peninsula, and the Antil-
lean populations from Cuba, Jamaica,
and Puerto Rico. The remaining main-
land populations of the species, which
occupy more optimal tropical environ-
ments, are grouped (Fig. 20) by a
couplet of lesser magnitude (distance
coefficient, 2.000). However, within
this group there are two distinct units
that seem to correspond to the wet and
dry extremes expressed in the environ-
ments of Middle and South America.
(See figure 21 for selected distance co-
efficients between samples of P. par-
nellii. )

There are two distinct populations
of P. parnellii south of the Isthmus of
Tehuantepec in Middle America. One
of these is described as a new subspe-
cies and occupies northern Middle
America and the Pacific versant. The
other is a northern extension of P. p.
ruhiginosiis from the Amazonian low-
lands of South America and occupies
the humid Caribbean versant from
Central Honduras through eastern Pan-
ama. In South America, there are two
distinct populations; one of these as
mentioned above is the Amazonian P. p.
ruhiginosus. The other, P. p. fuscus, is
restricted to the semiarid Caribbean

coast of Colombia and Venezuela and
closely resembles the new subspecies
from northern Middle America and the
Pacific versant farther to the south (see
account of P. p. fuscus for further dis-
cussion). Particular features of the geo-
graphic variation within and between
the above-mentioned populations are
considered in somewhat more detail in
the accounts of the subspecies.

Key to localities used in statistical analysis
of geographic variation. — 1, Sonora; 2, north-
em Sinaloa; 3, Panuco, Sinaloa; 4, vicinity of
San Bias, Nayarit; 5, vicinity of Jamay and
Chapala, Jalisco; 6, Colima; 7, southern Ta-
maulipas; 8, Cueva del Coyote, 6 km SSE
Tonatico, Mexico; 9, Cueva Poza de Monte-
zuma, 1 km E Oaxtepec, Morelos; 10, vicinity
of Teloloapan, Guerrero; 11, vicinity of
Colotlipa, Guerrero; 12, Yerbabuena, Guerrero;

13, vicinity of San Andres Tuxtla, Veracruz;

14, Teapa, Tabasco; 15, vicinity of Escarcega,
Campeche; 16, southeastern Campeche; 17,
northern British Honduras; 18, Gruta de
Balankanche, 5 km E Chichen Itza, Yucatan;
19, vicinity of Tehuantepec, Oaxaca; 20,
vicinity of Arriaga and Tonala, Chiapas; 21,
vicinity of Tuxtla Gutierrez, Chiapas; 22, vi-
cinity of Palenque, Chiapas; 23, Cueva de
Zapaluta, 1.3 mi SE Zapaluta, Chiapas; 24, La
Democracia, Huehuetanango, Guatemala; 25,
vicinity of San Cristobal Verapaz, Alta Vera-
paz, Guatemala; 26, vicinity Las Quebradas,
Izabal, Guatemala; 27, vicinity of Yepocapa,
Chimaltenango, Guatemala; 28, Finca El Car-
nero, Jutiapa, Guatemala; 29, Las Flores
Gracias, Santa Barbara, Honduras; 30, vicinity
of Comayagua, Comayagua, Honduras; 31,
San Jose, Rio Tinto, Olancho, Honduras; 32,
vicinity of La Libertad, El Salvador; 33, vicin-
ity of Divisadero, La Union, EI Salvador; 34,
Esteli and Dario, Nicaragua; 35, vicinity of El
Recreo, Zelaya, Nicaragua; 36, 3 mi S, 10 mi
W Las Caiias, Guanacaste, Costa Rica; 37,
Almirante, Bocas del Toro, Panama; 38, Cerro
Hoya, Los Santos, Panama; 39, Penonome,
Code, Panama; 40, Paraiso, Canal Zone, Pana-
ma; 41, Bas Obispo, Canal Zone, Panama; 42,
Summit, Canal Zone, Panama; 43, Cerro Azul,
Panama, Panama; 44, Chilibre, Panama, Pana-
ma; 45, Buena Vista, Colon, Panama; 46,
Coloso, Las Campanas, Bolivar, Colombia; 47,
Minas de Aroa, Bolivar R. R., Yaracuy, Vene-
zuela; 48, vicinity of Urama, Yaracuy, Vene-
zuela; 49, vicinity of San Esteban, Carabobo,
Venezuela; 50, Curapao, 19 km E Caracas,
Miranda, Venezuela; 51, northeastern (Sucre)
Venezuela; 52, Isla de Margarita, Nueva
Esparta, Venezuela; 53, Trinidad; 54, eastern
Bolivar, Venezuela; 55, vicinity of Maripa,

SMITH: SYSTEM ATICS OF MORMOOPID BATS

63

20

10

Pp.mexiconus

Pp. mesoame

0 200

Scale of Miles

'00

90

Figure 22. Distribution of Pteronotus parnellii in North America and northern South America. A
few localities of record elsewhere in South America (Brasil, the Guianas, and Peru) are not shown

(see account of P. p. rubiginosus).

Bolivar, Venezuela; 56, vicinity of Punto Paez,
Apure, Venezuela; 57, Rio Cunucunuma, Belen,
Amazonas, Venezuela; 58, Rio Mavaca, 108
km SE Esmeralda, Amazonas, Venezuela; 59,
Rio Negro, Serra de Cucuhy, Rio Bronco,
Brasil; 60, vicinity of Tingo Maria, Huanuco,
Peru; 61, Cuba; 62, Jamaica; 63, Hispaniola;
64, vicinity of En Cafe, La Gonave Island,
Haiti; 65, Puerto Rico.

Pteronotus parnellii parnellii (Gray)

Phyllodia Parnellii Gray, 1843, Proc. Zool. Soc.
London, p. 50, October.

Chilonycteris osbumi Tomes, 1861, Proc. Zool.
Soc. London, p. 66, May. Type locality.
Sportsman Hall and Oxford Cave, Man-
chester, Jamaica; syn types, BM 7.1.1.626
(male) and 7.1.1.632 (sex unknown).

Chilonycteris boothi Gundlach, 1861, Monatsb.
k. preuss. Akad. Wiss., Berlin, p. 154. Type
locality, "In Fundador, auch in Guines"
[Matanzaz and Habana provinces, Cuba,
respectively]; present disposition of holo-
type unknowTi.

Syntijpes. — Two adults, sex unknown,
skins, skulls not removed, BM 43.6.15.6-
7; from unspecified locality in Jamaica;
obtained by Richard Parnell.

Distribution. — Cuba and Jamaica
(Fig. 22); no information presently
available suggests that this race is
limited to specific regions on either
island, but future investigation may re-
veal such restriction.

Description. — Largest Antillean race
both externally and cranially. Standard
external measurements (average, with
extremes in parentheses) of eight males
and four females, respectively, from St.
Clair Cave, St. Catherine Parish, Jamai-
ca, as follows: total length, 79.4 (77.0-
81.0); length of tail vertebrae, 20.4
(18.0-22.0); length of hind foot, 13.0
(13.0); length of ear from notch, 20.7
(20.0-22.0); see table 1 (localities 61
and 62), for additional selected external
and cranial measurements.

The pelage is generally "tricolored"
dorsally. As mentioned previously, color
bands are not readily discernible, but
three zones of color are more or less
distinguishable. The basal fifth of the

64

MISCELLANEOUS PUBLICATION MUSEUM OF NATURAL HISTORY

TABLE 1. Selected external and cranial measurements of several populations of Pteronotiis parnel-
lii. Each measurement is represented by the obser\ed range, below which is the mean, plus or minus
two standard errors of the mean; superscript numbers indicate sample size if different from that

given in left-hand colmun. See text for key to localities.

Locality

N

Length of
Forearm

Zygomatic
Breadth

Rostral
Breadth

Length of
Maxillary
Toothrow

Condylobasal
Length

Depth of
Cranium

Fteronot

us parnellii mexicanus

1

10

55.1+0.72^

11.6+0.18

7.5+0.14

8.6+0.13

18.9+0.17

8.9+0.12

(53.3-56.3)

(11.0 11.9)

(7.2-7.9)

(8.2-9.0)

(18.5-19.3)

(8.6-9.2)

6

22

55.6±0.52

11.7+0.09^

7.6+0.07

8.8+0.07

19.3+0.15

8.9+0.09

(53.1-57.9)

(11.3-13.1)

(7.4-8.0)

(8.5-9.2)

(18.8-20.0)

(8.4-9.3)

7

27

57.4±0.33-«

11.8+0.09

7.6+0.08

8.8+0.05

19.5+0.09^"

8.9+0.07

(56.2-60.0)

(11.5-12.4)

(7.3-8.2)

(8.5-9.1)

(19.1-19.9)

(8.6-9.4)

8

14

58.1+0.64

11.8+0.11

7.7+0.09

8.8+0.06

19.4+0.12

9.1+0.05

( 55.2-59.4 )

(11.5-12.1)

(7.4-8.0)

(8.6-9.0)

(19.1-19.8)

(8.9-9.3)

12

14

57.5+1.82

11.9+0.11"

7.8+0.08

8.9+0.10

19.5+0.15"

9.1+0.09"

(55.3-59.7)

(11.4-12.1)

(7.6-8.0)

(8.6-9.3)

(19.0-20.6)

(8.8-9.4)

19

11

55.7+0.51

11.8+0.12

7.6+0.10

8.7+0.03

19.2+0.14

8.9±0.07

(54.2-56.6)

(11.5-12.2)

(7.3-7.9)

(8.5-9.1)

(18.9-19.7)

(8.8-9.1)

Pteronotus parnellii mesoainericanus

20

11

58.7+1.95

12.4+0.12^°

8.0+0.09

8.9+0.19

20.1+0.17"

9.3+0.12^»

(56.2-61.9)

(12.0-12.6)

(7.8-8.2)

(8.1-9.3)

(19.7-20.4)

(9.1-9.8)

15

13

58.3+0.51

12.2+0.12

7.9+0.10

9.0+0.07

19.7+0.14

9.1+0.13

(56.5-59.7)

(12.0-12.4)

(7.6-8.1)

(8.9-9.2)

(19.4-20.1)

(8.8-9.3)

18

4

55.3+0.68

11.9+0.14

7.6+0.13

8.8+0.10

19.2+0.17

8.9+0.22

(54.6-56.2)

(11.7-12.0)

(7.5-7.8)

(8.7-8.9)

(19.0-19.4)

(8.6-9.1)

27

14

58.7+0.54"

12.4+0.11

8.0+0.08

9.2+0.06

20.3+0.13

9.4+0.10

(57.5-60.0)

(12.1-12.7)

(7.7-8.2)

(9.0-9.4)

19.8-20.7)

(9.0-9.7)

33

21

59.2+0.52^"

12.3+0.07

8.0+0.09

9.2+0.06

20.3+0.11

9.4+0.07

(57.4-61.2)

(12.1-12.7)

(7.6-8.4)

(9.0-9.5)

(20.1-20.9)

(8.9-9.6)

30

5

59.9+0.75

12.6+0.26

8.2+0.12

9.3+0.17

20.3+0.37

9.5+0.10

(.58.8-60.8)

(12.2-12.9)

(8.0-8.3)

(9.0-9.5)

(19.6-20.6)

(9.3-9.6)

38

19

59.5+0.59

12.7+0.12

8.2+0.05

9.4+0.07

20.5+0.10

9.5+0.10

(56.8-61.9)

(12.3-13.2)

(7.9-8.4)

(9.1-9.5)

(20.2-21.2)

(9.1-10.0)

Pteronotus parnellii rubiginosus

31

9

60.2+0.75

12.6+0.12

8.3+0.09

9.4+0.11

20.9+0.13

9.6+0.08

(58.7-61.6)

(12.3-12.8)

(8.1-8.5)

(9.2-9.7)

(20.6-21.2)

(9.4-9.8)

35

12

61.2+0.66"

12.9+0.17

8.3+0.12

9.6+0.09

20.9+0.18

9.7+0.09

(59.6-63.0)

(12.4-13.3)

(8.0-8.6)

(9.3-9.8)

(20.3-21.3)

(9.4-9.9)

37

10

62.0+0.86

12.9+0.11

8.4+0.13

9.6+0.08

21.1+0.23

9.6+0.12

(59.6-64.1)

(12.7-13.2)

(8.1-8.7)

(9.3-9.7)

(20.5-21.9)

(9.4-10.0)

40

26

61.0+0.39'"

12.7+0.10

8.2+0.07

9.5+0.06

21.0+0.12

9.5+0.09

(59.7-62.9)

(12.2-13.2)

(7.9-8.5)

(9.2-9.8)

(20.3-21.4)

(9.1-9.9)

SMITH: SYSTEM ATICS OF MORMOOPID BATS

65

TABLE 1. (Continued)

Locality

N

Length of
Forearm

Zygomatic
Breadth

Rostral
Breadth

Lengtl'i of
Maxillary
Toothrow

Condylobasal
Length

Depth of
Cranium

56

9

61.1+0.77

12.8+0.17

8.4+0.12

9.5+0.14

20.7+0.32

9.6+0.11

(59.5-62.S)

(12.4-13.2)

(8.1-8.7)

(9.2-9.7)

(19.6-21.1)

(9.5-9.9)

54

15

61.3±1.90"

13.0+0.13"

8.5+0.08

9.6+0.08

21.3+0.14"

9.6+0.09'^'

(58.4-64.3)

(12.6-13.3)

(8.3-8.8)

(9.5-9.9)

(20.9-21.7)

(9.4-9.9)

57

5

61.2+0.55

13.0+0.12

8.5+0.19

9.8+0.10

21.2+0.16

9.7+0.13

(60.7-62.2)

(12.8-13.1)

(8.2-8.7)

(9.6-9.9)

(21.0-21.5)

(9.5-9.9)

60

23

63.0+1.00"

13.1+0.07

8.4+0.08

9.5+0.07

21.0+0.11

9.6+0.09

(60.2-65.4)

(12.7-13.4)

(8.0-8.8)

(9.1-9.7)

(20.5-21.6)

(9.1-10.0)

53

12

59.7+ 1.6&'

12.7+0.12"

8.5+0.11"

9.6+0.10

20.8+0.17^

9.6+0.12"

(58.3-61.6)

(12.5-13.0)

(8.3-8.8)

(9.3-9.9)

(20.4-21.0)

(9.4-9.8)

Pteronotus painellii f

ISCUS

46

15

59.7+1.83"

12.5+0.13'-

8.3+0.05

9.4+0.06

20.4+0.13'^'

9.4+0.10'-

(57.2-62.7)

(12.0-12.8)

(8.1-8.4)

(8.9-9.6)

(20.0-20.9)

(9.1-9.6)

49

19

59.9+0.67''

12.3+0.12

8.0+0.08

9.2+0.06

20.3+0.18'^

9.3+0.05

(57.5-62.5)

(12.0-12.6)

(7.8-8.4)

(8.9-9.5)

(19.0-20.8)

(9.2-9.5)

51

10

59.8+0.62

12.5+0.20'-'

8.3+0.11

9.3+0.11

20.5+0.20"

9.4+0.16"

(58.5-61.4)

(12.1-13.1)

(8.0-8.6)

(9.0-9.5)

(20.2-21.1)

(9.1-9.8)

52

15

58.0+0.52

12.1+0.13"

8.1+0.07

9.2+0.08

20.3+0.13'^'

9.2+0.09"

(56.3-60.0)

(11.7-12.6)
Pterunc)

(7.8-8.3)
tiis parnellii pa

(9.0-9.4)
rnellii

(19.8-20.6)

(9.0-9.5)

61

12

51.6+0.62^

11.1+0.13"

7.5+0.12

8.7+0.12

19.1+0.16

8.5+0.10'"

(50.0-52.5)

(10.8-11.6)

(7.2-7.9)

(8.2-9.0)

(18.7-19.4)

(8.3-8.8)

62

16

53.2+0.45'"'

11.3+0.05'-^

7.6+0.07

8.9+0.05

19.2+0.08

8.6+0.08

(51.7-54.7)

(11.1-11.5)

(7.4-7.8)

(8.7-9.1)

(18.9-19.6)

(8.1-8.8)

65

4

50.5+0.22

11.0+0.22

7,6+0. 19-'

8.8+0.12'

18.9+0.14

8.6+0.17

(50.2-50.7)

(10.8-11.3)

(7.3-7.8)

(8.6-8.9)

(18.8-19.1)

(8.3-8.7)

Pteronotus painellii pi

tsillus

64

2

10.1+0.10

6.6+0.07"'

7.5+0.08

16.4

8.1+0.18"

(10.1-10.2)

(6.5-6.7)

(7.3-7.6)

(7.9-8.2)

Pteronotus parnellii gonavensis

63

8

10.2+0.13

7.0+0.11

7.9+0.08

17.5+0.23^

8.2+0.12

(10.0-10.5)

(6.7-7.2)

(7.7-8.1)

(17.0-17.9)

(7.9-8.2)

individual hairs is dark brown (Front's
Brown) grading to a paler (whitish)
central zone. The tips of the filaments
are generally Snuff Brown. On top of
the head, the central zone of the hairs
expands and nearly obliterates the
darker basal zone, imparting an overall

pale appearance. Ventrally, the pelage
on the pectoral and abdominal regions
is sharply bicolored; the basal half of
each hair is dark brown, whereas the
terminal half is whitish. On the throat,
the hairs are almost completely white or
have a slight brownish cast.

66

MISCELLANEOUS PUBLICATION MUSEUM OF NATURAL HISTORY

A

B

C

D

F

G

H

J

Figure 23. Dorsal views of skulls of Pteronotus parneUii gonavensis (A, La Gonave Island, Haiti),
P. ;;. pusillus (B, Haiti), P. p. portoricensis (C, Puerto Rico), P. p. parneUii ( D, Jamaica; E, Cuba),
P. p. mexicanus (F, Sinaloa), P. p. mesoamericanus (G, Yepocapa, Guatemala), P. p. fitsciis (H, Aroa,
Venezuela), P. p. ruhiginosiis (I, Bonanza, Nicaragua; J, Maripa, Venezuela). All specimens are
adult males; horizontal line above A is 10 mm in length.

Comparisons. — For comparison with
P. p. portoricensis see account of that
subspecies. From both P. p. pusillus and
P. p. gonavensis, P. p. parneUii is easily
distinguished on the basis of larger ex-
ternal and cranial size (Table 1). In
addition, the rostra of both pusillus and
gonavensis are relatively shorter and
broader than that of parneUii (Fig. 23).

Remarks. — J. Knox Jones, Jr., kindly
examined the syntypes of Chilonijcteris
osburni Tomes, 1861; although the
length of the forearm (52.3 in each) is
slightly smaller than usual, they are
clearly within the range of variation of
Jamaican P. p. parneUii.

There is nothing in Gundlach's
(1861:154-155) description of the Cu-

ban Chilonycteris boothi that clearly
diagnoses that taxon. Peters (1872:360)
and Dobson (1878:452) synonymized
both osburni and boothi under Chilo-
nycteris parneUii. On the other hand,
Miller (1902:401) compared portoricen-
sis with both parneUii and boothi, which
he recognized as distinct species. Miller
alleged that the lower premolars were
so crowded in parneUii that the second
premolar was excluded lingually from
the toothrow, and the first and third pre-
molars were in contact, but that crowd-
ing was not particularly apparent in
either boothi or portoricensis; the latter
was further distinguished on the basis
of shorter ears. Miller (1904:341-342)
reiterated the differences between Cu-

SMITH: SYSTEMATICS OF MORMOOPID BATS

67

ban boothi and Jamaican parnellii, but
also stated that otherwise the two spe-
cies were "closely similar." Rehn (1904:
197) was the first to apply boothi as a
trinomen. In addition to Miller's char-
acteristics of the lower premolars, Rehn
further distinguished the Cuban bats by
a "more depressed rostrum and the more
robust form." After studying these two
insular populations, it is apparent to me
that the degree of crowding of the
lower premolars is variable and that
there are no marked differences in the
"depression" of the rostrum. Therefore,
because of the great degree of overlap
in external and cranial dimensions
(Table 1), I regard Chilonijcteris boothi
Gundlach, 1861, as a junior synonym of
P. p. parnellii Gray, 1843.

Specimens examined (62). — CUBA. Ha-
hana: Cueva de Cotilla, 4 nii S San Jose de
Las Lajas, 2 (AMNH); Cueva de Numancia,
Aguacate, 2 (AMNH). Las Villas: Cueva del
Agua, Sagua La Grande, 5 (ROM); Cueva de
Los Masones, Trinidad, 1 (ASC). Oriente:
Cueva de Las Majaes, Caney, 6 ( 1 AMNH, 5
ROM); Baracoa, 4 (USNM). Pinar del Rio.
San Vicente, 1 (ASC); Pan de Gnajaihon, 1
(AMNH). JAMAICA. Hanover: Lucca, 3
(AMNH). St. Catherine: Healthshire Hills
Caves, 1 (AMNH). St. Elizabeth: Walling-
ford Cave, near Balachva, 4 (AMNH); Oxford
Cave, near Balaclava, 10 (AMNH); Perti
Cave, near Goshen, 1 (AMNH); Sf. Clair
Cave, 12 (ROM). Trelawmj: Windsor Cave,
7 (2 AMNH, 5 ASC). No specific locality, 2
(USNM).

Pteronotus parnellii pusillus

(G. M. Allen)

Chilonijcteris parnellii pusillus G. M. Allen,
1917, Proc. Biol. Soc. Washington, 30:168,
23 October.

Holotype. — Adult female, skin and
skull, MCZ 16468; from Arroyo Salado,
Dominican Republic, Hispaniola; ob-
tained by James L. Peters on 7 March,
1916.

Distribution. — Haiti and Dominican
Republic (Fig. 22).

Description. — Size, smallest of races
of parnellii definitely still living, both
externally and cranially. Standard ex-
ternal measurements for an adult female

from Cueva Duran, 4 km SW Moncion,
Dominican Republic, are as follows:
total length, 73; length of tail vertebrae,
20; length of hind foot, 12; length of ear
from notch, 18; length of forearm of
three adult females, 48.9, 50.2, and 50.5.
Additional selected cranial measure-
ments of combined material from Haiti
are given in table 1 (locality 64).
Viewed in profile, the forehead of pu-
sillus rises more abruptly and the brain-
case is shorter and more globular than
in other Antillean races, except P. p.
gonavensis (Fig. 23).

The color of the pelage is Snuff
Brown dorsally becoming paler on the
head and nape. Like P. p. parnellii
there is a discernible central zone on the
individual filaments of hair from the
back. Ventrally, the pectoral and ab-
dominal regions are brown with a drab
wash, becoming paler on the throat.

Comparisons. — This race is easily
distinguished from all other subspecies
(mainland and insular) except P. p.
gonavensis, purely on the basis of exter-
nal and cranial size. From P. p. gona-
vensis, the subspecies pusillus is recog-
nizable by larger cranial size (Table 1,
see especially length of maxillary tooth-
row, and condylobasal length ) . Further-
more, pusillus has a relatively longer,
narrower rostrum (Fig. 23).

Remarks. — The distinctness of P. p.
pusillus and P. p. gonavensis from all
other races of parnellii is represented by
a distance coefficient of 7.800 (Fig. 20).
The magnitude of this coefficient, in
comparison to those between other pop-
ulations of parnellii, is so great that
pusillus and gonavensis might well rep-
resent distinct species. The relation-
ships between these two taxa and the
larger Antillean races (parnellii and por-
toricensis) are not clearly understood
and, as will be discussed beyond, the
evolutionary and zoogeographic rela-
tionships of the larger subspecies also
are poorly known. In view of these con-
siderations, and because pusillus and
gonavensis resemble parnellii morpho-

68

MISCELLANEOUS PUBLICATION MUSEUM OF NATURAL HISTORY

logically, I here retain them as sub-
species of the latter.

Specimens examined (15). — HAITI. San
Michel, 6 (USNM); Diquini, 1 (USNM); San
Raphael, 1 (USNM). DOMINICAN REPUB-
LIC. Maria Trinidad Sanchez: Arroyo Salado,
2 (MCZ). Samana: Cave near Sanchez, 1
(USNM). San Juan: Rancho La Guardia, 1
(AMNH). Santiago Rodriguez: Cueva Du-
ran, 4 km SW Moncion, 3 (1 AMNH, 2
ROM).

Pteronotus parnellii gonavensis
(Koopman)

Chilonycteris parnellii gonavensis Koopman,
1955, Jour. Mamm., 36:110, 25 February.

Holotijpe. — Cranium only, lacking
incisors and auditory region, USNM
295009; from an owl pellet deposit found
in a cave near En Cafe, La Gonave
Island, RepubHc of Haiti, Hispaniola;
collected by A. J. Poole and W. M. Per-
rygo in March 1929.

Distribution. — Known only from type
locality as a sub-Recent fossil ( Fig. 22 ) .

Description. — Size, cranially the
smallest known subspecies of P. par-
nellii; see table 1 (locality 63) for se-
lected cranial measurements. Rostrum
noticeably short (Fig. 23); forehead
rising abruptly over short, rounded
braincase as in P. p. pusillus.

Comparisons. — As mentioned in the
previous account, P. p. gonavensis is
easily distinguished from all other races
of parnellii, except pusillus, solely on
the basis of being smaller cranially.
From pusillus, it differs in having a
shorter, broader rostrum, and generally
shorter, slightly narrower skull (Fig.
23).

Remarks. — This subspecies is based
entirely on subfossil material from an
owl pellet deposit found in a cave on
La Conave Island, which is located in
the Gulf of Gonaives off western His-
paniola. This subspecies is closely re-
lated to the Hispaniolan subspecies pu-
sillus. In addition to cranial remains,
Koopman (1955:112) reported the pres-
ence of pusillus-\ike radii in the En Cafe
material. These wing elements were,

however, slender in comparison to those
of pusillus from St. Michel, Haiti. A
case could be made for elevating pusil-
lus to specific rank, in which event
gonavensis should remain associated
with pusillus as a distinct subspecies.

Specimens examined (11). — HAITI. Cave
near En Cafe, La Gonave Island, 11 (USNM).

Pteronotus parnellii portoricensis

(Miller)

Chilomjcteris portoricensis Miller, 1902, Proc.
Acad. Nat. Sci. Philadelphia, 54:400, 12
September.

Holotijpe. — Adult female, skin and
skull, USNM 102358; from cave near
Pueblo Viejo [Cueva di Fari], Puerto
Rico; obtained by L. Stejneger and
C. W. Richmond.

Distribution. — Known only from
Puerto Rico (Fig. 22).

Description. — Size, only slightly
smaller, both externally and cranially,
than P. p. partiellii from Cuba and Ja-
maica. Standard external measurements
of two males from Cueva Trujillo Alto
are as follows: total length, 78, 83;
length of tail vertebrae, 20, 20; length
of hind foot, 12, 13; length of ear from
notch, given as 19 and 21 by Miller
(1902:401); see table 1 (locality 65) for
additional selected external and cranial
measurements. It is impossible to de-
scribe the color of the pelage of this
subspecies because most of the material
examined was preserved in spirits and
badly bleached. Miller (1902) and
Rehn ( 1904 ) suggest that the color is
not greatly different from that of P. p.
parnellii.

Comparisons. — From both P. p. pu-
sillus and P. p. gotuivensis, this subspe-
cies can be recognized by larger cranial
size (Fig. 23). The differences between
portoricensis and parnellii of Cuba and
Jamaica are less striking in that por-
toricensis averages only slightly smaller
than parnellii (Figures 19, 23, and
Table 1).

Remarks. — In his description of
Chilonycteris portoricensis, Miller (1902:

SMITH: SYSTEMATICS OF MORMOOPID BATS

69

401-402) noted its close resemblance to
CJuIonycteris hoothi (=:P. p. parnellii)
from Cuba; the former being charac-
terized by short ears. However, there is
nothing in my data to suggest any dif-
ferences in the length of the ears of
these two geographically separated
populations.

Several points must be considered in
order to interpret the observed simi-
larity between parnellii and portoricen-
sis. The problem would be less com-
plex if Cuba, Jamaica, and Puerto Rico
were adjacent islands. However, the
island of Hispaniola, through which
gene How should logically pass, is geo-
graphically interposed between Puerto
Rico and the other two islands. Fur-
thermore, the Hispaniolan subspecies
piisiUus as well as ii,onavemis are nota-
bly dissimilar from all other Antillean
races of P. parnellii (Fig. 21) and can-
not be construed to represent interme-
diates in a continuum between the pop-
ulations of the islands in question. It is
of course possible, owing to an inade-
quate knowledge of the Hispaniolan bat
fauna, that there is an as yet undis-
covered population of more typical par-
neUii-like bats on Hispaniola. On the
other hand, the similarities between
parnellii and portoricensis may have
been independently acquired, or per-
haps these two represent the primitive
condition of the Antillean parnellii stock
from which pusillus was derived. Until
the situation is better understood, I
elect to recognize the Puerto Rican pop-
ulation as a distinct subspecies.

Specimens examined (7).— PUERTO RICO.
Pueblo Viejo, 1 (USNM); Cueva di Fari near
Pueblo Viejo, 4 (USNM); Cueva de Trujillo
Alto, 2 (AMNH).

Pteronotus parnellii mexicanus (Miller)

Chilontjcteris mexicana Miller, 1902, Proc.

Acad. Nat. Sci. Philadelphia, 54:401, 12

September.
Pteronotus par7wlUi mexicana Hall and Dal-

quest, 1963, Univ. Kansas Publ. Mus. Nat.

Hist., 14:217, 20 May.

Holotype. — Adult male, skin and
skull, USNM 89277; from San Bias,
Nayarit; obtained by E. W. Nelson and
E. A. Goldman on 9 June, 1897; original
no. 11132.

Distribution. — Mexico, north of the
Isthmus of Tehuantepec along the Paci-
fic lowlands to central Sonora (Cueva
de la Tigre, 15 mi SSE Carbo, Cockrum
and Bradshaw, 1963:3) and north along
the Gulf coast to southern Tamaulipas
( Sierra de Tamaulipas, 2 mi S, 10 mi W
Piedra); also known from subtropical
regions of the plateau states of Chihua-
hua, Durango, and San Luis Potosi ( Fig.
22).

Description. — Size small for a main-
land race of the species, both externally
and cranially. Standard external meas-
urements (average, with extremes in
parentheses) of eight males and five
females, respectively, from 2 mi SE
Jalcocotan, 3000 ft, Nayarit, are as fol-
lows: total length, 89.2 (88.0-94.0),
89.9 (83.0-90.0); length of tail vertebrae,
21.6 (20.0-24.0), 21.0 (19.0-22.0); length
of hind foot, 13.6 (13.0-14.0), 13.8 (13.0-
14.0); length of ear from notch, 22.0
(21.0-23.0), 21.6 (21.0-22.0). See table
1 (especially localities 1, 6, and 7), for
additional selected external and cranial
measurements.

The pelage of mainland subspecies
of Pteronotus parnellii is less distinctly
tricolored dorsally than that of insular
races. A nearly monocolored pelage re-
sults primarily from a weak expression
of the whitish central zone on the hairs
of the back; in P. p. mexicanus, this
central zone is Cinnamon-Buff with
basal and terminal portions brownish.
Specimens examined from near the type
locality and from western Mexico are
noticeably paler (Snuff Brown) in color
than those of other mainland races. The
head and nape are the same color.
Ventrally, the pectoral and abdominal
regions are pale drab gray with a buffy
tinge, becoming slightly paler on the
throat and chin.

Comparisom. — For comparison with

70

MISCELLANEOUS PUBLICATION MUSEUM OF NATURAL HISTORY

the adjacent subspecies in Middle Amer-
ica, see the following account. From the
Antillean races P. p. parnellii and P. p.
portoricensis, the subspecies mexicanus
is distinguishable by its larger size, both
externally and cranially (Table 1), and
less strikingly tricolored pelage. The
breadth and length of the skull are
greater in mexicanus than in any of the
Antillean populations (Fig. 23).

Remarks. — Unlike insular areas,
zones of intergradation are more easily
discernable on the mainland where con-
tinuous gene flow obtains in the absence
of barriers. The typical population of
P. p. mexicanus inhabits western Mex-
ico; bats of this race become progres-
sively larger (Table 1) and darker as
they approach the region of the Isthmus
of Tehuantepec. A rather narrow zone
of intergradation seems to exist in this
region between mexicanus and the sub-
species in southern Mexico and northern
Middle America. Trans-isthmian resem-
blances, based on distance coefficients
(Fig. 21), are greater in the north and
central parts of the Isthmus than along
the southern. Pacific portion. In fact,
bats from the Sierra de Los Tuxtlas re-
semble those from Guatemala to such
an extent as to justify their assignment
to the same subspecies. The large dis-
tance coefficient (2.077) between locali-
ties 19 and 21 (Fig. 21) may reflect the
relatively more arid conditions which
could constitute a partial barrier in that
part of the Isthmus. The northern part
of the Isthmus may act less like a bar-
rier to gene flow due to the more mesic
conditions in that region. The influence
of such gene flow seems to extend along
the eastern coast of Mexico with repre-
sentatives of mexicanus being slightly
larger and somewhat darker than those
in western Mexico. The darkness of
Tamaulipan specimens described by Al-
varez (1963:398) is related to seasonal
change of color.

As mentioned previously, small size
and pale color probably relate to some,
as yet poorly understood, factors of the

suboptimal environment in which some
parnellii live. In addition, the composi-
tion of the insectivorous chiropteran
fauna and intraspecific competition may
have influenced the evolution of P. p.
mexicanus in western Mexico as well as
the Antillean races. In the Pacific low-
lands of Sinaloa, there are 20 to 25 spe-
cies of insectivorous bats, including
other mormoopids (Pteronotus persona-
tus, P. claviji and Mormoops megalo-
phylla). The majority of these species
are small, slow-flying bats that probably
do not directly compete with P. p. mexi-
canus (or other mormoopids, for that
matter). However, at least eight (pri-
marily molossids and lasiurines) are
large, swift-flying bats that may com-
pete directly with P. p. mexicanus and
other mormoopids. In this arid sub-
tropical habitat, where environmental
stresses may be more stringent on prey
and predator, the more temperate spe-
cies may have a selective advantage over
the mormoopids that also inhabit the
region. Therefore, in such regions,
smaller size in P. parnellii could be in
response to this competitive pressure.
In more tropical environments, although
the total number of insectivorous bats
may increase, there are comparatively
fewer large, swift-flying species.

Spccimei^s examined (460). — MEXICO.
Chihuahua: Carimeche, Rio Mayo, 10
(UMMZ); 1.5 mi SW Tocuina, 1500 ft, 1
(KU). Colima: Plava de Oro, 8 km W, 2 km
S Santiago, 1 (KU); 3 mi S Santiago, 1 (KU);
7 mi W, 0.5 mi S Santiago, 15 (KU); 4 km S
Coquimatldn, 5 (LACM); Pueblo Juarez, 6
(LACM); Rancho Tahamillas, 6 km N Agua
Zarca, 6 (IB); North Slope La Media-Luna, 8
(LACM). Distrito Federal: Wards Natural
History Establishment (presumably Tacubaya,
D. F.), 2 (USNM). Durango: Santa Ana (12
km E Cosala, Sinaloa), 1300 ft, 2 (KU).
Guerrero: 12 km N Zacatula, 6 (ENCB);
Puente de Dios, Yerbabuena, 1700 m, 16
(KU); Ctieva Laguna Honda, Yerbabuena, 12
km NNW Teloloapan, 1800 m, A (IB); Ojo de
Agua de Mexicapan, 4.3 km N Teloloapan,
1480 m, 4 (IB); Puente de Los Sabinos, 16
km E Teloloapan, 1040 m, 5 (IB); Ojo de
Agua de Chapa, 5 km SE Teloloapan, 1520
m, 14 (IB); 10 mi E, 2 mi S Teloloapan, 1
(KU); Grutas Cacahuamilpa, 4 (ENCB); 3

SMITH: SYSTEMATICS OF MORMOOPID BATS

71

I

mi W Mexcala, 2000 ft, 3 (TCWC); 3 km N
Agita de Obispo, Chilpancingo, 1 (KU); 1 mi
N, 6 mi E Ttixtla, 2 (TCWC); Grittas de
Juxlahitaca, 5 km N Colotlipa, 800 m, 3
(TCWC); 4 km NE Colotlipa, 1100 m, 11
(ENCB); 2 mi NW Acapulco, 50 ft, 8 (KU).
Jalisco: Bolanos, 1 (USNM); 1.5 mi WNW
Amatitan, 4100 ft, 1 (KU); 15 km W Ameca,
4200 ft, 3 (KU); 10 mi SE Talpa de Allende,
5350 ft, 1 (KU); El Zapote, 3 (IB); 5 mi W
Chapala, 5000 ft, 13 (KU); 4 mi SE Ocodan,
4 (MCZ); 0.5 km N, 3 km W Jammj, 1650 m,
26 (7 ENCB, 19 IB); 5 mi S Gnillo, 3100 ft,

1 (KU); 6 77n E Limon, 2700 ft, 3 (KU); 15
km NW Cihuatlan, 3 (KU). Mexico: 7.5 km
S, 10 km E Valle de Bravo, 2100 m, 2
(ENCB); Cueva del Coyote, 6 km SSE Tona-
tico, 1500 m, 16 (AMNH). Michoacdn: 5 km
SW Turandeo, 1900 m, 7 (ENCB). Morelos:
Cueva Poza de Montezuma, 1 km E Oaxtepec,
1450 m, 32 (IB); 0.5 mi NW Huajintlan,
3410 ft, 2 (TCWC). Natiarit: San Bias, 7 (1
MVZ, 6 USNM); 8 mi E San Bias, 7 (LACM);

2 mi SE Jalcacotdn, 3000 ft, 13 (KU); 6 mi
SSE Las Varas, 3 (KU). Oaxaca: 3 km WNW
Dominguillo, 730 m, 5 (KU); 13 mi E Juchi-
tan, Isthmus of Tehuantepec, 1 (LACM);
Tehuantepec, 1 (USNM); 6 mi NW Mixte-
quilla, 8 (TCWC); 18 mi N Matias Romero,
1 (LACM); 8 km NW Salitm Cruz, 11
(ENCB); Huilotepec, 1 (USNM). Puehla:
1 mi E Raboso, 4350 ft, 1 (KU); 1 7ni SE
Rahoso, 14 (KU). Sinaloa: 12 mi NE San
Benito, 1000 ft, 1 (KU); 20 km N, 5 km E
Badiraguato, 1800 ft, 1 (KU); 23 km W
Pericos, 1 (LACM); 4 mi S Casa Blanca, 1200
ft, 1 (KU); 3 mi SE Camino Real, 500 ft, 10
(KU); 1.5 km W Copala, 1400 ft, 4 (KU);
Pdnuco, 22 km NE Concordia, 2050 ft, 33
(KU); 1 mi E Santa Lucia, 3650 ft, 2 (KU);
Rancho Batel, 6 7»/ E Santa Lucia, 6300 ft, 2
(LACM); 1.5 km S Santa Lucia, 3200 ft, 11
(KU); 2.5 mi NE Concordia, 1 (KU); 1.5 mi
E Concordia, 250 ft, 2 (MSU); 12 mi N Villa
Union, 400 ft, 3 (KU); 15 mi N Rosario,
Chele, 300 ft, 1 (UMMZ); 5 mi WSW Plomo-
sas, 800 ft, 3 (KU); 0.5 mi S Concepcion,
250 ft, 1 ( KU ) . Sonora: 1 mi E Buena Vista,
on Rio Yaqui Reservoir, 1000 ft, 1 (KU);
Tesia, between Navajoa and Camoa, 7
(AMNH); Alamos, 1 (ROM); Prothomtorio
Mine, 3 mi W La Aduana, 1 (LACM); 0.25
mi W Aduana, 600 ft, 3 ( KU ) ; Aduana, 4 mi
SW Alamos, 2 (CAS); no specific locality, 1
(KU). Tamaulipas: Sierra de Tamaulipas, 2
mi S, 10 mi W Piedra, 1200 ft, 1 (KU); Sierra
de Tamaulipas, 3 mi S, 16 mi W Piedra, 1400
ft, 3 (KU); 10 km N, 8 km W El Encino, 400
ft, 22 (KU); Rancho Pano Ayuctle, 25 mi N
El Mante, 3 km W Pan American Highway,
300 ft, 3 (KU); Ojo de Agua, 20 mi N El
Mante, 3 km W Pan American Highivay, 300
ft, 2 (KU); 7 mi W Ocampo, 2400 ft, 1 (KU).

Veracruz: Mirador, 2 (USNM); 8 km NW
Poterero, 1700 ft, 1 (KU); 38 km SE Jesus
Carranza, 500 ft, 2 (KU).

Pteronotus parnellii mesoamericanus

new subspecies

Holotype. — Adult male, skin and
skull, KU 64713; from 1 mi S and 0.75
mi E Yepocapa, 4280 ft, Chimaltenango,
Guatemala; captured by James W. Bee
and J. A. Ibarra on 7 January 1955;
original no. 550107-31 of Bee.

Distribution. — Mexico south of the
Isthmus of Tehuantepec, including Sier-
ra de Los Tuxtlas and the Yucatan
Peninsula, westward into Guatemala,
British Honduras, and El Salvador to
central Honduras (west of Sierra de
Esperanza ) , thence southward along the
Pacific versant of Nicaragua and Costa
Rica to the Province of Code, west of
the Cordillera de Talamanca, in western
Panama (Fig. 22).

Description. — Size medium, both
cranially and externally, for mainland
races of parnellii. Standard external
measurements for the holotype, followed
by the average (extremes in paren-
theses) for 11 males and 29 females re-
spectively, from the type locality are as
follows: total length, 96, 95.6 (93.0-
98.9), 95.9 (93.0-98.0); length of tail
vertebrae, 26, 24.5 (23.0-26.0), 24.5
(23.0-27.0); length of hind foot, 14 for
all specimens of series; length of ear
from notch, 22, 21.8 (21.0-22.0), 21.5
(20.0-22.0). Additional selected exter-
nal and cranial measurements are given
in table 1 (see especially localities 27
and 33).

Individuals in the series from the
type locality, all obtained in early
January, are not in fresh pelage; all are
dark brown (near Prout's Brown) dor-
sally. There is no marked change in
the color of the pelage on the nape and
head. Ventrally, in the pectoral and
abdominal regions, the basal portion of
the hairs are nearly Prout's Brown giv-
ing an overall dark brown appearance
to the pelage; the tips of the hairs are

72

MISCELLANEOUS PUBLICATION MUSEUxM OF NATURAL HISTORY

pale Cinnamon-Buff and give a "frosted"
appearance to the venter; the throat and
chin regions usually are somewhat paler.
Specimens captured in early March
from near Chinaja and Toocog, Peten,
Guatemala, are in fresh pelage and are
somewhat darker than those from the
type locality. Three December-taken
males from Puente Punta Gordo, Guate-
mala, are brilliant Ochraceous Tawny in
color, both dorsally and ventrally.

Comparisons. — From P. p. mexican-
us, the new subspecies can be distin-
guished by larger external and cranial
size (Table 1). This is especially true
in comparison with specimens of mexi-
canus from the northern parts of its
range; individuals of mexicanus are gen-
erally paler in color.

Compared with P. p. riihiginosus,
P. p. mesoamericanus averages smaller
in size, both cranially and externally.
The color of mesoamericanus, although
darker on the average than that of mexi-
canus, is paler compared to the pelage
of ruhiginosus. In the zone of inter-
gradation between mesoamericanus and
ruhiginosus — eastern Honduras, south-
ward more or less along the continental
divide — these two races become increas-
ingly more difficult to distinguish. As a
rule, bats from the Pacific versant, al-
though increasing somewhat in overall
size in Panama, are usually smaller and
paler than bats of the Caribbean coastal
lowlands, herein assigned to ruhiginosus.

Remarks. — Specimens assigned to
this new subspecies have been variously
referred to in the past as Chilonycteris
parnellii fusca (Koopman, 1955; Hall
and Kelson, 1959; Ryan, 1963), Chilo-
mjcteris ruJnginosa fusca ( Hershkovitz,
1951; Felten, 1956), and Pteronotus par-
nellii fuscus (Jones, 1966 and Jones, et
al. 1971). The relationship between
fuscus and mesoamericanus will be con-
sidered in more detail below.

Typical populations of mesoameri-
canus are found in south-central Guate-
mala, and mesoamericanus gradually in-
tergrades with mexicanus in northern

Chiapas, Tabasco, and southern Vera-
cruz. As was mentioned earlier, gene
flow between these two races evidently
decreases, probably owing to adverse
ecological conditions, across the south-
ern Pacific coast of the Isthmus of Te-
huantepec (see figure 21 for distance
coefficients). In a similar fashion,
southern populations of mesoamericanus
become involved in a zone of inter-
gradation with ruhiginosus. These
trends are normal and expected in con-
tinuous, freely interbreeding popula-
tions of the same species. On the Yuca-
tan Peninsula there is a marked reduc-
tion in external and cranial size at
progressively more northern localities
which can easily be seen by comparing
localities 20, 15, and 18 (Table 1). Fur-
thermore, it will be noted that localities
15, 16, and 17 become increasingly more
distinct from locality 18. This trend
seems to be related to the increasingly
suboptimal or xeric conditions on the
northern part of the peninsula. Perhaps
ecological factors that influence this
"peninsular effect" are similar and in-
dependently involved in the small size
and pale color of typical mexicanus in
western Mexico. Another instance in
which reduced size and pale color seem
to be correlated with xeric conditions
occurs in P. p. fuscus from Margarita
Island off the northeastern coast of
Venezuela.

There are several important aspects
concerning the zone of intergradation
between mesoamericanus and ruhigino-
sus. First, geographically distinct popu-
lations of P. parnellii appear to be segre-
gated into east and west versants in
Middle America. This situation is by
no means novel, having been recognized
by herpetologists for some time; recently
Davis (1968,^1969, and 1970) considered
geographic variation in Vroderma and
ArfiJ)eus from this point of view. Ryan's
( 1963 ) Biotic Provinces of Central
America, based on mammalian distri-
butional patterns, also reflect Pacific
and Caribbean versants. The distinct-

SMITH: SYSTEMATICS OF MORMOOPID BATS

73

ness of populations occupying these ver-
sants no doul)t depends a great deal on
the ecological tolerance of the particu-
lar species. In the case of P. parnellii,
the versants seem to have had a mod-
erate effect in the differentiation and
isolation of mesoamericaniis and ruhig-
inosiis.

Secondly, the size and shape of the
landmass in relationship to the ecology
and proximity of the two versants to
each other seems to play an important
role in their overall effectiveness. The
Middle American landmass is widest in
southern Honduras and northern Nica-
ragua, and gradually becomes wedge-
shaped in northern Costa Rica. From
this point southward to northern South
America, the landmass is extremely nar-
row. The transverse mountain masses
in central Honduras and northern Nica-
ragua seem to form a partial barrier be-
tween mesoamericanus to the north and
west and ruhi^inosus to the southeast.
However, south of this mountain mass,
the versants with their corresponding
populations of P. parneUii are separated
only by a broad ecotonal belt and in
some places by mountains. Gene ex-
change may take place more or less
freely across this landmass. The small
size of mesoamericanus in northwestern
and central Nicaragua (as compared to
the larger rubiginosus of the Miskito
lowlands of eastern Nicaragua) seems
to be an indication of the relative ef-
fectiveness of the broad ecotone in main-
taining the two subspecies where inter-
posed mountainous areas are lacking.
As the landmass becomes increasingly
narrower there is a concurrent reduc-
tion in the breadth of the ecotone and
the two versants come into closer prox-
imity. The overall effect is the develop-
ment of a broader zone of intergrada-
tion in which both versant populations
of P. panieUii become gradually larger
in size, but with mesoamericamis always
averaging slightly smaller.

In eastern Panama and northwestern
Colombia (the area from the Canal

Zone to the Bolivar Geosynchne) the
zone of intergradation between meso-
americanus and rubiginosus becomes
more complex with the incorporation of
gene flow from western populations of
P. p. fiiscus. Phenotypically, ruhigino-
sus dominates the intergradation in this
region. Pteronotus p. rubiginosus is con-
tinuous from eastern Nicaragua to the
Amazon Basin and effectively separates
P. p. fuscus and P. p. mesoamericanus,
which are similar in appearance and in-
deed may be closely related (see ac-
count of P. p. fuscus).

Dunn (1940:156), reporting on the
herpetofauna of Panama, noted what he
called a "faunistic crossing over" in the
Canal Zone and Darien. This idea was
founded on apparent affinities between
the fauna that occurs in the Pacific ver-
sant of Panama and that of the Carib-
bean coast of South America; the fauna
of the wetter Atlantic side (Caribbean
versant) reportedly had its closest af-
finities with the fauna of the Pacific coast
of Colombia. Although, this apparent
"crossover" has not yet been clearly de-
fined for mammals, it is strongly sug-
gested in this analysis of geographic
variation of P. parneUii. A somewhat
similar situation seems to exist in Davis's
(1968, 1969, and 1970) data for Uro-
cJernui and Artibeus. I suspect that with
further investigation this pattern of dis-
tribution may be found to be relatively
common among bats as well as other
small mammals.

Haffer (1967a, 1967b and 1970), in
his study of the zoogeography of the
Colombian bird fauna, discussed the
probable effects of climatic changes
during the Pleistocene on the tropical
forests of Central America and northern
South America. With alternating wet
glacials and drv interglacials, Haffer
suggested an alternating expansion and
contraction of tropical forests in this
vast region. In eastern Panama and
northwestern Colombia, each of these
circumstances would have allowed for a
differential continguity between Middle

74

MISCELLANEOUS PUBLICATION MUSEUM OF NATURAL HISTORY

American populations of P. parnellii
and those of South America and thus a
differential in gene flow at different
times. During dry climatic periods,
when the tropical forest contracted into
respective Middle and South American
nuclei, gene flow between populations
of P. parnellii from the Pacific versant
of Middle America and the Caribbean
coast of northern South America no
doubt prevailed through the narrow
causeway in eastern Panama. On the
other hand, with wet climatic periods
and resultant expansion of tropical for-
est, the above-mentioned gene flow
could have been cut off or at least re-
placed by panmixia between popula-
tions of P. parnellii from the Caribbean
versant of Middle America and those of
the Amazon Basin.

The subspecific name mesoameri-
canus was selected to describe the geo-
graphic region of Middle America in
which bats of this subspecies live.

Specimens examined (517). — MEXICO.
Campeclie: 5 km S Champoton, 1 (KU); 1
km N, 13 km W Escarcega, 65 m, 7 (KU);
12 km W Escarcega, 7 (KU); 7.5 km W Es-
carcega, 65 m, 3 (KU); 105 km E Escarcega,

1 (KU); 103 km SE Escarcega, 7 (KU); 65
km S, 128 km E Escarcega, 4 (KU). Chiapas:
10 km N, 1.5 km E Raudales, 150 m, 3
(ENCB); 16 mi NW Palenque, 100 ft, 1
(TCWC); Riiinas de Palenque, 140 in, 4
(ENCB); 21 km WSW Teapa, 200 ft, 1
(TCWC); approx. 8 km S SoIosucJiiapa, 400
ft, (Camp on Rio Teapa), 1 (LACM); 12 mi
W Mai Paso, 400 ft, 2 (TCWC); 7 mi WSW
Ocozocoautla, 2500 ft, 2 (TCWC); Cerro
Hiieco Cave, 2 mi SE Tuxtla Gutierrez, 266
ft, 5 (TCWC); 2 mi S Tuxtla Gutierrez, 2
(MCZ); 4 mi NE Chiapa de Corzo, 3000 ft,
6 (TCWC); 6 mi SE San Cristobal, 7300 ft,

2 (MVZ); Cueva de Zapaluta, 1.3 mi SE
Zapaluta, 5700 ft, 53 (TCWC); 2 km NW San
Lucas, 3300 ft, (33 km SE Zapaluta), 3
(TCWC); 7 km W Los Amates, 1800 ft, 5
(TCWC); Finca Oquilapa, 8 mi SE Tonala,
100 ft, 3 (LACM); 15 km SE Tonohi, 100 ft,
6 (TCWC); 5 km SE Pijijiapan, 100 ft, 5
(TCWC); 11 km NW Escuintla, 100 ft, 4
(LACM). Quintana Boo: Aeropuerto, 4 km
WSW Puerto Juarez, 5 m, 1 (KU). Tabasco:
2 mi SW Teapa, 1 (KU); 5 mi SW Teapa, 9
(KU). Veracruz: Rio Basura, 3 km N Zonte-
comapan, 2 (ENCB); Zontecomapan, 4
(ENCB); 2 mi WNW San Andres Tuxtla,

1000 ft, 3 (TCWC); J mi N San Andres
Tuxtla, 8 (TCWC); Cueva Laguna Encan-
tada, 3 km ENE San Andres Tuxtla, 1000 ft,
1 (TCWC). Yucat/m: Gruta de Balenkanche,
5 km E Chichen Itza, 9 (KU). BRITISH
HONDURAS. Augustine, 1 (ROM); Rock-
stone Pond, 2 (ROM). GUATEMALA. Alta
Verapaz: Rio San Simon, 6 km NE Raxruja,
140 m, 2 (TCWC); Lanquin Cave, 1022 ft,
4 (KU); 1 mi W Lanquin, 2000 ft, 1
(TCWC); 3 km ENE San Cristobal Verapaz,
1370 m, 22 (TCWC). Chimaltenango: Rio
Queleya, 0.5 mi E Yepocapa, 4300 ft, 4 ( KU ) ;
1 mS, 0.75 mi E Yepocapa, 4230 ft, 41 (KU).
Chiqidmula: Jocotan, near Chiquimula, 1350
ft, 1 (KU). Escuintla: San Luis, Escuintla, 2
(LIMMZ). Huehuetenango: La Democracia,
3300 ft, 8 (TCWC). Isabal: 22 km SSW
Puerto Barrios, 200 ft, 11 (TCWC); Quebra-
das, 1000 ft, 8 (FMNH); Puente Punta Gordo,
km 51 on Gnat. City-Puerto Barrios Hwy.,
2050 ft, 3 (KU). Jutiapa: Finca el Carnero,
Jutiapa, 8 (UMMZ). Peten: Toocog, 15 km
SW La Libertad, 540 ft, 2 (KU); 12 km
NNW Chinaja, 5 (KU). HONDURAS. At-
lantida: 7 mi E La Ceiba, 100 ft, Atlantida, 3
(TCWC); Balfate, 2 (FMNH). Comat/agua:
Comayagua, 580 m, 3 (TCWC); 3 km W
Comayagua, 580 m, 4 (TCWC); 1 km W
Comayagua, 580 m, 1 (TCWC); 17 km S
Comayagua, 580 m, 1 (TCWC). Copdn: 6
km ESE Copan, 900 m, 7 (TCWC). Cortes:
8 mi SW San Pedro Sula, 500 ft, 2 (TCWC).
Intibucd: 5 km NE Jesiis de Otoro, 640 m, 1
(TCWC). Lempira: Las Floras, approx. 10
km N Gracias, 1850 m, 32 (AMNH). Oco-
tepeque: Nueva Ocotepeque, 3 (TCWC).
Santa Barbara: 2 km S San Nicolas, 600 m, 7
(TCWC). EL SALVADOR. Cuscatldn: 2
km W Suchitoto, 390 m, 2 (TCWC). La
Libertad: 20 km W La Libertad, 250 m, 2
(TCWC); 16.8 km W La Libertad, 30 m, 1
(TCWC); 13.1 km W La Libertad, 15 m, 1
(TCWC); 3.5 km E La Libertad, 15 m, 1
(TCWC); 5.7 km E La Libertad, 15 m, 14
(TCWC). La Union: Tabanco, 650 ft, 3
(MVZ). Morazan: Gigante Mine, approx. 6
mi NE Divisadero, 3 (MVZ); 2 mi N Divisa-
dero, 1 (MVZ); Graveyard Mine, 1.5 mi NW
Divisadero, 600 ft, 3 (MVZ); Encuentros
Mine, 3 mi W Divisadero, 700 ft, 16 (MVZ);
Monte Cristo Mine, 1.5 mi W Divisadero, 750
ft, 2 (MVZ); San Jose Mine, approx. 1 mi W
Divisadero, 1 (MVZ); Misericordia Mine, 1
(MVZ). San Miguel: Comacaran, 700 ft, 2
(MVZ). NICARAGUA. Chinandega: 6.5 km
N, 1 km E Cosiguina, 10 m, 1 (KU). E.steli:
2 km S Esteli, 3300 ft, 3 (TCWC); 50 mi N
Sebaco, 1 (AMNH). Matagalpa: 2 mi SE
Dario, 1500 ft, 4 (TCWC). Rio San Juan: 1
km S El Castillo, 130 m, 1 (TCWC). COSTA
RICA. Guanacaste: 9 km N Liberia, 4 km E
Pan American Highway, 12 (LACM); Rio Te-

SMITH: SYSTEMATICS OF MORMOOPID BATS

75

noiia, 3 mi S, 10 mi W Las Canas, 10 ft, 3
(TCWC). Puiitarenas: 9 mi ENE Puerto
Golfito, 100 ft, 1 (TCWC). PANAMA. Chi-
liqui: Cuesta de Piedra, 3000 ft, 1 (USNM);
i mi E Cuesta de Piedra, 2800 ft, 2 (USNM);
Guabala, 50 ft, 2 (USNM). Code: Peno-
nome, 45 (9 USNM, 36 UMMZ). Los Santos:
Cerro Hoya, 14 (USNM); Gudnico, 6
(USNM). Veragiias: Isla de Cebaco, 1
(USNM).

Pteronotus parnellii rubiginosus

(Wagner)

Chilonycteris rubiginosus Wagner, 1843, Wieg-
mann's Archiv fijr Natiirgesch., 9:367, 31
December (International Rules of Zoologi-
cal Nomenclature, Art. 21b, ii, p. 19).

Holotype. — Adult male, skin, skull
)iot removed, Zoologisches Staats-Sam-
mulung, Miinchen, West Germany no.
45; from Caicara, Mato Grosso, Brasil;
obtained by Johann Natterer on 19
August 1S2S; original no. 115.

Distrilnition. — Central America, from
central Honduras southward along the
humid Caribbean versant through Nica-
ragua, Costa Rica, and Panama into
South America (Amazonian lowlands to
as far south as the Mato Grosso region
of Brasil, Tingo Maria, Peru, and east to
Surinam); also on Trinidad. The range
of this subspecies is not well-defined in
eastern Panama or in northwestern and
eastern Colombia (Fig. 22).

Description. — Size, largest for the
species P. parnellii, both cranially and
externally. Standard external measure-
ments (average, with extremes in paren-
theses) for 12 males and four females,
respectively, from Rio Cunucunuma, Be-
len, 150 m, Amazonas, Venezuela, are
as follows: total length, 95.4 (88.0-
102.0), 91.2 (87.0-95.0); length of tail
vertebrae, 26.2 (23.0-28.0), 25.5 (24.0-
27.0); length of hind foot, 16.3 (15.0-
17.0), 16.2 (24.0-27.0); length of ear
from notch, 25.2 (24.0-28.0), 25.5 (25.0-
26.0). Length of forearm for the holo-
type (measured by Dilford C. Carter)
and a female from 264 km N (by road)
Xavantina, Mato Grosso, Brasil, are 64.2
and 61.8, respectively; see table 1, for
additional external and cranial measure-

ments of representative populations of
this subspecies. Members of this sub-
species are the darkest in color of all
subspecies of P. parnellii. Dorsally,
fresh pelage is extremely dark brown
(near Clove Brown). An August-taken
female in fresh pelage, from Santa Rosa,
17 km N and 15 km E Boaco, Nicaragua,
was Fuscous Black dorsally. The over-
all darker appearance of nihiginosus
can be attributed to the complete ob-
literation of the central zone of pale
coloration on the hairs of the back. This
zone, if at all discernible, is only slightly
paler in color than the terminal zones.
Ventrally, bats of this subspecies are
darker, owing to a more extensive Bister-
colored basal band on the hairs, which
have brownish-white tips. The increased
pigmentation of the hairs may account
for the brilliant cinnamon color of old
pelage, which usually is a rich Ochrace-
ous Orange in other geographic races.

Comparisons. — For comparison with
P. p. mesoamericanus, see account of
that subspecies. From P. p. fiiscus, the
subspecies rubiginosus may be distin-
guished by its larger external and crani-
al size ( Table 1 ) and darker pelage.

Remarks. — This subspecies occurs
from the northern part of the Amazon
Basin northward in the east to Trinidad
and Surinam, and northward in the
west to Central America, where, as was
mentioned in the previous account,
rubiginosus occurs in the Caribbean ver-
sant and seems to interbreed freely
along a north-south line (roughly co-
incident with the continental divide)
with P. p. mesoamericanus.

Koopman (1955:112) indicated that
P. p. fuscus probably was a synonym of
rubiginosus. However, after examining
a considerable number of specimens
from Middle and South America, par-
ticularly Venezuela (referable to rubig-
inosus), I conclude that the trinomen
fuscus is valid and should be retained
for the populations of P. parnellii from
along the Caribbean coast of northern
South America (excluding Trinidad).

76

MISCELLANEOUS PUBLICATION MUSEUM OF NATURAL HISTORY

The resemblance between ru])i^inosus
and fuscus increases ( lower distance co-
efficients, Fig. 21) in the region of
northeastern Venezuela. Physical and
ecological barriers posed by an eastern
extension of the Andes and the Vene-
zuelan llanos (an extensive east-west
grassland area that extends from south-
eastern Colombia, north of the Rio Ori-
noco, and eastward to the Delta Ama-
curo), probably separate fuscus and
rubiginosus in northwestern South
America. Limited gene flow between
these two subspecies probably occurs in
eastern Venezuela where the major
route for genetic exchange may be
through the more mesic Delta Amacuro.
The populations of rubiginosus on Trini-
dad with those of the Amazon Basin
are probably continuous through the
delta region of eastern Venezuela.

In figure 21, it will be noted that
localities 55 and 56 are linked by a low
distance coefficient and that localities
57, 58, 59, and to a lesser extent 54, like-
wise are connected by low coefficients.
However, these two groups of localities
are separated from each other by slightly
higher values. These differences are in-
teresting because localities 55 and 56
are situated near the southern edge of
the Venezuelan llanos where the tropical
forest is not well developed. Localities
57, 58, 59, and 54 are, however, situated
in areas of higher humidity and rainfall
where the tropical forest is more exten-
sive. In the material examined in this
study, specimens of rubiginosus were
found to be largest in southern Vene-
zuela and Guyana. Unfortunately, in-
sufficient material exists to fully charac-
terize the geographic variation in the
Brasilian portion of the Amazon Basin.

Specimens examined (.526). — HONDURAS.
Olancho: San Jose, Rio Tinto, 340 m,
3 (TCWC); N£ Catacamas, 500 m, 14
(TCWC). NICARAGUA. Boaco: Santa
Rosa, 17 km N, 15 km E Boaco, 300 m, 1
(KU); Los Cocos, 14 km S Boaco, 220 m, 1
(KU). Chontale.s: No specific locality, 1
(KU). Zelaya: Bonanza, 850 ft, 1 (KU);
Cara de Mono, 50 m, 4 (KU); El Rccreo, 1
(LACM); S side Rio Mico, El Recreo, 25 m.

1 (KU); Rio Mico, 6 (UMMZ); 10 km W
Rama, 40 m, 12 (TCWC); 2 mi ESE Rama,
200 ft, 7 (TCWC). COSTA RICO. Cartago:
3 km S Moravia, 1 (UMMZ). PANAMA.
Bocas- del Tow: Almirante, 27 (USNM); 7
km SSE Changuinola, 1 (USNM). Canal
Zone: Rio Pequeni, Salamanca Hydrographic
Station, 7 (MCZ); 3 mi from Cristobal, 2
(FMNH); Bas Obispo, 12 (USNM); Sardi-
nilla Mine no. 1, 3 mi E Camboa, 5 (MVZ);
Barro Colorado, 4 (UMMZ); Paraiso, 53
(USNM); Summit, 12 (USNM). Colon:
Btiena Vista, 12 (USNM). Daiien: El Real,

1 (USNM); Rio Chiicunaque, 2 (USNM).
Panama: Cerro Azul, 16 (USNM); Rio Chili-
brillo. Bat Caves South of Alhapiiila, 1
(USNM); Chilihri River, 7 (USNM). Son
Bias: Head of Rio Pito, Armila, 2 (USNM).
VENEZUELA. Apure: Hato Cariben (La
Villa), 60 km NE Pto. Paez, 76 m, 5 (USNM);
Cerro de Murcielago, 8 km NW Pto. Pdez, 76
m, 6 (USNM). Bolivar: Ciudad Bolivar, I
(MLS); Peru Mine, El Callao, 16 (1 FMNH,

2 USNM, 13 AMNH); Hato Florida, 14 km
S, 45 km E Caicara, 50 m, 17 (USNM);
Maripa, 1 (KU); Los Patos, 28 km SE El Man-
teco, 3 (USNM); Rio Siipamo, 56 km SE El
Manteco, 3 (USNM); Hato San Jose, 146 km
S, 7 km E Ciudad Bolivar, 306 m, 4 (USNM);
Boca de Parguaza al Orinoco, 3 (IZT); Isla
Oriipe, Rio Orinoco (Cerca Pto. Pdez), 1
(IZT); El Manaca, .59 km SE El Dorado, 6
(USNM); 10 km NE Icabaru, 1 (USNM).
Tcrritorio Amazonas: Chaparito, 9 km SE
Puerto Ayacucho, 119 m, I (USNM); Raija,
32 km SSE Puerto Aijacucho, 135 m, 5
(USNM); Rio Manapiare, San Juan, 155 m,
17 (USNM); Rio Cunucunuma, Belen, 150 m,
17 (USNM); Tamatama, Rio Orinoco, 135 m,

3 (USNM); Boca Mavaca, 68 km SE Esme-
ra'da, 138 m, 7 (USNM); Rio Mavaca, 108
km SE Esmeralda, 140 m, 22 (USNM).
GUYANA, [localities not mapped]: Kana-
slien, upper Essequiho Wai-Wai village, 1
( ROM ) ; Kuitaro River, Rupuntini Dist., I
(ROM); Nappi Creek, Kanuku Mts., I
(ROM). BRASIL. Rio Bronco, Rio Negro,
Sena de Cucuhy, 64 (AMNH); [localities not
mapped] Taparinha (on the Amazon River), I
(MPEG); 264 km N (by road) Xavantina (12°
51' S, 51° 46' W Serra do Roncador), approx.
1750 ft, Mato Grosso, 1 (USNM). PERU
[localities not mapped]. Tingo Maria, 1
(MCZ); 3 mi S Tingo Maria, 2000 ft, Hu4-
nuco, .38 (TCWC). TRINIDAD. Port of
Spain, 2 (USNM); Zacaya Cave, Montevideo,
11 (USNM); Gudcharo Cave, 4 (MCZ);
Oroi>ouche Heights, 8 (FMNH); Guanapo
Heights, 17 (1 AMNH, 4 LACM, 12 USNM);
Las Cueva^, 1 (AMNH); Mt. Tamana Caves,
IS (2 TCWC, 4 ROM, 12 USNM); Tainana
Hill, I (AMNH).

Additional marginal record. — SURINAM.

SMITH: SYSTEMATICS OF MORMOOPID BATS

77

Tafelljerg, Anton van Aerde Cave ( Husson
1962:74).

Pteronotus parnellii fuscus (J. A. Allen)

Chiloiujcteris- mhiginosa fusca ]. A. Allen,
1911, Bnll. Amer. Mus. Nat. Hist., 30:262^
2 December.

Holotijpe.—kdnXi male, skin and
skull, AMNH 31561; from Las Qniguas,
5 mi S Puerto Cabello, 650 ft, Venezue-
la; obtained by M. A. Carriker on 7
September 1910.

Distribution.— ^oxihexn South Amer-
ica, along the semiarid Caribbean coast
of Colombia and Venezuela (Fig. 22).

Description. — Size, medium for spe-
cies and similar to that of P. p. meso-
americanus. Standard external measure-
ments ( average, with extremes in paren-
theses) for seven males and seven fe-
males, respectively, from the vicinity of
Urama, Yaracuy, Venezuela, are as fol-
lows: total length, 95.3 (93.0-98.9), 87.5
(83.0-94.0); length of tail vertebrae,
24.6 (20.0-27.0), 24.1 (22.0-26.0); length
of hind foot, 14.9 (14.0-16.0), 14.7 (13.0-
16.0); length of ear from notch, 22.1
(18.0-25.0), 23.1 (23.0-24.0). Additional
selected external and cranial measure-
ments are given in table 1.

Fresh pelage of fuscus, as the name
implies, is dark grayish-brown dorsally.
Two February-taken females from 1.5
km S Cariaco, Sucre, Venezuela, are
nearly Clove Brown dorsally. Individual
hairs from the back of bats from this
region are blackish-brown basally and
tipped with a zone of Prout's Brown;
the central zone is almost indistinguish-
able from the base, being only slightly
paler in color. Ventrally, the pectoral
and abdominal regions are dark brown,
owing primarily to the dark brown basal
bands on the hairs. Each hair has a
whitish or buffy tip. Bats of this sub-
species on Margarita Island, Nueva Es-
parta, Venezuela, are somewhat paler
(Sayal Brown) in color than adjacent
mainland populations.

Comparisons. — For comparison with
P. p. rubiginosus, see account of that
subspecies.

Remarks. — The subspecific name fus-
cus has been variously applied to the
Middle American individuals of P. par-
nellii herein designated mesoamericanus
( Hershkovitz, 1951; Koopman, 1955;
Felten, 1956; Hall and Kelson, 1959;
Jones, 1966, Jones, et al, 1971; and
others ) .

The variation within this subspecies
is similar to that discussed in each of
the above accounts, that is, geographic
departure from the typical population
results in a certain degree of phenotypic
deviation. In this case, the trend is gen-
erally toward larger size either to the
westward or eastward from northwest-
ern Venezuela. To the west, fuscus may
intergrade with rubiginosus and perhaps
mesoamericanus in eastern Panama and
northwestern Colombia as discussed
previously. To the east, due to an ap-
parent breakdown of the mountain and
grassland barriers that apparently par-
tially isolate coastal populations of
fuscus, there seems to be limited gene
exchange with rubiginosus.

Of interest in northeastern Venezue-
la is a notable reduction in overall size
of specimens of fuscus from Margarita
Island (Table 1, locality 52). Bats from
there were netted on the northeastern
end of the island where the habitat is
somewhat mesic and on the western end
where conditions are extremely xeric.
The point to be made here is that, again,
in an area where less than optimal habi-
tat prevails, P. parnellii deviate pheno-
typically from the typical population.

The close resemblance between fus-
cus and the Middle American subspecies
herein designated mesoamericanus is
one of the most interesting problems en-
countered during this study. Previously
bats of these two subspecies were as-
signed to fuscus by various authors.
This assignment evidently resulted from
a paucity of material formerly available
from the region of Guatemala south to
northern South America. Climatic and
ecological fluctuations in eastern Pana-
ma and northwestern Colombia during

78

MISCELLANEOUS PUBLICATION MUSEUM OF NATURAL HISTORY

the Pleistocene seem to have been key
factors in the evokition of these two
races; the continued maintenance of this
resemblance may relate to similarity in
the habitats occupied by each.

Typically, fuscus averages slightly
larger than mesoamericanus (Fig. 19),
although there is broad overlap. The
differences are such that if the two
populations were contiguous I would
not have proposed recognition for meso-
americanus. However, in view of the
fact that both fuscus and mesoameri-
canus enter into a zone of intergradation
with rubiginosus, in eastern Panama and
adjacent regions of Colombia, and di-
verge from this zone as distinct units,
indicates a certain degree of indepen-
dence, possibly including independent
adaptation to a similar environment. I
therefore propose that they be recog-
nized as separate subspecies to better
represent the observed geographic vari-
ation.

Specimens examined (221). — COLOMBIA.
Bolivar: Cartagena, 1 ( ROM ) ; Coloso, Las
Campanas, 250 ni, 16 (FMNH); Norosi,
Mompas, 1 (USNM); Magdelena: Colonia
Agricola de Caracalicito, Santa Marta, 400 m,

4 (USNM). VENEZUELA. Aragua: Ran-
cho Grande, 1100 m, 24 (1 UMMZ, 4 MLS,
6 USNM, 13 IZT). Carahobo: Rio Yaracuy,
10 km NW Urama, 25 m, 4 (USNM); 6 km
NW Urama, 60 m, 3 (USNM); San Esteban,
12 (2 USNM, 10 AMNH); Las Quiguas, 5 mi

5 Puerto Cahello, 650 ft, 6 (1 FMNH, 5
AMNH). Distrito Federal: El Limon, 3 km
S, 46 km W Caracas, 398 m, 2 (USNM);
Trapiche (Cerca de El Limon), 1 (IZT).
Gudrico: 57 km S, 49 km E Caracas, 630 m,
3 (USNM); Cueva de Los Murcielagos, Loma
de Medio, 460 m, 1 (MLS); Cueva Loma del
Medio, 7 km S San Juan de Los Morros (Car-
retera a stir de Sebastian), 3 (IZT). Miranda:
Los Teques, 1700 m, 1 (IZT); Curapao, 19
km E Caracas, 1160 m, 5 (USNM); Cueva
Walter Dupuij-Capaya, 1 (MLS); Cueva Al-
fredo jahn, 70 m, 20 (9 IZT, 11 MLS). Mona-
gas: 8 km NW (by road) Caripito, 1 (KU);
San Agustin, 2 km N, 3 km W Caripe, 1165
m, 1 (USNM); Caripe, 900 m, 20 (IZT);
Cueva del Gudcharo, 1050 m, 1 (MLS).
Nueva Esparta: Cerro Mata Siete, 2 km N, 2
km E La A.suncion, 305 m, 1 (USNM); La
Agtiada, 3 km S La Asuncion, 53 m, 12
(USNM); Cueva Atagua, Cerro Atagua, 5
( MLS ); Cueva Honda del Piache, SE El Valle,

2 (RMNH); El Valle, 50 m, 2 (KU); Cueva
El Convento, 1 km N, 1 km W San Francisco
de Macanao, 100 m, 9 (KU). Sucre: Caserio
Santa Fe, 1 (MLS); Tacal, 11 km SSW Cti-
mand, 2 (KU); Rio Grande, Cerca de Juan
Antonio, 10 m, 1 (MLS); 1.5 km S Cariaco, 3
(KU); 1.5 km NW El Pilar, 1 (KU); Rio
Salado, 9 km N, 4 km E Giiiria, 90 m, 9
(USNM); Ensenada Cauranta, 7 km N, 5 km
E Giiiria, 1 m, 2 (USNM). Trujillo: Agna
Santa, 23 km NW Valera, 90 m, 2 (USNM).
Yaracuij: Minas de Aroa, Bolivar R. R., 11 (1
FMNH, 4 MLS, 6 AMNH); 19 km NW
Urama, 7 (USNM).

Subgenus Chilonycteris Gray

Type species. — Chilonycteris maclea-
yii Gray.

Distribution. — From southern Sonora
and Tamaulipas, Mexico, southward
(excluding portions of the Mexican
Plateau and most of the Yucatan Penin-
sula) into South America, including
Trinidad, to Mato Grosso, Brasil. The
subgenus is represented on the Greater
Antilles by two endemic species — Ptero-
notus fuliginosus and P. macleayii. For
additional information concerning dis-
tribution, see accounts beyond and fig-
ures 24, 27, and 28.

Description. — Size small, forearm
usually less than 50; wing membranes
attached high on sides of body, leaving
a narrow, hair-covered band on back;
labio-nasal plate simple to moderately
complex, margin above each nostril
smooth with irregular, rounded, wart-
like tubercles (three to four tubercles
over each nostril) or with an elongate,
squared excrescence, lateral spikes ab-
sent in P. personatus but present in both
fuliginosus and nmcleayii (Fig. IIB, and
D-E); tragus relatively simple, spatu-
late, with a moderately well-developed
and shelf-like secondary fold or with
prominent secondary fold covered with
many long and short hairs (Fig. 10, B-
D); antero-medial edge of pinna of ear
with two to four, or occasionally as
many as six, tooth-like serrations; ears
connected by two low, inconspicuous
ridges that meet on top of the muzzle
and form a prominent rostral tubercle

SMITH: SYSTEMATICS OF MORMOOPID BATS

%

in persomitus and a low rostral tubercle
in fuliginosus and madeayii; pelage dis-
tinctly tricolored dorsally (more promi-
nent in fuliginosus and macleayii than
in personatus); profile of skull relatively
flat, although rostrum may be slightly
upturned; forehead slopes gradually in
personatus, but rather abruptly in the
Antillean species; braincase ovoid in
shape in personatus and globular in fu-
liginosus and macleayii; rostral breadth
almost always less than, or equal to, the
length of the maxillary toothrow; upper
incisors reduced in size, inner pair bi-
furcate and with a narrow rounded heel
directed laterally toward canine; inner
and outer incisors of each premaxillary
compressed together, forming two
groups with just the inside tips of the
inner two incisors touching (this char-
acter shared with the subgenus Ptero-
notus); diastema usually present be-
tween outer incisors and canine (more
prominent in Antillean species); lower
incisors reduced in size, all trilobed and
with short, rounded heel; tympanic ring
large, almost completely covering audi-
tory bulla; basioccipital region broad
between bullae, with two large ex-
tremely shallow, oval-shaped pits, into
which the M. longus capitus inserts.

Remarks. — This subgenus and the
subgenus Pteronotus are similar and
probably more closely related than
either is to the subgenus Phylloclia. Ex-
ternally Chilonycteris is easily distin-
guished from Pteronotus by the absence
of a naked back, structure of the tragus,
and structure of the labio-nasal plate.
Cranially, the two subgenera are simi-
lar. They differ, however, in shape and
proportions of the braincase and rostral
regions; Pteronotus being broader in
both respects (Figs. 25 and 34).

Pteronotus macleayii (Gray)
MacLeay's Mustached Bat

( Synonymy under subspecies )

Distribution. — Cuba and Jamaica

(Fig. 24).

De.scription. — Size small, forearm
usually measuring 40 to 45; wing mem-
brane and uropatagium attached to
ankle by way of a long ligament that is
tightly bound to the distal half of tibia;
tragus long and spatulate, secondary
fold prominent, sparsely clothed by
short bristle-like hairs, with longer hairs
on cranial and distal edges (Fig. lOD);
labio-nasal plate distinct, a squared
lappet above each nostril and prominent
lateral spikes to either side of nostrils
(Fig. IID); one or two (occasionally
three) tooth-like serrations on antero-
medial edge of pinna; ears connected
by an extremely low, inconspicuous
ridge that is fused on top of the nose,
rostral tubercle not prominent; clitoris
unusually long; pelage tricolored dor-
sally; profile of skull relatively flat, ros-
trum slightly elevated, forehead rises
abruptly onto round, high braincase;
rostrum longer than braincase (Fig.
25); marked diastema between outer
upper incisor and canine; inner incisors
distinctly bifurcate; lower incisors re-
duced in size and trilobed with short,
rounded heels.

Relationships. — This species is most
closely related to P. fuliginosus, which
also inhabits the Greater Antillean re-
gion. Both are small species, but may
be readily separated by length of fore-
arm as well as condylobasal length,
macleayii being the larger in both re-
spects. In addition, nuicleayii has a
longer and broader rostrum than fu-
liginosus (Fig. 25). Both Antillean spe-
cies seem to be closely allied to the
mainland P. personatus, although differ
from it in construction of the tragus and
labio-nasal plate, and in shape of the
braincase. Along with Mormoops hlain-
villii, P. fuliginosus and macleayii may
represent a relatively early invasion of
the Greater Antilles by mormoopids as
discussed in a later section.

Geographic variation. — The popula-
tions of P. macleayii from Guba and Ja-
maica are currently allocated to the sub-
species macleayii and grisea, respective-

80

MISCELLANEOUS PUBLICATION MUSEUM OF NATURAL HISTORY

ly. Analysis of geographic variation of
this species verifies the distinctness of
these two populations in that significant
differences were computed between
them for every character examined, ex-
cept breadth across the post-palatal
extension. Cuban representatives of the
species are consistently smaller in over-
all size than those from Jamaica.

Remarks. — There has been some con-
fusion in the literature with respect to
the identity and distinctness of the
species inacleayii and fiiUfiinosiis, evi-
dently stemming from mislabeling of
the holotypes of the two species. I am
grateful to J. Knox Jones, Jr., Karl F.
Koopman, and especially John Edwards
Hill, who examined the holotypes and
contributed to resolution of this con-
fusing problem.

Gray (1839:4) described ChiJomjc-
teris macleaijii based on material from
Cuba presented to him by W. S. Mac-
Leay. According to Hill (personal com-
munication), records at the British Mu-
seum (a manuscript catalogue prepared
by Gray) showed that MacLeay gave
Gray three specimens of Chilonycteris
from Cuba. These appear in the British
Museum register under the number BM
38.6.21.S. Gray's (1839:4) description
was extensive and delimits accurately
the species herein designated P. mac-
leaijii. Salient points in his description
were "edge of the nose with two tooth-
like expansions on each side, one over
each nostril," and that the length of the
forearm was 1.5 inches (approximately
38 mm). The labio-nasal plate illus-
trated by Gray (1839: pi. 1, fig. 2)
agrees exactly with that illustrated in
figure IID. Apparently, the skull was
not removed at the time of Gray's de-
scription.

Gray's (1843a: 20) description of
Chilonycteris fuliginosa is not nearly so
kjng nor so detailed as that for niac-
leayii. Gray noted that fuli<finosa agreed
in many respects with Chilonycteris
macleayii, but differed from the three
specimens of that species in the collec-

tion of the British Museum in being
darker in color and having larger and
rather narrower ears. No mention was
made regarding the labio-nasal plate.
The holotype of this species bears the
register number BM 42.11.24?, and the
skull has not been removed from this
specimen.

In September of 1968, J. Knox Jones,
Jr. compared several specimens of sup-
posed macleayii and fuliginosus with
mormoopid holotypes housed in the
British Museum. His conclusions were
that the labeled holotypes of both mac-
leayii (skull removed) and fuliginosa
belonged to one and the same species,
and agreed completely with a supposed
individual of fuliginosus from Cuba. In
addition, the labio-nasal region of both
holotypes agreed with that shown in
figure HE. On the other hand, Chilo-
nycteris grisea, as described by Gosse
(1851), agreed with a supposed speci-
men of inacleayii and figure IID. These
results were indeed perplexing, since
Gray (1839) and later Dobson (1878)
clearly illustrated the labio-nasal plate
that agreed precisely with material cur-
rently assigned to macleayii. I asked
Karl F. Koopman, who visited the Brit-
ish Museum shortly afterward, to re-
examine these particular holotypes using
his knowledge of the group. His con-
clusions verified those of Jones. This
led me to re-examine all of the descrip-
tions and pertinent literature concerning
these holotypes; later I corresponded
with J. E. Hill, asking if he could shed
some light on this subject. Hill ap-
proached this matter by trying to trace
the three specimens originally given to
Gray. He commented (personal com-
munication), as follows:

"One, a male, is the reputed type
specimen. It is also in a bottle labeled
comparatively recently by G. W. C. Holt
and lacks any identifying mark except
a loose label in the bottle, written by
R. W. Hayman and bearing only the
registration number BM 38.6.21.8. Its
identity with the type of fuliginosa is

SMITH: SYSTEMATICS OF MORMOOPID BATS

81

cy-":

30

- 20

80

70

Figure 24. Distribution of Pteronotus macleayii.

indicated by a remark on the bottle
label of the latter by Holt which reads
'= C. nuicleayi.'

"The other male example is in a bot-
tle first labeled by Gray and subse-
quently by Dobson. It is evidently the
original container: inexplicably, there
is also a specimen of Pteronotus daviji
in the bottle. Dobson has labeled the
bottle 'Type' and ' 5 and S ' but the
word type and the symbols have been
obliterated at some time. However, the
bottle also contains a loose label written
by Dobson, which reads 'Chilonycteris
macleayi type.' This male specimen is
clearly referable to the species which
you characterize as macleayii.

"To some extent the female speci-
men eludes me. However, the collec-
tions include a female Chilonycteris,
BM 7.1.1.700, from the Tomes Collec-
tion, its bottle labeled by Gray as Chilo-
nycteris macleayi, from Macleay. An
addition to the register entry for this
specimen reads 'No doubt typical: Ap-

parently from Brit. Mus.' It seems to me
that this may be the female listed by
Gray and Dobson and that it found its
way into the collection of R. F. Tomes,
returning here when that collection was
accessed in 1907. The specimen agrees
with the type of fuliginosa."

At the beginning of World War II,
type specimens were segregated from
the general collections at the British
Museum. Hill is of the opinion that
probably the two males described above
were in the same bottle and that the
wrong specimen later was segregated as
the holotype. This error may then have
been compounded by subsequent label-
ing. Hill also wrote that the specimens
may have been incorrectly segregated
by either Thomas or Dobson.

The solution to this problem sug-
gested by Hill, and one that is accept-
able to me, is to take the second male as
the holotype of Chilonycteris macleayii.
Gray apparently had two kinds of mor-
moopid bats in the MacLeay accession

82

MISCELLANEOUS PUBLICATION MUSEUM OF NATURAL HISTORY

— a large species and a smaller one.
Gray's (1839) description and illustra-
tion of Chilonycteris macleaijii clearly
refer to the larger bat here designated
Pteronotus macleaijH.

Pteronotus macleayii macleayii (Gray)

Chilonycteris macleayii Gray, 1839, Ann. Nat.
Hist., 4:5, September.

Holotype. — Adult male, preserved in
alcohol, skull apparently not removed,
BiM 38.6.21.8; from Cuba; collected by
W. S. MacLeay. See remarks in species
account of P. macleayii above.

Distribution. — Cuba (Fig. 24).

Description. — Smaller of the two
subspecies of the species, both cranially
and externally. Most specimens exam-
ined were preserved in alcohol, thereby
making difficult an accurate assessment
of color. A December-taken male, from
Soledad, Las Villas, Cuba, has relatively
fresh pelage that is tricolored dorsally.
Individual hairs from the back have
olive-brown basal bands and dark brown
to blackish brown tips. The central
band is whitish to grayish white. The
head does not differ appreciably in color
from the back. Ventrally, the pelage is
grayish white with a pale buffy tinge.
The hairs from the pectoral and ab-
dominal regions are sharply bicolored,
being Clove Brown at the base and
tipped with grayish white; on the throat
and chin, hairs are more nearly mono-
colored or have pale Snuff Brown bases.
Of three individuals, from San Vicente,
Pinar del Rio, captured on 26 July, one,
a male, resembles the specimen de-
scribed above, being only slightly
darker. The other two specimens are
females that are Ochraceous Orange in
color.

Measure ments . — Selected external
and cranial measurements (average,
with extremes in parentheses, plus or
minus two standard errors of the mean)
of 15 adult males from Cuba, are as
follows: length of forearm (12 speci-
mens only), 42.1 (41.2-43.4) ± 0.37;
zygomatic breadth, 7.8 (7.6-8.0) d=

0.07; rostral breadth, 6.1 (5.9-6.5) ±
0.09; length of maxillaiy toothrow, 6.6
(6.4-6.8) ± 0.05; condylobasal length,
14.5 (14.1-15.0) ± 0.15; depth of crani-
um, 6.9 (6.4-7.1) ± 0.09.

Standard external measurements
( average, with extremes in parentheses )
of four males from Baracoa, Oriente,
Cuba, are as follows: total length, 62.7
(61.0-65.0); length of tail vertebrae, 21.7
(20.0-23.0); length of hind foot, 13.7
(12.0-15.0); length of ear from notch,
9 for each individual.

Comparisons. — From P. m. grisea of
Jamaica, P. m. macleayii may be dis-
tinguished on the basis of smaller exter-
nal and cranial size ( see measurements ) .
In addition, the cranium of macleayii is
noticeably narrower and shorter than
that of grisea ( Fig. 25 ) . These two sub-
species do not appear to differ in color,
although additional material will be
necessary before this can be precisely
judged.

Remarks. — In the past, some authors
have placed the name Lohostoma quad-
ridens Gundlach, 1840, in synonymy
under P. m. macleayii. Close examina-
tion of Gundlach's description indicates
that specimens of both the species
macleayii and fuliginosus probably were
present in his sample; no holotype or
type series was designated. I therefore,
regard Gundlach's quadridens as a
nomen diibium.

Specimejis examined (82). — CUBA. Ha-
hana: Cueva de Mudo, Giiines, 6 (ROM);
Bucuranao, I (AMNH). Oriente: Cueva
Zuntiaga, Barajagua-Mayari, 2 (MCZ); Bara-
coa, 44 (2 FMNH, 10 MCZ, 32 USNM);
Cueva de Majana, Baracoa, 11 (4 AMNH, 7
MCZ); Santiago, 1 (USNM). Pinar del Rio:
Cueva del Rio, San Vicente, 1 (TCWC); San
Vicente, 12 (11 AMNH, 1 LACM); Banos
San Vicente, 2 (USNM). Las Villas: Soledad,
Hai-vard Botanical Gardens, 1 (ASC). Isla de
Finos: Cueva de Pedernales, south coast, 1
(AMNH).

Pteronotus macleayii grisea (Gosse)

Chilonycteris grisea Gosse, 1851, A naturalist's
sojourn in Jamaica. Mem. Council Royal
Soc. Agric, Jamaica, p. 326.

SMITH: SYSTEMATICS OF MORMOOPID BATS

83

A

B

C

D

F

Figure 25. Dorsal \iew oi skulls oi Pteronotus fulifiinosiis torrei (A, Cuba), P. f. fuliginosus (B,

Jamaica; C, Dominican Republic; D, Puerto Rico), P. m. itiacleayii (E, Cuba), and P. m. grisea (F,

Jamaica). All specimens are adult males; horizontal line above B is 10 mm in length.

Holotype. — Adult male, preserved in
alcohol, .skull not removed, BM (no
registration number); from Phoenix
Park, near Savanna la Mar, Westmore-
land Parish, Jamaica.

Distribution. — Jamaica (Fig. 24).

Description. — Size large for the spe-
cies. Eight January-taken specimens
from Oxford Cave, Balaclava, Jamaica,
are in relatively fresh pelage which is
dark brown dorsally. The basal and
terminal color bands of hairs on the back
are dark brown ( Bister Brown to Prout's
Brown); the central color band is gray-
ish white. Ventrally, the pectoral and
abdominal regions are grayish white
tinged with buff, the throat and chin
being somewhat paler.

Measurements. — Selected external
and cranial measurements (average,
with extremes in parentheses, plus or
minus two standard errors of the mean)
of 19 adult males (sample size in
brackets if different) from Jamaica are
as follows: length of forearm, 43.5
(41.7-45.1) ±: 0.33 [33]; zygomatic
breadth, 8.2 (8.0-8.4) ± 0.05; rostral
breadth, 6.4 (6.3-6.6) ± 0.04; length of
maxillaiy toothrow, 6.8 (6.7-7.0) ±
0.03; condylobasal length, 15.2 (14.8-
15.7) ± 0.10 [17]; depth of cranium,
7.1 (6.9-7.3) ± 0.06.

Standard external measurements
( average, with extremes in parentheses )
of six males and six females from Provi-
dence Cave, Montego Bay, Jamaica, are

as follows: total length, 74.2 (73.0-
75.0), 74.3 (72.0-78.0); length of tail
vertebrae, 23.7 (22.0-25.0), 23.3 (22.0-
25.0); length of hind foot, 9.4 (9.0-10.0),
9.9 (9.5-10.0).

Comparisons. — For comparison with
P. m. macleaijii, see account of that sub-
species.

Remarks. — The degree of difference
between the Jamaican subspecies, gri-
sea, and macleaijii from Cuba seems to
indicate the relative effectiveness of the
water barrier between the two islands.
In view of their small size, which may
greatly affect their ability to cross the
strait between the two islands, it is
doubtful that any substantial gene flow
now exists between these two races.

Specimens examined (69). — JAMAICA.
Hanover: Lucea, 4 (AMNH). St. Catherine
St. Clair Cave, 12 (3 LACM, 9 USNM)
Spanish Town, 1 (USNM). S^ Elizabeth
Oxford Cave, Balaclava, 8 (AMNH). St.
James: Providence Cave, Montego Bay, 22
(AMNH); Montego Bay, 22 (USNM).

Pteronotus fuliginosus (Gray)
Sooty Mustached Bat

( Synonymy under subspecies )

Distribution. — Cuba, Jamaica, His-
paniola, and Puerto Rico (Fig. 27).

Description. — Size small, forearm
usually less than 40.0; wing membrane
and uropatagium attached to ankle by
way of a long ligament that is tightly
bound to the distal half of the tibia;

84

MISCELLANEOUS PUBLICATION MUSEUM OF NATURAL HISTORY

36.0 37,0 38.0 39.0 40.0 4 1.0
I r 1 1 — ~T — I r — I 1 i 1

A

7.2

.4 7.6 7.8 8.0

"1 r

1 I r

"1 1

C

5,6 5.

6.0 6.2 6.4

"1 I r

B

5.6 5.8 6.0 6.2 6.4
I 1 I I 1 I I 1 1 —

D

Figure 26. Dice-squares illustrating geographic variation of Pteronotus fuliginosus from the Greater
Antillean islands of Cuba, Jamaica, Hispaniola, and Puerto Rico ( top to jjottom ) . The measurements
illustrated (in mm) are: A, length of forearm; B, rostral breadth; C, zygomatic breadth; D, length
of maxillary toothrow. The thin horizontal hne represents the range of variation and the thin vertical
line the mean; the blackened part of the horizontal bar equals two standard errors on either side of
the mean and the white part indicates one standard deviation on either side of the mean. See text

for discussion.

tragus long and spatulate, secondary
fold prominent, sparsely covered with
short bristle-like hairs, with longer hairs
on cranial and distal edges (Fig. IOC);
labio-nasal plate moderately complex,
margin above each nostril with three to
four wart-like tubercles, nostrils sep-
arated by slight emargination, promi-
nent lateral spike on either side of
nostrils (Fig. HE); three to four tooth-
like serrations on antero-medial edge of
long, lanceolate portion of the ear; ears
connected by two extremely low, incon-
spicuous ridges that fuse on top of nose,
rostral tubercle not prominent; pelage
distinctly tricolored dorsally, hairs of
back having three well-marked bands,
pelage bicolored ventrally; profile of
skull relatively flat, rostrum slightly ele-
vated, forehead rises abruptly onto

round, high braincase; rostrum approxi-
mately same length as braincase (not
noticeably longer — Fig. 25); marked
diastema between outer upper incisor
and canine; inner incisors distinctly bi-
furcate; lower incisors reduced in size
and trilobed with short, rounded heels.

Relationships. — See discussion of re-
lationships in foregoing account of
Pteronotus rnacleayii.

Geographic variation. — As currently
understood, the representatives of this
species from Jamaica and Hispaniola
are assigned to the subspecies fuligino-
sus and those from Puerto Rico to
inftata; bats from Cuba have been re-
ferred to a distinct species, torrei. My
analysis of geographic variation, indi-
cates that the populations of P. fuligi-
nosus from Jamaica, Hispaniola, and

SMITH: SYSTEMATICS OF MORMOOPID BATS

85

- 30

G^-

Pf. fuliginosus
\

20

80

70

Figure 27. Distribution of Pteronotus fuliginosus.

Puerto Rico are more or less homoge-
neous. On the other hand, these popu-
lations are separated by non-overlapping
subsets of means from Cuban bats. The
only characteristic in which non-over-
lapping subsets of means was not ob-
served was the length of the forearm.
In external and cranial size, the Cuban
bats are smaller than bats from the other
Greater Antillean islands (Figs. 25-26).
Generally, the largest representatives of
this species are from Hispaniola and
Puerto Rico. The bats of the Jamaican
population are intermediate in size be-
tween those of Cuba and Hispaniola
and Puerto Rico. Therefore, it seems
that only two populations merit recog-
nition at the subspecific level.

Pteronotus fuliginosus fuliginosus (Gray)

Cliilontjcteris ftiliginosa Gray, 1843, Proc. Zool.
Soc. London, p. 20, 28 February.

Cliilomjcteris macleayii inflata Rehn, 1904,
Proc. Acad. Nat. Sci. Philadelphia, 56:190,
26 March. Type locality, Cueva di Fan,
near Pueblo Viejo, Puerto Rico. Holotype,

adult male, preserved in alcohol, skull re-
moved, ANSP 6234; obtained on 19 March
1900, by Dr. C. W. Richmond and Dr. L.

Steneger.

Holotype. — Adult male, preserved in
alcohol, skull not removed, BM 42.11.
24? (registration number currently in
doubt); from Port au Prince, Haiti; pre-
sented by J. N. Tweedy.

Distribution. — Greater Antillean is-
lands of Jamaica, Hispaniola, and Puerto
Rico (Fig. 27).

Description. — Size large for the spe-
cies. The pelage of these bats is dis-
tinctly tricolored dorsally. However, the
majority of specimens of this species
that I examined were preserved in alco-
hol, making accurate color descriptions
difficult. A series of males and females
from Montego Bay, captured on 31 De-
cember, were in relatively fresh pelage.
Dorsally, specimens were grayish brown.
Hairs of the back have three distinct
color bands, the broadest of these being
a central band that is grayish white to
buffy white; the basal and terminal

86

MISCELLANEOUS PUBLICATION MUSEUM OF NATURAL HISTORY

bands are dark brown or Fuscous Black.
Ventrally, the overall color is grayish
white in the pectoral and abdominal re-
gions; the individual hairs are bicolored
\vith Clove Brown basal bands and
terminal bands that are whitish or gray-
ish white. The throat and chin are
slightly paler than other parts of the
venter owing to a reduction of the dark
basal portion of the hairs. An adult
male from the Dominican Republic, cap-
tured on 17 October, is in fresh pelage
that is extremely dark brown, almost
black in color.

Measu re me nts . — Selected external
and cranial measurements (average
with extremes in parentheses, plus or
minus two standard errors of the mean)
of eight adult males from Jamaica, nine
from Hispaniola, and 18 from Puerto
Rico (sample size in brackets if differ-
ent) respectively, are as follows: length
of forearm, 3S.0 (37.0-39.5) + 0.34 [16],
39.8 (38.5-40.8) ± 0.83 [5], 38.3 (30.1-
39.9) ±: 0.59 [31]; zygomatic breadth,
7.6 (7.3-7.8) ± 0.11, 7.6 (7.4-7.9) ±
0.12, 7.6 (7.3-7.9) ± 0.08; rostral
breadth, 6.1 (5.9-6.2) ± O.OS, 6.1 (6.0-
6.3) ± 0.08, 6.1 (5.9-6.4) ± 0.07; length
of maxillary toothrow, 6.0 (5.9-6.2) ±
0.07, 6.0 (5.9-6.2) ± 0.07, 6.0 (5.8-6.1)
± 0.05 [19]; condylobasal length, 13.5
(13.2-13.8) ± 0.15 [7], 13.6 (13.3-14.1)
± 0.15, 13.6 (13.2-13.9) ± 0.10 [17];
depth of cranium, 6.8 (6.7-7.1) ± 0.10,
7.1 (6.7-7.3) ±: 0.13, 7.1 (6.9-7.7) ±
0.09 [17].

Standard external measurements
(average, with extremes in parentheses)
of 13 males and 13 females, respectively,
from St. Clair Cave, St. Catherine Par-
ish, Jamaica, are as follows: total
length, 60.2 (58.0-64.0), 61.1 (59.0-
66.0); length of tail vertebrae, 18.8
(16.0-21.0), 19.3 (18.0-22.0); length of
hind foot, 8.8 (8.0-10.0), 9.0 (8.0-10.0);
length of ear from notch, 15.3 ( 15.0-
16.0), 15.2 (14.0-16.0).

Comparisons. — From P. /. torrei of
Cuba, P. /. fuliginosus is readily distin-
guished on the basis of larger external

and cranial size (Figs. 25-26), although
there may be slight overlap in the range
of any one measurement.

Remarks. — Rehn (1904) distinguished
"Chilomjcteris macleaijii in^ata" from
fuliginosus primarily on the basis of a
more inflated rostrum in the former.
The breadth of the rostrum of the holo-
type (ANSP 6234) is 6.4, which falls at
the upper end of variation observed in
populations of P. fuliginosus. However,
in view of the other similarities between
populations of P. fuliginosus from Ja-
maica, Hispaniola, and Puerto Rico, I
regard the subspecies inflata as a junior
synonym of P. f. fuliginosus.

Specimens examined (265). — JAMAICA.
St. Catherine: St. Clair Cave, 29 (27 ROM, 2
USNM). St. Elizabeth: Oxford Cave, Bala-
clava, 8 (AMNH); Balaclava, 28 (26 AMNH,
2 MCZ). HAITI. San Michel, 2 (USNM); ^
mi SE Cerca La Source, 52 (USNM); no
specific locality, 1 (AMNH). DOMINICAN
REPUBLIC, bistrito Nacional: Caiia Honda,
Santo Domingo, 1 (AMNH). Pedernales: 6
mi NE Oviedo, 1 (ASC). Puerto Plata: Sosua,
2 (1 ASC, 1 MCZ). Santiap,o Rodriguez:
Cueva Clavijo, 12 km W Moncion, 2 ( 1
AMNH, 1 FMNH). No specific locality, 4
(ANSP). PUERTO RICO. Cueva di Fari,
near Pueblo Viejo, 46 ( 12 AMNH, 2 ANSP, 9
UMMZ, 23 USNM); 7.5 km E Guaniana, 1
(ASC); Cueva de Burro, between Vega Aha
and Vega Baja, 25 (10 UMMZ, 15 USNM);
Cueva de Trujillo Alto, 59 (AMNH); San
Juan, 1 (USNM); Mayaguez, 2 (USNM);
Toro Negro Doiia Juan Recreation Area, 1
(ROM).

Pteronotus fuliginosus torrei

(G. M. Allen)

Chdontjcteris torrei G. M. Allen, 1916, Proc.
New England Zool. Club, 6:4, 8 February.

Holotype. — Adult female, preserved
in alcohol, skull removed, MCZ 11672;
from La Cueva de La Majana, Baracoa,
Oriente Province, Cuba; obtained by Sr.
Victor Jose Rodriguez y Verrier, on 15
June 1915.

Distribution. — Cuba (Fig. 27).

Description. — Size small, both ex-
ternally and cranially, for species (see
measurements). In addition to small
size, torrei has a narrow rostrum that is

SMITH: SYSTEMATICS OF MORMOOPID BATS

87

noticeably attenuate at the tip; the
nasal aperture also is narrow (Fig. 25).
The pelage is short and dark brown in
color dorsally. The central color band
of hairs from the back appears to be
somewhat more restricted and more
brownish in December-taken specimens
of this race compared with individuals
of P. f. fuliginosus from Jamaica taken
at the same time of year. Ventrally, the
pelage is pale Snuff Brown or bulfy
brown.

Measurements. — Selected external
and cranial measurements (average
with extremes in parentheses, plus or
minus two standard errors of the mean)
of 15 adult males from Cuba (sample
size in brackets, if different) are as fol-
lows: length of forearm, 37.3 (35.9-
38.5) ± 0.26 [30]; zygomatic breadth,
7.4 (7.1-7.7) ± 0.08; rostral breadth, 5.9
(5.8-6.1) ± 0.05; length of maxillary
toothrow, 5.8 (5.6-6.0) ± 0.05; condylo-
basal length, 13.2 (12.8-13.4) ± 0.09;
depth of cranium, 6.7 (6.4-7.2) ± 0.11.
No reliable standard external measure-
ments are available.

Comparisons. — For comparison with
P. f. fuliginosus, see account of that sub-
species.

Remarks. — The subspecies torrei has
been regarded by most authors as a full
species since its description by G. M.
Allen (1916). However, after corre-
sponding with Oldfield Thomas, Allen
(1917:168) noted that torrei "should
best be considered as a subspecies of
fuUginosa," a point of view with which
I fully agree.

Specimens examined (119). — CUBA. Ha-
hana: Ciieva de Mudo, Giiines, 6 (ROM);
Cueva del Infreno, near Tapaste, 1 (USNM).
Las Villas: Cueva de Colon, Yaguajay, Punta
Cagiianes, 2 (AMNH). Oriente: Baracoa, 15
(USNM); 8 mi E Baracoa, 24 (USNM); Cue-
va de Majana, 6 (USNM); San Carlos Estate,
Guantanamo, 1 (USNM); Banabacoa, Dos
Caminos, 1 (USNM). Pinar del Rio: Cueva
del Rio, San Vicente, 9 (7 ASC, 2 TCWC);
Cueva del Indio, San Vicente, 44 (AMNH);
San Vicente, 1 (LACM); Guanajay, 9
(USNM).

Pteronotus personatus (Wagner)
Wagner's Mustached Bat

( Synonymy under subspecies )

Distribution. — From southern Sono-
ra and Tamaulipas, Mexico, southward
(excluding portions of the Mexican
Plateau and most of the Yucatan Penin-
sula) through Central America (al-
though not yet recorded from British
Honduras or Nicaragua) into South
America (including Trinidad) as far as
the Mato Grosso region of Brasil (Fig.
28). The geographic range of this spe-
cies in the Amazon Basin is poorly
known.

Description. — Size small; uropata-
gium and wing membranes attach to
ankle by way of a short ligament that
is tightly bound to tibia; tragus spatu-
late, secondary fold moderately well de-
veloped and shelf-like, clothed with
short hairs ( Fig. lOB ) ; labio-nasal plate
simple, margin above nostrils smooth or
with several wart-like tubercles, nostrils
not separated by deep emargination,
lacking lateral spikes on either side of
nostrils (Fig. IIB); anterio-medial edge
of pinna with three to four (occasion-
ally as many as six) tooth-like serra-
tions; ears united by two low, incon-
spicuous ridges that fuse on top of muz-
zle and form a prominent rostral tuber-
cle; profile of skull relatively flat, ros-
trum slightly elevated, forehead sloping
gradually onto long, oval-shaped brain-
case; upper incisors reduced, slight dia-
stema between outer incisor and canine;
inner incisors distinctly bifurcate, with
a narrow, rounded heel; lower incisors
reduced and having short rounded heel,
inner pair distinctly trilobed, outer pair
weakly trilobed; small, shelf-like cus-
pule on anterior lingual surface of post-
centrocrista (terminology of Van Valen,
1966:9) on Ml and M2 (this cuspule,
found in no other mormoopid, is in a
position that roughly approximates that
of the metaconule); infraorbital foramen
opening just slightly anterior to maxil-
lary root of zygomatic arch, canal

88

MISCELLANEOUS PUBLICATION MUSEUM OF NATURAL HISTORY

20

10

0 200

Figure 28. Distribution of Pteronotus personatus. The type locality, San Vicente, Mato Grosso,

Brasil, is not shown.

formed by a thin spicule of bone (a
similar condition exists in Mormoops;
the infraorbital canal opens at the end
of a short tubular canal in all other
species of the genus Pteronotus, Fig.
18).

Relationships. — Aside from the
unique characteristics described above,
personatus, more than any other species
of the genus Pteronotus, approximates a
morphologically intermediate link be-
tween the subgenera Phyllodia and
Pteronotus. Externally, personatus most
closely resembles the Antillean species
fuliginosus and niacleayii that are here
included with it in the subgenus Chilo-
mjcteris. Cranially, personatus resem-
bles P. clavyi, but differs from that
species in having a narrower, less up-
turned rostrum; more prominent mastoid
flanges; and a comparatively smaller
braincase. The skull of P. parnellii also
resembles that of personatus, but is
more massive in structure, has a more
inflated braincase, and markedly heavier
dentition.

Geographic variation. — Unfortunate-

ly, most of the available material of this
species is from Mexico, thereby making
difficult an accurate assessment of geo-
graphic variation within the species as a
whole. Generally, as was the case in P.
parnellii, there is a trend toward in-
creased size, cranially as well as ex-
ternally, in the southern part of the
range (Table 2 and Fig. 29). The
"break" between small bats in the north
and larger bats of Panama and the
Caribbean coast of South America (in-
cluding Trinidad) is emphasized by the
paucity of specimens from certain areas
in Central America. Future investiga-
tions no doubt will reveal a more exten-
sive zone of intergradation between the
two subspecies here recognized. The
somewhat larger size of specimens from
El Salvador (locality 10, here assigned
to the subspecies P. p. psilotis) may re-
sult from limited gene flow along the
Pacific versant of Middle America. Fur-
ther comment on this point, however,
must await specimens from Nicaragua
and additional material from Costa
Rica. Contrary to Handley (1966:759),

SMITH: SYSTEMATICS OF MORMOOPID BATS

89

ri^.

A

B

Figure 29. Dice-squares illustrating geographic variation in Pteronottis personatus from selected

mainland localities. The measurements illustrated (in mm) are: A, length of maxillary toothrow;

B, zygomatic breadth. See text for key to localities and figure 26 for explanation of symbols.

there seem to be sufficient differences
between northern and southern popula-
tions to merit tlie recognition of two
geographic races, as suggested by Good-
win and Greenhall (1961:323).

Key to localities used in statistical analysis
of geographic variation. — 1, Sinaloa; 2, Naya-
rit; 3, Ocampo, Tamaulipas, and vicinity of El
Salto and Valle, San Luis Potosi; 4, Jalisco and
Colima; 5, vicinity of Tequesquitengo, Morelos,
and Teloloapan, Guerrero; 6, 3 km E San
Andres Tuxtla, Veracruz; 7, Oaxaca; 8, Lan-
quin Cave, Guatemala; 9, western Honduras;
10, El Salvador; 11, San Jose, Rio Tinto, Hon-
duras; 12, Penonome, Panama; 13, Cartagena,
Colombia; 14, northwestern Venezuela.

Pteronotus personatus personatus

(Wagner)

Chilonycteris personata Wagner, 1843, Wieg-
mann's Archiv fiir Naturgesch., 9, p. 367,
31 December, in accordance with the In-
ternational Code of Zoological Nomencla-
ture (Art. 21b, ii, p. 19).

Holotype. — Adult male, preserved in
alcohol, skull not removed, Naturhis-
torisches Museum, Vienna, Austria (no

registration number); from Mato Gros-
so, Brasil; obtained by Johann Natterer.
Wagner (1847:185) restricted the type
locality to St. Vicente, Mato Grosso,
Brasil.

Distribution. — From western Costa
Rica south into South America, where it
occurs along the Caribbean coast, in-
cluding Trinidad. Only one specimen
was examined from south of the coastal
region, this being from Los Patos, 28
km SE El Manteco, Bolivar, Venezuela.
No other specimens, aside from the
holotype, are known from the Amazon
Basin (Fig. 28). Two specimens, re-
ported as this species by Da Cunha
Vieira (1942:272) from Cuiaba and
Tapirapoa, Mato Grosso, Brasil, are
Pteronotus parneUii ruhiginosus (see Da
Cunha Vieira, 1955:358).

Descriptio n . — Wagner ( 1843 : 367 )
described the holotype as follows: "Ch.
fusca, suptus dihitior, dorso piJoso." The
length of forearm was given as 1.8
inches [about 46.0]. In a later paper,
Wagner (1847:185-186) compared "Ch.

90

MISCELLANEOUS PUBLICATION MUSEUM OF NATURAL HISTORY

TABLE 2. Selected external and cranial measurements of several populations of Pteronotus personatus.
Each measiuement is represented hy the obser\ed range, below which is the mean, plus or minus
two standard errors of the mean; superscript numbers indicate sample size if diflerent from that

given in left-hand column. See text for key to localities.

Locality

N

Length of
Foreaini

Zygomatic
Breadth

Rostral
Breadth

Length of
Maxillary
Toothrow

Condylobasal
Length

Depth of
Cranium

Pteronotus personatus

psilotis

1

7

42.1+0.53"

8.3±0.11

5.8+0.10

5.9+0.06

13.7+0.06

7.4+0.08

(41.1_42.9)

(8.1-8.5)

(5.7-6.1)

(5.8-6.0)

(13.6-13.8)

(7.2-7.5)

3

6

42.1±0.43

8.4+0.07

6.0+0.08

5.9+0.06

14.0+0.14

7.5+0.14

(41.5-42.9)

(8.3-8.5)

(5.8-6.1)

(5.8-6.0)

(13.7-14.2)

(7.2-7.7)

5

23

42.7±0.33^

8.3+0.07^'

5.9+0.04^

5.9+0.04^'

13.9+0.10

7.6+0.07

(41.1-44.4)

(8.0-8.6)

(5.7-6.0)

(5.8-6.1)

(13.4-14.4)

(7.2-7.9)

6

47

43.2±0.24

8.3+0.05

5.9+0.05

5.9+0.04

13. 9 ±0.07

7.4+0.05

(40.8-45.0)

(8.1-8.7)

(5.6-6.2)

(5.7-6.2)

(13.4-14.4)

(7.0-7.9)

8

29

42.7+0.35-'^

8.3+0.06

5.9+0.04"'"

5.9+0.03

13.8+0.08

7.2+0.04

(40.8-43.9)

(7.9-8.5)

(5.7-6.1)

(5.7-6.0)

(13.2-14.1)

(7.0-7.5)

10

9

44.0+0.62

8.5+0.15^

6.1+0.06

6.0+0.05

14.1+0.14

7.5±0.08

(42.6-45.2)

(8.3-8.9)

(6.0-6.2)

(5.9-6.1)

(13.8-14.5)

(7.3-7.7)

11

3

42.5±0.71

8.3+0.13

5.9+0.20

6.0+0.18

13.9+0.35

7.3+0.24

(41.8-43.0)

(8.2-8.4)

(5.7-6.0)

(5.8-6.1)

(13.6-14.2)

(7.0-7.5)

Pteronotus personatus ])crsonatus

12

16

44.9±0.36="

8.6+0.11^"

6.1+0.07

6.2+0.06

14.4+0.11

7.5+0.10'-^

(43.1-46.8)

(8.4-9.0)

(6.0-6.4)

(6.0-6.5)

(14.0-14.7)

(7.1-7.8)

13

n

45.8±0.54'^

8.9±0.14

6.3+0.08

6.4+0.04

14.6+0.15

7.6+0.10

(44.5-47.4)

(8.5-9.3)

(6.1-6.5)

(6.3-6.5)

(14.3-15.0)

(7.3-7.9)

14

2

44.6±1.30

6.1+0.50

6.2+0.20

14.1+0.30

7.5+0.30

(44.0-45.3

(5.9-6.4)

(6.1-6.3)

(14.0-14.3)

(7.3-7.6)

persotiata" to several other mormoopids
and distinguished it on the basis of its
small size and hair-covered back. Be-
cause Natterer's specimen had been
preserved in alcohol, Wagner ( 1847 )
commented only that dorsally it was
black or dark brown and ventrally it
was pale brown and that the tips of the
hairs of the venter were pale. Selected
external measurements of the holotype,
taken from Wagner ( 1847 ) unless other-
wise noted are as follows (measure-
ments originally given in inches and
tenths are here converted to milli-
meters ) : length of body, 54; length of
tail, 19.5; length of ear, 13; length of
forearm (measured by Dilford C.

Carter), 47.2; length of tibia (measured
by Carter), 17.0; approximate length of
skull (measured by Carter), 16.0.

Specimens herein assigned to P. p.
personatus are large for the species, both
cranially and externally. Standard ex-
ternal measurements of three males and
one female, respectively, from Armila,
San Bias, Panama, are as follows: total
length, 72, 68, 75, 73; length of tail
vertebrae, 18, 18, 20, 20; length of hind
foot, 11, 10, 12, 11; length of ear from
notch, 17, 17, 18, 17. For additional se-
lected external and cranial measure-
ments, see table 2.

Dilford C. Carter (personal commu-
nication) described the color of the

SMITH: SYSTEMATICS OF MORMOOPID BATS

91

holotype as "reddish brown" dorsally
and paler ventrally, but also noted that
it was faded. Most of the specimens of
this subspecies that I have examined
were in old pelage that was Ochraceous
Tawny to Ochraceous Orange. Several
adults from northwestern Colombia,
captured in May, June, and July, were
in various stages of molt. The fresh
pelage is dark brown (Front's Brown)
dorsally, slightly darker (Bister) on top
of the head. Ventrally, these specimens
are grayish white with a buffy tinge to
Cinnamon-Buff; the throat and chin are
white to buffy white.

Comparisons. — From P. p. psilotis,
the subspecies P. p. persoiuitus can be
distinguished on the basis of its larger
cranial and external size (Table 2 and
Fig. 30). Goodwin and Greenhall
(1961:223) stated that their three speci-
mens from Trinidad were smaller than
Mexican specimens of this species. The
two specimens from which they reported
measurements are females and fall at
the lower end of variation of P. p. per-
sormtus. However, the length of the
forearm, which shows the least amount
of secondary sexual variation, for these
females was given as 45.7 and 44.0,
which places them well within the ob-
served variation of P. p. personatus.
Based on the few specimens in fresh
pelage, personatus appears to be only
slightly darker than psilotis.

Remarks.— Be la Torre (1955:697)
argued that until it could be demon-
strated that psilotis and "persormta"
were indeed conspecific, that the two
taxa be maintained as full species, and
this suggestion has been followed by
most mammalogists. In a paper report-
ing on mammals from Venezuela, J. A.
Allen (1911:264-265) employed the spe-
cific name "per sonata" for specimens
from Aroa, Bolivar Railroad, Yaracuy,
and discussed these in relation to psilo-
tis Dobson, 1878, from an unknown
locality (subsequently restricted to Te-
huantepec, Oaxaca, Mexico, by de la
Torre, 1955:696). Allen (1911:265)

quoted a portion of a letter from G. S.
Miller, Jr., who had compared some of
the Venezuelan material with specimens
from Mexico. In this quotation. Miller
indicated that the Venezuelan material
was slightly different and that he was
inclined to recognize the two groups at
least as subspecies. Goodwin and Green-
hall (1961:223) recognized the distinct-
ness of South American bats and applied
the name Chilonycteris persoimta per-
sonata to specimens from Trinidad.
They did not elucidate, however, other
than to point out that '^ Chilonycteris
personata Wagner, and Chilonycteris
psilotis Dobson, have been considered
as a conspecific by some authors."

Wagner (1847), in further describ-
ing Natterer's holotype of personatus
from San Vicente, Mato Grosso, Brasil,
noted its small size in comparison to
"Ch. ruhi<!,inosa" and its apparent simi-
larity to "Ch. Mac-Leayii" and "Loho-
stonia quadriclens," of Gundlach (see
account of P. m. macleayii for discus-
sion of quadridens) in having four
tooth-like serrations on the ears. Meas-
urements given by Wagner (1847) for
the holotype of personatus, some of
which have been converted to milli-
meters and listed above, are only slightly
larger than those of specimens formerly
assigned to psilotis from northern South
America, as are measurements of the
holotype taken by Dilford C. Carter.
There is little question in my mind that
personatus of Wagner should be ap-
plied to specimens from northern South
America, which, as noted above, are dis-
tinctly larger than bats from Mexico and
adjacent Middle America. The alterna-
tive would be to retain personatus as a
species, known only from the holotype,
and to apply the specific name psilotis
in the manner suggested by de la Torre
(1955:697), which would necessitate
proposing a new subspecific name for
the populations from Costa Rica, Pana-
ma, northern South America, and Trini-
dad. This would lead to considerable
nomenclatural confusion if I am correct

92

MISCELLANEOUS PUBLICATION MUSEUM OF NATURAL HISTORY

in

the assumption that personatus and

psilotis are conspecific.

Specimens examined (91). — COSTA RICA.
Guanacaste: AA ini S Liberia, 1 (LACM).
Puntarenas: Rincon de Osa, Osa Productos
Forestales, Camp Seattle, 35 m, 1 (LACM).
PANAMA. Code: Penonome Caves, 23 (3
MCZ, 20 USNM). San Bias: Armila, Qiie-
brada Venado, 4 (USNM). COLOMBIA.
Antiogua: Turbo, 5 (ROM). Bolivar: Carta-
gena, 28 (8 ROM, 20 USNM). Cordoba:
Planeta Rica, S of Monteria, 4 (ROM). Giia-
jira: Nazaret, 4 (ROM). Sucre: Tola Viejo,
5 (ROM). VENEZUELA. Bolivar: Los Patos,
28 km SE El Manteco, I (USNM). Carahoho:
Rio Yaracuy, 10 km NW Urama, 25 m, 1
(USNM). Sucre: La Biea, 1 (AMNH). Yara-
cuy: Minas de Aroa, Bolivar R. R., 12
(AMNH). TRINIDAD. No specific locality,
1 (USNM).

Pteronotus personatus psilotis (Dobson)

Chilonycteris psilotis Dobson, 1878, Catalogue
of the Chiroptera in the . . . British Mu-
seum, London, p. 451.

Chilonycteris torrei continentis Sanborn, 1938,
Occas. Papers Mus. Zool., Univ. Michigan,
373:1, 26 May. Type locality, Laguna de
Zotz, Peten, Cuatemala. Holotype, adult
male, preserved in alcohol, skull removed,
UMMZ 77085; obtained on 4 March 1935
by C. L. Hubbs and H. Van de Schalie.

Sijntypes. — Adult male, preserved in
alcohol, skull removed, BM 50.8.29.3,
and adult female, preserved in alcohol,
skull not removed, BM 50.8.29.4; from
an unknown locality (restricted to Te-
huantepec, Oaxaca, Mexico, by de la
Torre, 1955:696); obtained by Lt.
Stricklan.

Distrilmtion. — From southern parts
of Sonora and Tamaulipas, Mexico,
southward (excluding portions of the
Mexican Plateau and Yucatan Penin-
sula) to eastern Honduras and El Sal-
vador (Fig. 28).

Description. — Contrary to the state-
ment of Goodwin and Greenhall (1961:
223), this subspecies averages smaller
externally and cranially than does P. p.
personatus (Table 2 and Fig. 30).
Standard external measurements (aver-
age, with extremes in parentheses) of
18 males and 12 females, respectively,
from 3 km E San Andres Tuxtla, 1000

A

B

C

Figure 30. Dorsal view of skulls of Pteronotus
personatus psilotis (A, Sinaloa; B, Cuatemala)
and P. p. personatus (C, Colombia). All speci-
mens are adult males; horizontal line above A
is 10 mm in length.

ft, Veracruz, are as follows: total
length, 66.8 (64.0-74.0), 67.1 (63.0-
70.0); length of tail vertebrae, 17.0
(15.0-19.0), 17.6 (16.0-20.0); length of
hind foot, 10.9 (10.0-12.0), 11.2 (10.0-
12.0); length of ear from notch, 16.3
(16.0-17.0), 16.2 (16.0-17.0).

The length of forearm of the syn-
types (measured by J. Knox Jones, Jr.)
were approximately 41.3 for the male
and approximately 43.6 for the female
(measurements are approximate be-
cause forearms appeared to be partly
fractured). Selected cranial measure-
ments (taken by Jones) of the male
syntype are as follows: zygomatic
breadth, 8.7; breadth of braincase, 8.0;
mastoid breadth, 8.9; zygorostral length,
10.0; interorbital constriction, 3.6; ros-
tral breadth, 6.0; length of maxillary
toothrow, 5.8; condylobasal length, 13.4;
length of mandibular toothrow, 7.0. See
table 2 for additional selected external
and cranial measurements of this sub-
species.

Most of the specimens of psilotis
examined were in old pelage that ranged
in color dorsally from reddish brown to
Ochraceous Tawny or brilliant Och-
raceous Orange. Ventrally, the pale tips
of the hairs arc markedly reddish and
in some cases almost yellowish, evi-
dently owing to changes with age of
the pelage. In fresh pelage, these bats
are gray to brownish (Snuff Brown)
dorsally, with the top of the head being

SMITH: SYSTEMATICS OF MORMOOPID BATS

93

somewhat darker. A pale-colored cen-
tral zone is faintly discernible on the
hairs of the back. Hairs of the pectoral
an abdominal regions are strikingly bi-
colored, with brown bases (Front's
Brown to Snuff Brown) and tips that
are pale grayish white or Cinnamon
Buff. Two adult males from near Vaca,
Sinaloa, caught in November, are dark
brown dorsally and olive-brown ven-
trally (terminal parts of ventral hairs
grayish white).

Comparisons. — For comparison with
P. p. personatus, see preceding account.

Remarks. — The general resemblance
between P. personatus and the Antillean
P. fuliginosus led Sanborn (1938:1-3) to
describe Guatemalan specimens as only
subspecifically distinct from Chilomjc-
teris torrei of Cuba (herein assigned as
a subspecies of P. fuliginosus). He dis-
tinguished C. t. continentis from other
mormoopids on the basis of attachment
of the wing and uropatagium to the
ankle and the projection of the calcar
past the "free edge" of the uropatagium.
The projection of the calcar in this
fashion is more or less typical of the
Antillean species P. fuliginosus and
macleayii; it also is present, but con-
siderably more variable, in P. persona-
tus. The holotype of continentis and
other Guatemalan specimens examined
clearly are of the same subspecies as
those from farther north in Mexico.

Although there is some variation,
Mexican and northern Middle American
representatives of this subspecies are
uniformly small in size. In southeastern
Honduras and eastern El Salvador (lo-
cality 10), there is a rather marked in-
crease in overall size possibly indicating
intergradation between psilotis and per-
sonatus. Because there is a certain de-
gree of overlap in measurements of bats
from El Salvador and those from more
northerly locations, I have tentatively
assigned these to the subspecies P. p.
psilotis.

Specimens examined (426). — MEXICO.
Campeche: 65 km S, 1228 km E Escarcega, 2
(KU). Chiapas: 10 mi NE Mai Paso, 500 ft.

2 (TCWC); 3 mi SSE SoyokS, 3000 ft, 1
(TCWC); Santo Domingo, near Cnpia, 4 km
SSE Chiapa de Corzo, approx. 1300 ft, 1
(TCWC); 38 mi SSE Tuxtla Gutierrez, 1800
ft, 2 (TCWC); Rio Dorado, 40 mi SSE Tuxtla
Gutierrez, 1800 ft, 2 (TCWC); 5 mi N Ar-
riaga, 800 ft, 1 (TCWC); 10 mi S Zapaluta,
3000 ft, 2 (TCWC); 20 km SE Pijijiapan, 2
(LACM); 11 km NW Escuinth, 100 ft, 2
(LACM). Colima: 7 mi W, 0.5 mi S Santi-
ago, 6 (KU); Plaija de Ore, 8 km W, 2 km S
Santiago, 16 (KU); Cueva de La Fabrica, 4
mi S Coquimatkin, 18 (LACM); Pueblo Jua-
rez, 1 (AMNH). Guerrero: 12 km N Zaca-
tula, 1 (ENCB); Rio de Los Sabinos, 16 km
E (by road) Teloloapan, 1040 m, 6 (IB); 10
mi E, 2 mi S Teloloapan, 1 (KU); 2 mi NW
Acapulco, 50 ft, 16 (13 KU, 3 MSU). Jalisco:
5 mi S Grullo, 3100 ft, 2 (KU); AutUn, La
Grana, 12 (IB). Morelos: Alpuyeca, 3500
ft, 3 (TCWC); 2kmW Tequesquitengo, 1000
m, 7 (ENCB); Tequesquitengo, 4 (AMNH);
3.4 km S Tequesquitengo, 840 m, 16 (IB).
Nayarit: Huajicori, Rio del Bajar, 7 (LACM);
La Cuchara, approx. 40 mi E Acaponeta, 2
(LACM); San Bias, 5 (MVZ). Oaxaca: San
Geronimo, 1 (USNM); 6 mi NW Mixtequilla,

3 (TCWC); Tapanatepec, 375 ft, 55 (54
AMNH, 1 KU); JO mi S Tapanatepec, 10
(AMNH). San Luis Potosi: El Salto, 1750
ft, 3 (2 AMNH, 1 KU); 1 km S El Salto, 2
(KU); cave on Rancho Nacimiento de Rio
Coy, 16 mi S El Salto, 6 (4 TCWC, 2 USNM).
Sinaloa: 0.5 mi SE Vaca, 650 ft, 2 (KU); 1
mi S, 6 mi E El Cariizo, 1 (KU); 1.5 mi N
Badiraguato, 750 ft, 2 (KU); Cueva China-
caterra, 4.5 mi S, 8 mi W Pericos, 800 ft, 3
(TCWC); La Cruz, 30 ft, 1 (KU); 1.5 mi E
Concordia, 250 ft, 2 (MSU); Chupadero, Rio
Pdnuco, 26 mi NE Villa Union, 800 ft,
1 (LACM); 1 mi W Matatan, 2 (KU);
0.5 mi S Concepcion, 250 ft, 4 (KU).
Tamaulipas: 7 mi W Ocampo, 2400 ft, 1
(KU). Veracruz: Cueva de Laguna Encan-
tada, 3 km ENE San Andres Tuxtla, 1000 ft,
42 (TCWC); 3 km E San Andres Tuxtla, 1000
ft, 44 (KU). GUATEMALA. Alta Verapaz:
Lanquin Cave, 1022 ft, 42 (41 KU, 1 AMNH).
Chiquimula: Jocotan, near Chiquimula, 1350
ft, 1 (KU). Peten: Laguna Zotz, 1 (UMMZ).
HONDURAS. Comayagiia: 1 km W Coma-
yagua, 580 m, 19 (TCWC). Copdn: Ruinas
de Copan, 2 (TCWC). Ocotepcque: 4 km N
Nueva Ocotepeque, 840 m, 4 (TCWC). Olan-
cho: San Jose Rio Tinto, 340 m, 3 (TCWC).
Santa Barbara: 5 km NE Llama, 120 m, 4
(TCWC). Valle: Rio Lanre, 10 km E San
Lorenzo, 25 ft, 2 (TCWC). EL SALVADOR.
Cuscutldn: 2 km W Suchitoto, 390 m, 2
(TCWC); San Rafael Cedros, 710 m, 2
(ROM). Morazan: Graveyard Mine, 1.5 mi
NW Divisadero, 600 ft, 15 (MVZ). Finca El
Marne, 8 km S Santa Ana, 960 m, 1 (ROM).

94

MISCELLANEOUS PUBLICATION MUSEUM OF NATURAL HISTORY

Additional marginal record. — Sonora: Rio
Cuchujaqui, 8 mi by road SSE Alamos, Baker
and Christianson (1966:310).

Subgenus Pteronotus Gray

Type species. — Pteronotus davyi
Gray.

Distribution. — Central Sonora and
southern Tamaulipas, Mexico, south-
ward (excluding portions of the Mexi-
can Plateau and Baja California)
through Central America into South
America, south to the Mato Grosso re-
gion of Brasil and west to the Province
of Piura in northern Peru. The species
P. davyi also occurs on Trinidad and
the Lesser Antillean islands of Grenada,
Dominica, Marie Galante (off Guade-
loupe), and Martinique. For additional
information concerning distribution, see
accounts beyond and figures 33 and 35.

Description. — Size medium to large;
wing membranes fused dorsally at mid-
line, resulting in a naked-backed ap-
pearance; wing membrane and uro-
patagium attached to ankle by way of
a long ligament (nearly half as long as
tibia) that is tightly bound to tibia;
labio-nasal plate moderately complex in
structure, margin above each nostril
ornamented by a series of irregularly
shaped, wart-like tubercles, lateral
spikes present on either side of nostrils
(Fig. IIC); tragus moderately complex,
spatulate, with prominent secondary
fold and distal tip curled craniad, and
with moderately long hairs ornamenting
distal and cranial edges (Fig. 10, E-F);
antero-medial edge of pinna smooth
(lacking tooth-like serrations); ears not
noticeably connected by ridge, rostral
tubercle only weakly developed; pelage
monocolored or only slightly bicolored
dorsally (hair beneath dorsally fused
wing membranes noticeably longer than
that exposed in the scapular, nape, and
head regions); profile of skull relatively
flat, but rostrum short and noticeably
upturned; forehead relatively abruptly
elevated onto elongated braincase; prox-
imal nasal root distinctly "scooped out"

forming a pronounced basin; rostrum
extremely broad in contrast to length as
exemplified by relatively short maxillary
toothrow; upper incisors reduced in size,
inner pair bifurcate and with a narrow
rounded heel directed toward canine,
inner and outer incisors of each premax-
illary compressed together, forming two
groups with just the inside tips of the
inner two incisors touching (this char-
acter shared with the subgenus Chilo-
nycteris); diastema present between
outer incisor and canine; lower incisors
reduced in size and trilobed, with short
rounded heel; tympanic ring large,
nearly covering auditory bulla; basioc-
cipital region between bullae, extremely
l)road and cup-shaped, having two shal-
low, oval-shaped pits into which the M.
longus capitus inserts.

Remarks.— The fusion of the wing
membranes at the dorsal midline is a
peculiar feature of bats of this subgenus.
Such a condition is not found in any
other microchiropteran, but does exist
in several megachiropterans. The func-
tional significance of this curious fea-
ture is not known. Perhaps it is in-
volved in the increased dissipation of
heat during periods of activity. It may
also function, in several ways, during
flight; the most obvious of which might
be to contour the posterior portion of
the body or to increase the effective
lifting surface. Conceivably, this ar-
rangement also could act as a "drag
chute," giving the bat a greater degree
of maneuverability. These are only
some possibilities and close observation
of living naked-backed bats will be
necessary to reveal the true functional
significance of this condition.

The species of this subgenus, P.
davyi and suapurensis, are morpho-
logically similar. The primary differ-
ence between these two species is in
cranial and external size, although some
proportional and qualitative differences
also exist (see descriptions beyond).
With the exception of P. fuliginosus and
macleayii, the two naked-backed species

SMITH: SYSTEMATICS OF MORMOOPID BATS

95

represent the only case in the family
Mormoopidae where close morphologi-
cal similarity is coupled with sympatry
over a broad area; P. parnelUi and per-
sonatus are broadly sympatric, but, as
should be apparent from the foregoing
discussions, these two species are quite
distinct morphologically. The pattern
of intraspecific geographic variation in
davyi and suapurensis suggests the pos-
sibility of character displacement: in
general, davyi increases in overall size
in progressively more southern parts of
its range, whereas this trend is reversed
in suapurensis, which increases in over-
all size at progressively more northerly
latitudes. In every case of sympatry be-
tween these two taxa, suapurensis is the
larger of the two. G. M. Allen (1935:
227) apparently was unaware of these
trends in geographic variation and re-
ported specimens of suapurensis from
Penonome, Code, Panama, as Pterono-
tus davyi suapurensis because of the
similarity in size between his Pana-
manian bats, typical davyi from Trini-
dad, and J. A. Allen's (1904) descrip-
tion of suapurensis. This is the only case
known to me in which these two taxa
have been considered as conspecific.

Pteronotus davyi Gray
Davy's Naked-backed Bat

(Synonymy under subspecies)

Distribution. — From central Sonora
and southern Tamaulipas, Mexico,
southward (excluding portions of the
Mexican Plateau and Baja California),
including the Yucatan Peninsula,
through Central America into South
America, where its distribution is not
well documented. The species occurs
at least as far south as northern Peru
(Piura Province) and possibly, in the
Amazon Basin, to the Mato Grosso re-
gion of Brasil (see remarks below, con-
cerning Chilonycteris gymnonotus Wag-
ner, 1843). The geographic range of P.
davyi also includes northern Venezuela,
Trinidad, and Lesser Antillean islands
of Grenada, Dominica, Marie Galante

(off Guadeloupe), and Martinique (Fig.

33).

Description. — Size small to medium,
length of forearm less than 50; wing
membrane, which is fused dorsally,
sparsely covered by long, irregularly-
spaced hairs.

Relationships. — Pteronotus davyi is
closely related to suapurensis. How-
ever, these two taxa are easily distin-
guished by cranial and external size
(Tables 3 and 4 and Fig. 34). In addi-
tion, the pubescence of the dorsally
fused wing membrane — shorter and
denser in suapurensis — also separates
the two. Apparently, P. davyi has di-
verged less from the main line of evolu-
tion than has suapurensis within the
genus Pteronotus. As mentioned above,
davyi somewhat resembles P. (Chilo-
nycteris) personatus in size and general
shape of the skull. However, other fea-
tures, such as structure of the tragus
and labio-nasal plate, broader and more
elevated rostrum, and dorsally fused
wing membranes readily distinguish the
subgenus Pteronotus from Chilonycteris.

Geographic variation. — The paucity
of material from Central America to the
south of El Salvador and Honduras, and
from South America, limits an accurate
definition of geographic variation in this
species. Until now, there was a sizeable
gap in the known geographic range,
extending from Nicaragua into South
America, where the species was known
only from Grenada and Dominica in the
Lesser Antilles, Trinidad, and Brasil.
Starrett and Casebeer (1968:7) reported
material from Costa Rica and I have
examined specimens from western Vene-
zuela and Nicaragua, which suggests
that the species occupies a continuous
distribution from Mexico into South
America.

With the exception of the Sonoran
population, davyi is uniformly small,
both cranially and externally, through-
out Mexico. There is a gradual increase
in overall size that begins in southern
Mexico (localities 9-10) and continues

96

MISCELLANEOUS PUBLICATION MUSEUM OF NATURAL HISTORY

LMkl^

134 13.6 13,8 14.0 14.2 14.4 14.6 148 150 152 154 156

Figure 31. Dice-squares illustrating geographic variation of Pteronotus davyi from selected main-
land localities (see text for key to localities). The measurements illustrated (in mm) are: A, ros-
tral ]:)readth; B, condylobasal length. See figure 26 for explanation of symbols.

into the adjacent region of Guatemala.
In Honduras and El Salvador (localities
16-18), this increase in cranial and ex-
ternal proportions is considerably more
pronounced than in bats to the north
(Table 3 and Fig. 31). The Sonoran
population (locality 1) is of interest in
that these bats average comparatively
larger than those from most other Mexi-
can populations and resemble davyi
from northern Middle America. Sina-
loan bats, although generally more like
typical Mexican populations, are some-
what intermediate between these and
Sonoran bats.

The above features can be seen in
the distance phenogram ( Fig. 32 ) in
which large-sized bats from Sonora,
Morelos, and northern Middle America
are grouped together in the upper cou-
plet (localities 1, 9, and 15-18) which
has a distance coefficient of 0.980. The
remaining Mexican populations are
grouped in the second couplet, which
has a distance coefficient of 0.700. Both
of these couplets are in turn lumped
into a larger couplet with a coefficient
of 1.330, which is widely separated ( dis-

tance coefficient, 3.430) from the bot-
tom couplet that encompasses Nicara-
guan. Lesser Antillean, and South Amer-
ican populations (including Trinidad).
In the latter couplet, davyi from north-
western Peru (locality 20) are segre-
gated by a rather large distance co-
efficient of 1.890.

Basically, the analysis of geographic
variation indicates that there are two
main populations of davyi (contrary to
the findings of Starrett and Casebeer,
1968:8), which meet along a transverse
zone of intergradation that is roughly
located near the northern border of
Nicaragua. Judging from the material
at hand, this zone appears to be ex-
tremely narrow and may represent a
relatively recent zone of secondary con-
tact between the northern and southern
populations. In the northwestern ex-
treme of the Mexican portion of the
northern population, ecological as well
as physiographic factors seem to have
favored large size, independent of the
trend toward large size observed at
southern latitudes.

In the southern population, there is

SMITH: SYSTEMATICS OF MORMOOPID BATS

97

3.500 2.800 2.100 1.400 0.700

I ' I 1 r

I

15

9

16

- 17

HZ

2

3

4

fc

7

12

II

13

5

10

8

14

19

22

21

20

Figure 32. Phenogiam of geographic variation
of Pteronoius daviji computed from the distance
matri.x- on standardized locality means and
clustered by the unweighted pair-group method
using arithmetic averages (UPGMA). The
cophenetic correlation for this phenogram is
0.857. See text for key to localities.

a small amount of variation between
bats from Dominica, Trinidad, Costa
Rica, and Nicaragua. However, there is
considerable difference between these
bats and those from northwestern Peru.
In view of the distinctness of the Peru-
vian bats, and the possibility that they
represent an isolated population in the
semiarid Pacific lowlands, west of the
Andes, I have elected to recognize them
as a new subspecies.

Remarks.— Wagner (1843:367) pro-
posed the name Chilonycteris gymno-
nottis for a naked-backed bat sent to
him by Johann Natterer from Cuyaba
{= Cuiaba), Mato Grosso, Brasil. Later,
Wagner (1847:183-185) redescribed and
illustrated this species. In conjunction
with the description of Lonchorhina
ourita. Tomes (1863:83) commented
that "the bottle from which this speci-
men [holotype of L. aurita] was taken
contained several West Indian species,
in which the Mormops hlainvillii [prob-

ably Monnoops megalophylla] and the
Chilonycteris gymnonota of Wagner
were conspicuous. The latter is distin-
guished from other species of the genus
by having the wing-membrane spring-
ing from the middle of the back, instead
of the sides of the body; and there can
be but little doubt tliat it is the Ptero-
notus davyi of Dr. Gray." This is ap-
parently the first time gijmnonotus was
synonymized with davyi, an arrange-
ment that has prevailed until now.
Thomas (1892:162) restricted the type
locality of Lonchorhina aurita Tomes, to
Trinidad, which suggests that Tomes'
specimens of "C. gymnonota' also may
have come from there. At any rate, the
precedent seems to have been set and
C. gymnonotus has been consistently
regarded as a synonym of P. davi/i. J. A.
Allen (1904:229) described the large
naked-backed species Pteronotus sua-
purensis from Suapure, Bolivar, Vene-
zuela, but made no reference to C gym-
nonotus. Examination of Wagner's
(1847) comments and measurements of
gymnonotus reveals that it is somewhat
larger than typical P. davyi. Da Cunha
Vieira (1942:268-270) reported a naked-
backed bat, a female, from San Luis de
Caceres, Mato Grosso, Brasil, which he
assigned to P. davyi; length of forearm
was given as 50 for this specimen. The
length of forearm of the holotype of
gymnonotus was given by Wagner
(1843) as (converted) approximately
48. Dilford C. Carter (personal com-
munication), examined the holotype of
Wagner's naked-backed species and
found that the forearm was approxi-
mately 51, but that botii forearms were
broken. The approximate length of the
skull was 17.5, which is slightly larger
than the holotype of suapurensis. The
length of the maxillary toothrow of the
holotype of suapurensis (examined by
me) was 7.3, whereas Carter obtained
7.0 for the holotype of gymnonotus. At
this point, it seems highly probable that
Wagner's naked-backed bat, P. gymno-
notus, is not a synonym of P. davyi, but

98

MISCELLANEOUS PUBLICATION MUSEUM OF NATURAL HISTORY

20

10

0 200

Scale of Miles

Figure 33. Distribution of Pteronotus daviji. A few localities in central South America (Brasil and
Peru) are not shown, including the occurrence of the subspecies P. d. incae.

actually a prior description of the spe-
cies known as P. suapureiisis. However,
in view of the paucity of material from
the Amazon Basin, and because I have
not personally examined Wagner's holo-
type, I tentatively retain the name
suapurensis for the large naked-backed
species and regard gymnonotus as a
synonym of P. d. davyi. If the holotype
of gymnonotus is critically compared
and found to be conspecific with sua-
purensis, I would certainly advocate the
recognition of gymnonotus as the valid
name for the large naked-backed species
in the interest of stability within Chi-
ropteran nomenclature.

Key to localities used in statistical analysis
of geographic variation. — 1, Sonora; 2, Sina-
loa; 3, Tamaulipas; 4, EI Salto, San Luis Poto-
si; 5, Jalisco; 6, Colima; 7, vicinity of Zacatula,
Acapulco, and Acahuizotla, Guerrero; 8, vicin-
ity of Teloloapan and Colotlipa, Guerrero; 9,
vicinity of Tequesquitengo, Morelos; 10, 3 km
E San Andres Tuxtla, Veracruz; 11, vicinity of
Tehuantepec and Tapanatepec, Oaxaca; 12,
western Chiapas; 13, central Chiapas; 14, Cam-
peche; 15, Lanquin Cave, Guatemala; 16,
eastern Honduras; 17, western Honduras; 18,

El Salvador; 19, 3 mi E San Ramon, Nica-
ragua; 20, Piura Province, Peru; 21, Trinidad;
22, Dominica.

Pteronotus davyi davyi Gray

Pteronotus davyi Gray, 1838, Mag. Zool. Bot.,
2:500, February.

Chilonycteris gymnonotus Wagner, 1843, Wieg-
mann's Archiv fiir Naturgesch., 9, p. 367,
31 December (in accordance with the In-
ternational Code of Zoological Nomencla-
ture, Art. 21b, ii, p. 19). Type locality,
Cuyaba (=Cuiaba) Mato Grosso, Brasil.
Holotype, adult female, preserved in alco-
hol, skull removed, Naturhistorisches Mu-
seum, Vienna, Austria ( no registration
number); obtained by Johann Natterer in
February; original no. 88. See remarks in
species account of P. davyi above.

Holotype. — Adult, sex not deter-
mined, preserved in alcohol, skull re-
moved, BM 9.1.4.74; from Island of
Trinidad; collector unknown.

Distribution. — From Nicaragua south
through Central America into South
America, where it is known from several
localities along the Caribbean coast of
Venezuela to Trinidad; also on the
Lesser Antillean islands of Grenada,

SMITH: SYSTEMATICS OF MORMOOPID BATS

99

TABLE 3. Selected external and cranial measurements of several jDopulations of Pteronotus davyi.
Each measurement is represented by the observed range, ]:)elo\v which is the mean, plus or minus
two standard errors of tlie mean; superscript numbers indicate sample size if difl'erent from that
given in the left-hand column. See text for key to locaUties.

Locality

N

Length of
Forearm

Zygomatic
Breadth

Rostral
Breaddi

Length of
Maxillary
Too throw

Condylobasal
Length

DepUi of
Cranium

Pteronotus davyi fulvus

1

15

45.0±0.45^''

8.7+0.10

6.7+0.07''

6.4+0.05

14.3+0.10"

7.5+0.04

(43.3-46.4)

(8.3-9.0)

(6.5-6.9)

(6.3-6.6)

(13.9-14.6)

(7.4-7.6)

2

11

43.7±0.56"

8.5+0.12'"

6.5+0.08

6.2+0.07

14.1+0.14

7.3+0.15'"

(42.3-45.2)

(8.3-8.8)

(6.3-6.6)

(6.0-6.4)

(13.7-14.4)

(7.7-7.7)

3

10

44.0+0.56

8.5+0.07

6.5+0.06"

6.2+0.07

14.0+0.12

7.3+0.13

(42.5-45.2)

(8.4-8.7)

(6.4-6.7)

(6.0-6.4)

(13.7-14.3)

(7.0-7.6)

9

13

44.2±0.65

8.7+0.11

6.6+0.09

6.4+0.07

14.2+0.12'^

7.5+0.08'^

(42.4-46.7)

(8.3-9.0)

(6.3-6.9)

(6.2-6.6)

(13.9-14.6)

(7.3-7.8)

10

59

43.8±0.29

8.6+0.04-"'^

6.5+0.04

6.3+0.03

14.3+0.06-'"

7.4+0.04^

(40.6-46.2)

(8.3-9.0)

(6.2-7.0)

(6.0-6.5)

(13.7-14.7)

(7.1-7.7)

15

22

45.1+0.51'-'

8.7+0.08

6.7+0.06'^

6.4+0.04-*

14.3+0.07

7.4+0.06

(43.7-46.9)

(8.4-9.1)

(6.4-7.0)

(6.2-6.7)

(14.1_14.8)

(7.2-7.6)

17

17

45.2±0.49

8.8+0.07

6.8+0.08

6.5+0.04

14.6+0.07

7.5+0.09

(43.7_47.8)

(8.5-9.0)

(6.4-7.0)

(6.4-6.6)

(14.2-14.8)

(7.1-7.8)

18

12

45.3+0.52'"

8.8+0.14

6.7+0.08

6.5+0.09

14.5+0.14

7.5+0.19

(44.3_46.7)

(8.4-9.2)

(6.5-7.0)

(6.3-6.8)

(14.2-15.1)

(6.7-8.2)

Pteronotus davyi dc

ivyi

19

14

46.1 ±0.82

9.0+0.13

6.9+0.13

6.7+0.07

14.9+0.13

7.7+0.15

(44.0-49.2)

(8.7-9.5)

(6.5-7.3)

(6.4-6.9)

(14.4-15.3)

(7.2-8.1)

21

4

47.3+2.90^

9.1

7.0+0.13

6.8+0.10

14.7+0.13

7.8+0.24

(45.9-48.8)

(6.9-7.2)

(6.7-6.9)

(14.6-14.9)

(7.6-8.1)

22

10

46.1+0.64''

9.0+0.07

7.0+0.07

6.6d=0.08

14.9+0.13

7.7+0.06

(45.4-46.9)

(8.9-9.2)

(6.8-7.2)

(6.5-6.9)

(14.7-15.2)

(7.5-7.8)

Pteronotus davyi incae

20

8

49.3+0.24

9.3+0.08

7.2+0.16

6.9+0.08

15.4+0.16

8.1+0.11

(48.6-49.6)

(9.1-9.4)

(6.9-7.5)

(6.7-7.0)

(15.0-15.7)

(7.9-8.3)

Dominica, Marie Galante (off Guade-
loupe), and Martinique (Fig. 33).

Description. — Size medium, both ex-
ternally and cranially, for species ( Table
3 and Fig. 34). Standard external meas-
urements (average, with extremes in
parentheses) of 17 males and nine fe-
males, respectively, from 8 mi E San
Ramon, Matagalpa, Nicaragua, are as
follows: total length, 78.3 (73.0-81.0),
79.3 (76.0-83.0); length of tail, 22.1

(18.0-24.0), 22.2 (20.0-24.0); length of
hind foot, 11.2 (10.0-12.0), 11.3 (10.0-
12.0); length of ear from notch, 17.3
(16.0-19.0), 17.0 (16.0-18.0). See table
3 for additional selected external and
cranial measurements.

Selected cranial measurements (taken
by Dilford C. Carter) of the holotype
are as follows: zygomatic breadth, 9.0;
breadth of braincase, 7.9; mastoid
breadth, 8.7; rostral breadth, 7.2; length

100

MISCELLANEOUS PUBLICATION MUSEUM OF NATURAL HISTORY

of maxillary toothrow, 6.5; length of
mandibular toothrow, 7.9; condylobasal
lengtli, 14.1; depth of cranium, 8.2.

Individuals taken in March from
Dominica are brownish (Snuff Brown)
dorsally. There is only a faint indica-
tion of color zones on the exposed hairs
of the back, the basal portion being only
slightly darker than the remainder of
the hair. Ventrally, the pelage is mark-
edly bicolored; the bases of the hairs
are dark brown (Front's Brown) the
tips are paler, giving the venter an
overall pale brown appearance. A large
series collected on 9 May from near San
Ramon, Nicaragua, continued individ-
uals that were bright Ochraceous
Orange in color, as were specimens of
suapurensis captured there at the same
time.

Comparisons. — See following ac-
counts for comparisons with other sub-
species.

Remarks. — The paucity of specimens
of davyi from southern Middle America
and South America unfortunately limits
study of geographic variation within the
species. Although, there appears to be
a zone of intergradation located in
northern Nicaragua and eastern Hon-
duras and El Salvador, the Nicaraguan
bats are strikingly larger in overall size
than are those from more northerly
populations. In Central America, these
bats seem to occur mostly along the
drier Pacific coast and in South Amer-
ica, P. d. davyi appears to favor the
northern coastal region where condi-
tions also tend to be somewhat arid.

This race represents the only occur-
rence of mormoopid bats in the Lesser
Antillean islands. I suspect this invasion
is relatively recent (perhaps late Pleis-
tocene), judging from the absence of
any noteworthy morphological differ-
ences between Dominican specimens
and those from northeastern South
America. There can be little doubt that
the initial invasion of the Lesser An-
tilles occurred from northeastern South
America or from Trinidad.

Specimens examined (171). — NICARA-
GUA. Chontales: Cuapa, 4 (KU). Matagal-
pa: 3 mi E San Ramon, 126 (KU). COSTA
RICA. Alajela: 4.6 mi W Atenas, 1 (LACM);
road from Atenas to San Mateo, 1 (LACM).
Quanacaste: Playa de Coco, 1 (LACM); 5 mi
N Liberia, 1 (LACM). VENEZUELA. Cara-
boho: Rio Yaracuy, 10 km NW Urama, 25 m,
1 (USNM). Distrito Federal: El Limon, 3
km S, 46 km W Caracas, 380 m, 1 (USNM).
Sucre: Ensenada Cauranta, 7 km N, 5 km E
Guiria, 1 m, 1 (USNM). Yaracmj: Minas de
Aroa, Bolivar R. R., 1 (AMNH). TRINIDAD.
Port of Spain, 9 (1 AMNH, 7 FMNH, 1
MCZ); Careiiage, 2 (AMNH); Nariva, 1
(AMNH); Aripo, 1 (AMNH); Siparia, 1
(AMNH). DOMINICA. St. George: Roseau,
ca. 50 ft, 2 (USNM). St. Luke: South Chil-
tern Estate, ca. 1300 ft, 2 (USNM). St. Pat-
rick: Grand Bay, sea level, 7 (USNM). No
specific locality, 8 (5 FMNH, 3 USNM).
GRENADA. No specific locality, 1 (BM).

Additional marginal records. — LESSER
ANTILLES (Jones and Phillips, 1969). Marie
Galante (off Guadeloupe); Martinique. VEN-
EZUELA. Araguaimujo, Territorio Delta Ama-
curo, Pirlot (1965:379).

Pteronotus davyi fulvus (Thomas)

Ch[ilonycteris]. Davyi fulvus Thomas, 1892,
Ann. Mag. Nat. Hist., ser. 6, 10:410,
November.

Pteronotus davyi fulvus Miller, 1912, Bull.
U. S. Nat. Mus., 79:33, 31 December.

Pteronotus suapurensis calvus Goodwin, 1958,
Amer. Mus. Novit., 1871:1, 26 February.
Type locality, Tehuantepec, Oaxaca, Mex-
ico. Holotype, adult male, skin and skull,
AMNH 175100; obtained by Thomas Mac-
Dougall on 15 August 1956; original no.
1729.

Holotype. — Adult male, skin and un-
cleaned skull, BM 93.2.5.24; from Las
Perias, Jalisco, Mexico; obtained by A.
C. Buller on 20 November 1891.

Distribution. — From southern parts
of Sonora and Tamaulipas Mexico,
southward (excluding portions of Mexi-
can Plateau and Baja California), in-
cluding the Yucatan Peninsula, to east-
ern Honduras and El Salvador (Fig.
33).

Description. — Size, small for the spe-
cies (Table 3 and Fig. 34). Standard
external measurements (average, with
extremes in parentheses) of 13 males
and 11 females, respectively, from 3 mi

SMITH: SYSTEMATICS OF MORMOOPID BATS

101

B

C

D

F

Figure 34. Dorsal view of skulls of Pteronotus daviji fiilviis (A, Sinaloa; B, Guatemala), P. d. davyi

(C, Nicaragua; D, Dominica Island), P. d. incae (E, Suyo, Peru), and Pteronotus suapnrensis (F,

Nicaragua). All specimens are adult males; horizontal line aljove B is 10 nmi in length.

N El Jicaro, Juchitan, Oaxaca, are as
follows: total length, 73.1 (71.0-75.0),
72.7 (71.0-75.0); length of tail, 21.5
(20.0-23.0), 21.7 (20.0-23.0); length of
hind foot, 11.0 (all individuals); length
of ear from notch, 17.9 (17.0-18.0), 17.7
(17.0-18.0). Additional selected exter-
nal and cranial measurements are given
in table 8. Length of forearm of the
holotype is 43.5, and selected cranial
measurements (taken by J. Knox Jones,
Jr.) are as follows: zygomatic breadth,
8.2; mastoid breadth, 8.4; rostral
breadth, 6.5; length of maxillary tooth-
row, 6.3.

Dorsally the color of this subspecies
ranges from bright Ochraceous Orange
to dark brown. October-taken individ-
uals from 15 km W Ameca, Jalisco, are
in fresh pelage and the exposed pelage
on the back is dark brown (Bister to
Front's Brown ) ; color zones on the hairs
are not readily apparent. Pelage of the
pectoral and abdominal regions is gray-
ish vv'hite. The basal half of these hairs
is dark brown and the tips are grayish
white or buffy white.

Comparisons. — From P. d. davyi, the
subspecies P. d. fulvus may be recog-
nized easily on the basis of smaller size
even though fulvus becomes somewhat
larger in the southern part of its range
(Honduras and El Salvador). Skulls of
P. d. davyi are longer and broader than
those of P. d. fulvus and, in direct com-
parison, appear to be more robust ( Fig.
34).

Remarks. — After examining the holo-
type of Pteronotus suapurensis calvus
Goodwin, 1958, I concluded that it is a
large individual of P. d. fulvus. Good-
win evidently now shares this view, be-
cause he (1969:55) placed P. s. calvus
in the synonymy of P. d. fulvus.

Mexican populations (excluding So-
noran bats) show a relatively small
amount of geographic variation (Fig.
31, localities 2-13). This same situation
also seems to apply to populations from
the northern part of the Yucatan Penin-
sula, although only a few individuals
were available for study from that re-
gion. In northern Middle America, bats
of this subspecies gradually increase in
size as they approach a narrow zone of
intergradation with P. d. davyi in
northern Nicaragua and eastern Hon-
duras and El Salvador. The exact nature
of gene flow between these two races in
this region is not clear.

One of the most curious features of
the geographic variation within this
subspecies is the marked increase in
overall size of the bats from Sonora
(locality 1 in table 3 and figure 31),
where males and females also exhibit
the greatest amount of secondary sexual
variation ( Table 1 ) . As discussed in the
account of Pteronotus parnellii mexi-
canus, this part of the geographic range
of the genus Pteronotus is at the north-
ern and arid extreme of the Neotropical
environment. In this region, factors
such as suboptimal habitat and possibly

102

MISCELLANEOUS PUBLICATION MUSEUM OF NATURAL HISTORY

the presence (or the absence) of other
insectivorous competitors may in some
way influence the size of mormoopids.
In parneUii the trend was toward smaller
size, whereas with davyi the trend has
favored large size with a greater degree
of sexual dimorphism. The break be-
tween Sonoran and Sinaloan populations
may be somewhat exaggerated owing to
the way in which Sinaloan samples were
grouped but, in any event, the trend in
the geographic variation appears to be
real and certainly warrants further in-
vestigation.

Specimens examined ( 597 ) . — MEXICO.
Campeche: 5 km S Champoton, 10 mi, 2
(KU); Apozote, 1 (USNM); 1 km N, 13 km
W Escdrcega, 65 m, 2 (KU); 12 km W Escdr-
cega, 2 (KU); 105 km E Escarcega, 1 (KU);
La Tuxpena, 2 (USNM). Chiapas: Palenque,

1 (FMNH); 6 km NE Mai Paso, 400 ft, 1
(TCWC); Santo Domingo, near Cupia, 4 km
SSE Chiapa de Corzo, 1300 ft, 7 (TCWC);
Tuxtla Gutierrez, 1 (USNM); 3 mi SE Tuxtla
Gutierrez, 1 (AMNH); 35 mi SSE Tuxtla
Gutierrez, 2200 ft, 1 (TCWC); Rio Dorado,
40 mi SSE Tuxtla Gutierrez, 1800 ft, 1
(TCWC); 13 mi SW Las Cruces, 7 (KU);
15 mi SW Las Cruces, 1 (KU); 10 mi S
Zapaluta, 3000 ft, 1 (TCWC); 8 mi N Arri-
aga, 2500 ft, 1 (TCWC); 6 mi N Arriaga, 600
ft, 4 (TCWC); 5 mi N Arriaga, 800 ft, 3
(TCWC); 5 km SE Pijijiapan, 100 ft, 4
(TCWC); 11 km NW Escuintla, 100 ft, 1
(LACM); 1.8 mi W Ixhuatdn, 1 (AMNH).
Colima: 7 mi W, 0.5 mi S Santiago, 15 (KU);
Plaija de Oro, 8 km W, 2 km S Santiago, 2
(KU); Punta Perm Blanca, 3 (MSU); 4 km S
Coquimatla, 10 (LACM); El Mixcoate, 9 km
W Pueblo Judrez, 1347 m, 11 (IB). Guerrero:

2 mi S, 10 mi E Teloloapan, 20 (10 IB, 10
KU); 12 km N Zacatula, 21 (ENCB); Mex-
cala, 2 (TCWC); Grutas de Juxlahuaca, 4 mi
N Colotlipa, 3 (1 KU, 2 TCWC); 4 km NE
Colotlipa, 1100 m, 7 (ENCB); Acahuizotla,
2800 ft, 7 (TCWC); 2 mi NW Acapulco, 50
ft, 12 (11 KU, 1 MSU). Jalisco: 15 km W
Ameca, 4200 ft, 1 (KU); El Zapote, 3 (IB);
0.5 km N, 3 km W Jamay, 1650 m, 22 (18
ENCB, 4 IB); 5 mi S Grullo, 3100 ft, 1 (KU).
Michoacdn: 7 mi S Tumliiscatio, 2700 ft, 1
(KU); Lake Chapala, 6 (AMNH). Morelos:
2 km W Tequesquitengo, 1000 m, 10 (ENCB);
3.4 km S Tequesquitengo, 840 m, 6 (IB).
Nayarit: 2 mi S Compostela, 2900 ft, 1 (KU).
Oaxaca: 18 mi NW Sola de Vega, 1 (AMNH);
5 mi N Juchatengo, 2 (CAS); 3 mi NE Jicaro,
Juchitan, 24 (AMNH); Tehuantepec, 19
(AMNH); 6 mi 7VW Mixtequilla, 3 (TCWC);

4 mi E Tapanatepec, 800 ft, 2 (TCWC); 10
mi S Tapanatepec, 3 (AMNH). San Luis
Potest: El Salto, 7 mi N Naranjo, 26 (AMNH).
Sinaloa: 1 mi S, 6 mi E El Carrizo, 2 (KU);
1 mi E Sinaloa, 180 ft, 1 (KU); La China-
catera, Monte Largo, 23 km W Pericos, 11
(IB); San Juan, 8 mi S San Ignacio, 1 (KU);
Panuco, 2050 ft, 9 (KU); 3 mi SE Plomosas,
4000 ft, 1 (KU). Sonera: 13 mi S Carbo,
1200 ft, 7 (TCWC); 14 mi SE Carbo, 1500
ft, 4 (USNM); Cueva del Tigre, 14.9 mi SE
Carbo, 14 (8 LACM, 6 USNM); 20 mi SE
Carbo, 2700 ft, 2 (LACM); Rancho Bana-
chari, 10 mi N Matape, 2700 ft, 1 (MVZ).
Tamaulipas: Rancho Santa Rosa, 25 km N, 13
km W Ciudad Victoria, 260 m, 10 (KU);
Cueva de Los Trocones, 7.5 km W Ciudad
Victoria, 6 (IB). Veracruz: Cueva de Lagima
Encantada, 3 km ENE San Andres Tuxtla,
1000 ft, 46 (TCWC); 3 km E San Andres
Tuxtla, 1000 ft, 45 (KU); Achotal, 1
(FMNH). Yucatan: Hoctun Cave, 14 (MVZ);
Chichen Itza, 1 (FMNH). GUATEMALA.
Alta Verapaz: Lanquin Cave, 1022 ft, 50
(KU); 1 7ni W Lanquin, 2000 ft, 5 (TCWC).
Chiquimula: Jocotan, near Chiquimula, 2
(KU). HONDURAS. Comayagua: Coma-
yagua, 580 m, 10 (TCWC). Copdn: Ruinas
de Copan, 3 (TCWC). El Paraiso: 1 km SE
Danli, 780 m, 1 (TCWC); Chichicaste, 480
m, 1 (TCWC). Intibucd: La Esperanza, 1660
m, 1 (TCWC). La paz: Cueva del Viejo, 2
mi W La Paz, 1800 ft, 13 (TCWC). Oco-
tepeque: Ocotepeque, 840 m, 1 (TCWC).
Santa Barbara: 2 km S San Nicolas, 660 m, 12
(TCWC); 7 km N Santa Barbara, 120 m, 1
(TCWC). Yoro: Subirana, 2800 ft, 7 ( MCZ).
El SALVADOR. Cuscatldn: 2 km W Suchi-
toto, 390 m, 3 (TCWC); San Rafael Cedros,
no m, 2 (ROM). La Libertad: 5.1 km E
La Libertad, 15 m, 1 (TCWC). La Union:
Tabanco, 650 ft, 14 (MVZ). Morazdn: En-
cuentros Mine, 3 mi W Divisadero, 700 ft, 1
(MVZ). Santa Ana: Finca El Marne, 8 km
SW Santa Ana, 960 m, 3 (ROM).

Pteronotus davyi incae new subspecies

Holotype. — Adult male, skin and
skull, TCWC 11638; from 4 mi W Suyo,
1000 ft, Piura Province, Peru; obtained
by Dilford C. Carter on 28 July 1964,
original no. 5313.

Distribution. — Known only from
northwestern Peru [not mapped].

Description. — Size, externally and cra-
nially large for the species (Table 3 and
Fig. 34). Standard external measure-
ments of the holotype followed by those
of two other adult males from the type

SMITH: SYSTEMATICS OF MORMOOPID BATS

103

locality are as follows: total length, 84,
76, 82; length of tail, 25, 21, 25; length
of hind foot, 9, 9, 9; length of ear from
notch, 18, 17, 17. Selected external and
cranial measurements of the holotype
are as follows: length of forearm, 49.3;
zygomatic breadth, 9.4; breadth of
braincase, 8.4; mastoid breadth, 9.7;
zygorostral length, 11.8; interorbital
breadth, 4.0; rostral breadth, 7.5; length
of maxillary toothrow, 7.0; length of
mandibular toothrow, 8.1; condylobasal
length, 15.7; depth of cranium, 8.0.

The holotype and type series, cap-
tured on 28 July, appear to be in fresh
pelage, which is pale grayish brown
(Dusky Drab) dorsally; color zones are
not apparent on the exposed hairs of
the back. The pectoral and abdominal
regions are grayish white (Mouse
Gray), whereas the throat and chin are
somewhat paler. A December-taken
male and a January-taken female from
northwestern Peru, are Ochraceous
Orange in color.

Comparisons. — From P. d. davyi and
P. d. fulvus, the new subspecies is easily
discernible on the basis of larger cranial
and external size (Table 3 and Fig. 31).
In addition to being longer, the skull of
incae is decidedly broader (especially
the rostral portion) than either that of
davyi or fidvus. Although, it is difficult
at this time to evaluate color accurately
in these bats, incae appears to be some-
what paler than either of the other two
races.

From P. suapurensis, specimens of
P. d. incae can be easily distinguished
by smaller cranial and external size and
the long and scattered hairs on the dor-
sally fused wing membrane, in compari-
son to the short, densely spaced hairs
of suapurensis.

Remarks. — Bats of this subspecies
appear to be isolated on the western
side of the Andes in northwestern Peru
and there seem to be several possible
routes by which they could have arrived
in this region. One of these may have
been through migration along the nar-

row corridor formed by the Pacific ver-
sant of the Andes in Colombia and
Ecuador. Haffer (1967a: 327) and Chap-
man (1926) suggested that part of the
bird fauna of the arid Pacific coast of
Colombia and Ecuador may have used
this route before tropical forest devel-
oped at the beginning of the Pleistocene.
The development of such a forest prob-
ably blocked subsequent dispersal of
non-forest species along this route.
Other alternatives involve trans-Andean
crossings along several arid intermon-
tane valleys in Ecuador and northern
Peru. Of these, there seem to be only
two valleys through which these bats
could have passed in order to gain ac-
cess to low passes over which such a
crossing could have occvured. One of
these passes is located between the up-
per Patia and Cauca valleys in southern
Colombia and northern Ecuador. This
route may have been used by Mormoops
megalophyllo, but as yet P. davyi is not
known from this region. The other and
perhaps more feasible route in the case
of davyi may have been through the
valley of the Rio Marafion from the
Amazon lowlands during an arid epoch,
thence up the connecting valley of the
Rio Huancabamba to the low pass at
its headwaters. The passes in both of
these regions are relatively low and the
landmass over which a migrant would
have passed is rather narrow (Tuttle
1970:54). In addition, both regions are
low enough to have been equitable dur-
ing drier periods of the Pleistocene.
Considerablv more needs to be known
about the Chiropteran fauna of Colom-
bia and Ecuador, particularly in these
intermontane valleys, along the eastern
slopes of the Andes and the adjacent
Amazonian lowlands.

The subspecies name incae is pro-
posed to commemorate the civilization
of the Inca that was once centered in
the region of northwestern Peril.

Specimens examined (15). — PERU [locali-
ties not mapped]. Cajaniarca: Jaen, 1
(AMNH). Piura: 4 mi W Suyo, 1000 ft, 3

104

MISCELLANEOUS PUBLICATION MUSEUM OF NATURAL HISTORY

20

Scale of Miles

Figure 35. Distribution of Pteronottis stiaptirensis. A few locations in South America (Peru) are

not shown.

(TCWC); Hacienda Bigotes, Salitral, 7
(FMNH); Huancabamba, 2 (AMNH); Can-
chaque, 2 (FMNH).

Pteronotus suapurensis (J. A. Allen)
Big Naked-backed Bat

Dermonotus suapurensis J. A. Allen, 1904, Bull.
Amer. Mus. Nat. Hist., 20:229, 20 June.

Pteronotus suapurensis J. A. Allen, 1911, Bull.
Amer. Mus. Nat. Hist., 30:265, 2 Decem-
ber.

Pteronotus suapurensis centralis Goodwin,
1942, Jour. Mamm., 28:88, 16 February.
Type locality Matagalpa, 3000 ft, Mata-
galpa, Nicaragua. Holotype, adult, sex
unknown, skin and skull, AMNH 33193;
obtained by W. B. Richardson on 10
November 1911.

Holotype. — Adult female, skin and
skull, AMNH 17573; from Suapure,
Bolivar, Venezuela; obtained by S. M.
Klages on 26 October 1901.

Distribution. — From near San An-
dres Tuxtla, Veracruz, Mexico, south-
ward through northern Guatemala into
Central America and South America,
where it occurs, at least as far south as
Tingo Maria, Peru; possibly to the Mato
Grosso region of Brasil (see remarks

under Pteronotus daviji concerning Chi-
lonycteris gymnonotus Wagner, 1843);
and west of the Andes in Piura Prov-
ince, Peru (Fig. 35).

Description. — Size medium to large,
both externally and cranially. Standard
external measurements (average, with
extremes in parentheses), for 10 males
and three females, respectively, from 3
mi E San Ramon, Matagalpa, Nicaragua,
are as follows: total length, 89.7 (86.0-
93.0), 90.7 (89.0-92.0); length of tail
vertebrae, 24.8 (21.0-27.0), 24.7 (24.0-
26.0); length of hind foot, 12.1 (11-0-
13.0), 12.7 (12.0-13.0); length of ear
from notch, 18.6 (18.0-20.0), 18.3 (17.0-
19.0). Additional selected external and
cranial measurements are given in table
4. The wing membranes are fused on
the dorsal midline and are covered with
many short hairs giving the back a
"velvety" appearance.

Specimens collected on 9 May near
San Ramon, Nicaragua, were in bright
Ochraceous Orange pelage as were in-
dividuals of P. davyi captured at the
same time. Two females taken in Janu-

SMITH: SYSTEMATICS OF MORMOOPID BATS

105

TABLE 4. Selected external and cranial measurements of several populations of Pteronotus suapu-

rensis. Each measurement is represented by the observed range, below which is the mean, plus or

minus two standard errors of the mean; superscript numbers indicate sample size if different from

that given in left-hand column. See text for key to localities.

N
Locality and
Sex

Length of
Forearm

Zygomatic
Breadth

Rostral
Breadth

Length of
Maxillary
Toothrow

Condylobasal
Length

Depth of
Cranium

Pteronotus suapurensis

1

1 $ 53.7

10.5

8.0

7.5

16.4

8.8

2

2 9 54.1,53.5

10.4, 10.1

8.2, 7.6

7.3, 7.5

16.2, 16.2

8.6, 8.5

3

5 9 54.0±0.50

10.3+0.10

7.9+0.19

7.5+0.14

16.4+0.29

8.6+0.16

(53.1-54.5)

(10.1-10.4)

(7.7-8.2)

(7.4_7.7)

(16.1-16.8)

(8.4-8.8)

4

28 5 52.6±0.31^

10.4+0.06

7.9+0.06

7.5+0.05

16.4+0.06

8.8+0.06

(51.1-54.2)

(10.2-10.8)

(7.7-8.2)

(7..3-7.7)

(16.1-16.7)

(8.5-9.2)

37 9 53.6+0.39

10.4+0.06

7.9+0.06

7.5+0.04

16.4+0.06

8.8+0.06

(50.8-56.0)

(10.0-10.6)

(7.6-8.2)

(7.2-7.8)

(15.7-16.9)

(8.3-9.0)

5

10$ 53.1+0.57

10.5+0.04

8.0+0.05

7.6+0.06

16.5+0.11

8.9+0.19^

(52.1-55.0)

(10.4-10.6)

(7.9-8.2)

(7.5-7.8)

(16.3-16.8)

(8.7-9.5)

4 9 53.2+1.36

10.4+0.31'

7.9+0.16

7.5+0.06

16.4+0.14

8.8+0.20^

(51.4—54.7)

(10.1-10.6)

(7.7-8.1)

(7.5-7.6)

(16.2-16.5)

(8.6-8.9)

6

U 52.7,52.0,54.3-^

10.1

7.7

7.4

16.4

8.6

7 9 52.6+O.OF

10.3+0.13^'

7.7+0.20"

7.5+0.13

16.3+0.14

8.6+0.12

(50.9-53.6)

(10.1-10.5)

(7.5-8.2)

(7.2-7.7)

(16.0-16.5)

(8.4-8.9)

7

33 <$ 51.2+0.45"

10.2±0.06

7.8+0.06

7.4+0.04

16.2+0.09™

8.6+0.06

(49.0-54.0)

(9.9-10.5)

(7.4-8.1)

(7.1-7.7)

(15.6-16.7)

(8.3-9.1)

19 9 53.3+0.48^**

10.2+0.07

7.8+0.07

7.5+0.06

16.2+0.11

8.5+0.08

(51.2-54.7)

(9.9-10.6)

(7.6-8.0)

(7.3-7.8)

(15.8-16.6)

(8.1-8.8)

8

15 6 51.2+0.49'''

10.2+0.09

7.7+0.08'^

7.4+0.03'"

16.2+0.11'"

8.6+0.12

(50.2-53.5)

(9.9-10.5)

(7.5-8.0)

(7.3-7.5)

(15.7-16.6)

(8.2-9.0)

10 9 52.9+0.77'

10.1+0.10

7.7+0.10"

7.4+0.09"

16.1+0.14

8.3+0.13

(51.1-54.3)

(9.9-10.4)

(7.5-8.0)

(7.2-7.6)

(15.7-16.4)

(8.0-8.6)

9

1$ 53.0+0.87

10.4+0.14

7.9+0.13

7.6+0.11

16.5+0.11

8.7+0.13

(51.5-54.8)

(10.2-10.6)

(7.6-8.1)

(7.3-7.7)

(16.3-16.7)

(8.5-9.0)

59 53.8+0.93

10.3+0.17

8.0+0.17

7.6+0.04

16.4±0.05

8.6+0.10

(52.7-55.4)

(10.0-10.5)

(7.8-8.3)

(7.5-7.6)

(16.3-16.4)

(8.5-8.8)

10

4 5 51.2+1.59

10.1+0.07^

7.7+0.14

7.2+0.13

15.7+0.45

8.6+0.16

(49.8-52.7)

(10.1-10.2)

(7.5-7.8)

(7.1-7.4)

(15.2-16.1)

(8.4-8.8)

8 9 52.1+0.61

10.0+0.13

7.5+0.15

7.3+0.09

15.7+0.17'

8.4+0.17'

(50.9-53.2)

(9.6-10.2)

(7.2-7.8)

(7.1-7.5)

(15.5-16.1)

(8.0-8.5)

11

19 51.6

10.1

7.7

7.3

15.6

8.5

12

6<? 52.1 + 1.77

10.2+0.13

7.8+0.09^

7.4+0.08'

16.3±0.12

8.8+0.12

(50.8-53.3)

(10.1-10.5)

(7.6-7.9)

(7.3-7.6)

(16.1-16.5)

(8.6-9.0)

13

2 9 51.0,50.5

9.5, 9.7

7.2, 7.2

6.8, 6.9

15.3, 15.5

8.3,8.4

106

MISCELLANEOUS PUBLICATION MUSEUM OF NATURAL HISTORY

ary from Lanquin Cave, Alta Verapaz,
Guatemala, are dark brown dorsally and
ventrally. On the venter, there is a pale
buffy-white wash, the chin and throat
being somewhat darker.

Relationships. — Pteronotus suaptiren-
sis can be distinguished from P. clavyi,
the only other species of the subgenus,
on the basis of larger cranial and ex-
ternal size as well as noticeably greater
pubescence on the dorsally fused wing
membrane.

Geographic variation. — There are so
few specimens of this species presently
available in collections that an accurate
and meaningful analysis of geographic
variation is not yet possible. Several
sizable samples were examined from
Central America, but only a few indi-
viduals are known from South America.

Whether or not the holotype of P.
suapurensis (a rather small adult fe-
male) is truly representative of the
populations of the Amazon Basin can-
not be assessed in the absence of other
comparative specimens from this critical
region. It appears that there is a gen-
eral geographic trend in the species to-
ward larger overall size at progressively
more northerly localities, the largest in-
dividuals occurring in Nicaragua, El Sal-
vador, Honduras, and Guatemala. Bats
of this species from Costa Rica and
Panama are only slightly smaller than
those to the north, and slightly larger
than individuals in a small sample from
Minas de Aroa, Bolivar R. R., Yaracuy,
Venezuela. The Venezuelan bats, how-
ever, frequently fall into non-significant
subsets of means with both Nicaraguan
and Panamanian samples; in most in-
stances Panamanian specimens are in-
termediate in size between those from
Venezuela and Nicaragua. The few
specimens examined from Tingo Maria
and Loreto, Peru, closely resembled
Panamanian specimens although aver-
aging somewhat smaller in cranial and
external size.

Two females from Hviancabamba,
Piura, Peru, are slightly smaller than

bats from Tingo Maria and Loreto and
may represent an isolated population.
However, because the geographic varia-
tion is slight between the few sizable
samples of the species currently avail-
able, it seems best to treat the species as
monotypic, a course followed by Felten
(1956:77-78), Davis and Carter (1964:
377), Jones (1966:449), and Starrett
and Casebeer (196(S:9). Additional ma-
terial from South America, may indicate
that two or more geographic races war-
rant recognition.

Key to localities used in statistical analysis
of geographic variation. — 1, Cueva Laguna
Encantada, 3 km ENE San Andres Tuxtla,
Veracruz; 2, Lanquin Cave, Guatemala; 3,
Honduras; 4, vicinity of Di\isadero, El Salva-
dor; 5, 3 mi E San Ramon, Nicaragua; 6,
southern Costa Rica; 7, Penonome, Panama;
8, Chilibrillo, Panama; 9, Armila, Panama; 10,
Minas de Aroa, Venezuela; 11, Suapura,
Venezuela; 12, Tingo Maria, Peru; 13, Piura
Pro\'ince, Peru.

Specimens examined (344). — MEXICO.
Veracruz: Cueva Laguna Encantada, 3 km
ENE San Andres Tuxtla, 1000 ft, 1 (TCWC).
GUATEMALA. Alta Verapaz: Lanquin Cave,
1022 ft, 2 (KU). HONDURAS. El Paraiso:
Chichicaste, 480 m, 2 (TCWC). Francisco
Morazdn: 2 mi S Zamorano, 1 (TCWC).
Olancho: San Jose, Rio Tinto, 340 m, .5
(TCWC). EL SALVADOR. CuscatUn: 2
km W Suchitoto, 390 m, 45 (TCWC); Suclii-
toto, 400 m, 2 (ROM); San Rafael Cedros,
710 m, 1 (ROM). La Union: Tabanco, 650
ft, 3 (MVZ). Morazdn: Graveyard Mine, 1.5
mi NW Divisadero, 600 ft, 17 (MVZ). Santa
Ana: Finca El Marne, 8 km S Santa Ana, 960
m, 1 (ROM). NICARAGUA. Chontales:
Cuapa, 1 (KU). Matagalpa: Matagalpa, 1
(AMNH); 3 mi E San Ramon, 13 (KU).
Zelaya: Cara de Mono, 50 m, 4 (KU); 10
km W Rama, 40 m, 1 (TCWC). COSTA
RICA. Pnntarenas: 35 mi S San Isidro, Rio
Ceiba, 1 (LACM); Rincon de Osa, Oso Pro-
ductos Forestales airstrip, 35 m, 2 (UMMZ);
2.7 mi N Villa Neilly, 6 (LACM); Villa Neilly,
2 (LACM). PANAMA. Canal Zone: Madden
Dam, 1 (USNM). Chiriqiii: Cerro Punta,
Casa Tilley, 1 (USNM); Bamhito, 35 km N
Concepcion, 6200 ft, 2 (TCWC); 4 mi from
El Volcdn, 1 (ANSP); San Vicente, 1
(USNM). Code: Penonome, 118 (58 FMNH,
1 LACM, 11 MCZ, 23 UMMZ, 25 USNM).
Darien: Tacarcuna Camp, 3200 ft, 2 (USNM).
Panama: Chilibrillo Caves, 26 (USNM); Chi-
lihre, 8 (USNM). San Bias: Armila, Que-
brada Venado, 10 (USNM). Veragtias: 3 mi
E Sona, 100 ft, 1 (USNM). COLOMBIA.

SMITH: SYSTEMATICS OF MORMOOPID BATS

107

Bolivar: Cartagena, 36 (1 ROM, 35 USNM).
Sucre: Tolu Viejo, 3 (USNM). VENEZUE-
LA. Bolivar: Siiapure, 1 (AMNH). TriijiUo:
Agua Viva, 18 km N Valera, 164 m, 2
(USNM). Yaracuij: Minas de Aroa, Bolivar
R. R., 11 (AMNH). PERU [localities not
mapped]. Hudnuco: 2 mi N Tingo Maria
2000 ft, 6 (TCWC). Loreto: Yarinacocha,
Rio Ucayali, 1 (FMNH). Fiura: Huanca-
bamba, 2 (FMNH).

Genus Mormoops Leach
Ghost-faced Bats

Mormoops Leach, 1821, Trans. Linn. Soc, 13,
p. 76. See remarks below concerning the
use of the generic name Mormoops.

Aello Leach, 1821, Trans. Linn. Soc, 13, p.
70 (type Aello cuvieri Leach).

Type species. — Mormoops blainvil-
lii Leach, 1821.

Distribution. — Humid to arid and
semi-arid regions, usually below 3000
meters, from southwestern Texas and
southern Arizona, through Mexico (in-
cluding Baja Galifornia, Mexican Pla-
teau, and Yucatan Peninsula) into Cen-
tral America as far as eastern Honduras
and El Salvador. Members of this genus
have not yet been reported from
Nicaragua, Costa Rica, or Panama. In
South America, Mormoops is known
from the Caribbean coast of Colombia
and Venezuela, Trinidad, the Dutch
West Indies, and as far south as La
Paz, Carchi Province, Ecuador. The
genus is represented on the Greater An-
tilles by the small endemic species M.
hlalnviUii, which also is known by a
single lower mandible from a pre-Co-
lombian cave deposit on Exuma Island
in the Bahamas (Koopman, 1951:229).
For further information regarding dis-
tribution, see species accounts beyond
and figures 36 and 40.

Description. — Bats of the genus
Mormoops range in size from small to
large; Mormoops blainvilUi from Cuba
being the smallest representative of the
genus and Mormoops megalophylla from
Ecuador being the largest. Externally,
Mormoops is easily distinguished from
Pteronotus as follows: long, lax pelage;
short and rounded ears that are strongly

connected dorsally on the rostrum by
two, conspicuous ridges; lower edge of
pinna confluent with lower lip; orna-
mentation of lower lip complex, witli
many intricate folds and scallops, form-
ing a shield-like plate (covered with
many wart-like tul^ercles) in front of a
narrow central tubercle; a secondary
transverse ridge below lower lip further
complicates the structure ( when opened,
the mouth forms a funnel-shaped orifice
that, coupled with the long, stiff hairs
located at the sides of the mouth, prob-
ably form an efficient apparatus for
catching insects); labio-nasal plate com-
plex with nostrils surrounded by sep-
arate pads and separated by a long,
wart-like ridge; margin above and be-
tween nostrils with several long, irregu-
larly-shaped tubercles (Fig. IIF);
tragus complex, secondary fold promi-
nent, pubescence restricted to several
long hairs and numerous short, bristles
on the cranial edge (Fig. 10, G-H);
wing membranes attach relatively high
on sides of body but somewhat lower
than in Pteronotus; uropatagium and
wing membrane attach to ankle by way
of a short ligament that is not bound to
the tibia.

The cranium of Mormoops is un-
usual in shape, in that the rostrum is
strongly upturned, and the forehead
rises so abruptly over the braincase that
it is almost at a right angle to the rostral
part of the skull (Fig. 37). Additional
features of the cranium of Mormoops
are: zygomatic arches angled as a result
of rostro-braincase flexion; mastoid
flanges nearly absent and mastoid bone
little more than short spicules above and
behind auditory bullae; parietals notice-
ably inflated in area that encases cere-
belli; infraorbital foramen located at
maxillary root of zygomatic arch (Fig.
18); incisive foramina extremely large;
basioccipital region relatively broad be-
tween bullae and basisphenoid with
cup-like depression, into which the M.
/o/igws capitus inserts.

The molars are typically dilambdo-

108

MISCELLANEOUS PUBLICATION MUSEUM OF NATURAL HISTORY

.^^

30

- 20

80

70

Figure 36. Distribution of Morrnoops blainvillii.

dont and there are striking diastemas in
the upper toothrow between the outer
incisor and the canine as well as be-
tween the first and second premolars;
upper incisors thin and blade-like, heel
on inner pair completely absent; upper
canine deeply grooved on the antero-
medial surfaces; lower incisors small
and delicate, all four are trilobed, heels
absent; second lower premolar long and
narrow; posterior portion of ramus
markedly curved also reflecting cranial
flexion.

Remarks. — Bats of this genus are the
most specialized in the family Mormo-
opidae. Various morphological peculi-
arities seem to have been canalized to-
ward the adaptation of these bats to a
swift-flying insectivorous niche. The
structure of the humerus, long, relatively
narrow wings, funnel-shaped mouth, and
enlarged cerebelli are among a few of
these specializations. Inasmuch as these
and other features are not nearly so
markedly developed in the genus Ptero-
notus, it seems reasonable to suspect

that Morrnoops segregated from the
mainline of mormoopid evolution at a
relatively early stage,

I herein follow Opinion 462 ( 16 [pt.
1]:1-12, April 2, 1957) of the Inter-
national Commission on Zoological
Nomenclature in applying the generic
name Morrnoops in preference to Aello,
which has eight pages of priority.

Morrnoops blainvillii I.each
Antillean Ghost-faced Bat

Morrnoops Blainvillii Leach, 1821, Trans. Linn.
Soc, 13:77.

Aiillo Cuvieri Leach, 1821, Trans. Linn. Soc,
13:71. Type locality unknown, herein pre-
sumed to be Jamaica. Holotype, adult, sex
unknown, skin an skull, BM (no registra-
tion niunber); presented by W. Bullock to
D. Brooke's Museum.

L[obostoma]. cinnamonieum Gundlach, 1840,
Arch. Natvu-gesch., 6:357. Type locality,
Cafetal St. [San] Antonio El Fundador,
Matanzas, Cuba; present disposition of
holotype imknown.

Holotype. — Not certainly located,
presumably in the British Museum (Na-

SMITH: SYSTEM ATICS OF MORMOOPID BATS

109

A

B

C

D

G

H

Figure 37. Dorsal and profile views of skulls of Monnoops megalophylla carteri (A and F, La Paz,

Ecuador), M. jn. tumidiceps (B, near Cartagena, Colombia), M. m. intermedia (C, Curasao), M.

m. megalophylla (D and G, Purificacion, Jalisco), M. hlainvillii (E and H, Cuba). All specimens

are adult males; horizontal line above B is 10 mm in length.

tural History); from Jamaica; presented
by Dom. Lewis to D. Brooke's Musemn.

Distribution. — Greater Antilles, in-
cluding Mona Island between Hispan-
iola and Puerto Rico. Koopman (1951:
229) reported a single mandible found
in a pre-Colombian cave deposit on
Exuma Island in the Bahamas ( Fig. 36 ) .

Description. — Smallest species of the
genus Mormoops, forearm usually less
than 41 (see measurements). Rostrum
extremely upturned (as described
above) and relatively narrow; braincase
globular and with frontal region some-
what inflated, which emphasizes the
abrupt rise of the forehead (Fig. 37);
basioccipital and basisphenoid narrow,
with a wide trough-like depression that
lacks a longitudinal, medial septum; ears
rounded, but with an inconspicuous dor-
sal portion that is short and pointed;
large fleshy papilla on each side of up-
per lip; anterior folds of lower lip
squared and plate-like; tragus complex
and with secondary fold comprising

largest part of the structure; short,
lanceolate flap above secondary fold
(Fig. lOH); tail and hind limbs long
in proportion to length of body.

Most of the specimens of this species
examined in this study were preserved
in alcohol thereby limiting an evalua-
tion of color. In a series of December-
and January-taken males from Oxford
Cave, Jamaica, the pelage was deep
cinnamon brown dorsally and only
slightly paler ventrally. The bases of
individual hairs are pale reddish in
color. It is here presumed that these
individuals were in fresh pelage; paler
and somewhat browner individuals
have been taken at other times in the
year.

Me as u re m ents . — Selected external
and cranial measurements (average,
with extremes in parentheses, plus or
minus two standard errors of the mean)
of nine adult males from Cuba, eight
from Jamaica, six from Hispaniola, and
11 from Puerto Rico (sample size in

110

MISCELLANEOUS PUBLICATION MUSEUM OF NATURAL HISTORY

45.0 46.0 47.0 48.0 49.0
1 1 I I I I 1 1 \ I

A

7.0 7.2 7.4 7.6 7.8
I \ \ ] \ 1 \ 1 \ —

B

n^

.0 8.2 8,4 8.6

.8 9.0

I — r

1 r

1 r

C

Figure 38. Dice-squares illustrating geographic

Antillean islands of Cuba, Jamaica, Hispaniola,

illustrated (in mm) are: A, length of forearm; B

D, depth of cranium. See fig

variation of Mormoops blainvillii from the Greater
and Puerto Rico (top to bottom). The measurements
length of maxillary too throw; C, zygomatic breadth;
ure 26 for explanation of symbols.

brackets if different), respectively, are
as follows: length of forearm, 45.8
(44.8-46.6) ± 0.39, 46.6 (45.7-47.2) ±
0.40, 47.2 (46.2-49.0) ± 0.80. 46.7 (44.8-
48.2) ± 0.72; zygomatic breadth, 8.3
(8.0-8.6) ±: 0.16 [6], 8.5 (8.3-8.8) rb
0.11, 8.6 (8.4-8.9) ± 0.14, 8.6 (8.4-8.8)
± 0.07 [10]; rostral breadth, 6.5 (6.3-
6.6) ± 0.13 [6], 6.5 (6.1-6.7) ± 0.74,
6.7 (6.5-6.9) ± 0.11, 6.8 (6.6-7.0) ±
0.07; length of maxillary toothrow, 7.3
(7.2-7.5) ± 0.10 [6], 7.4 (7.2-7.7) ±
0.13, 7.6 (7.4-7.8) ± 0.11 [7], 7.6 (7.5-
7.8) ± 0.06; condylobasal length, 12.8
(12.6-13.0) ± 0.14 [6], 13.0 (12.8-13.4)
± 0.17, 13.3 (12.8-13.7) ± 0.27, 13.3
(13.0-13.5) ± 0.09 [10]; depth of crani-
um, 8.2 (8.1-8.4) ± 0.11 [6], 8.1 (7.9-
8.3) ± 0.11, 8.4 (8.3-8.7) ± 0.13, 8.4
(8.2-8.7) ± 0.09 [10]. No reliable
standard external measurements are
available.

Relationships. — Bats of this species
are clearly related to the mainland
species Mormoops megalophtjUa, but
the Antillean bats evidently have been
separated from the mainland species for
quite some time. Mormoops bJainviUii,
along witli P. fuliginosus and macleayii,
probably represents a relatively old
mormoopid lineage on the Greater An-
tilles.

Geographic variation. — Analysis of
geographic variation of this species has
been somewhat limited owing to a
paucity of material. Two subspecies,
M. h. blainviUii and M. h. cinnamome-
um (= M. /;. ciwieri Hall and Kelson,
1959:95), have been recognized in the
past, based primarily on the shape of
the first upper premolar. This character,
Iiowever, is variable and appears to be
of little taxonomic importance. The re-
sults of my analysis indicate a more or
less continuous dine from Cuba to

SMITH: SYSTEM ATICS OF MORMOOPID BATS

111

30

20

0 200

Scale of Miles

100

90

FiGXJRE 39. Distribution of Mormoops megalophylla.

Puerto Rico with a slight break or step
between Jamaica and Hispaniola (Fig.
38). Cuban bats are the smallest in
overall size and those from Hispaniola
frequently are the largest, with repre-
sentatives from Puerto Rico equal to, or
slightly smaller than, those from His-
paniola. The greatest differences ob-
served between bats from Cuba and Ja-
maica on the one hand and those from
Hispaniola and Puerto Rico on the
other, were in depth of cranium and
length of maxillary toothrow (Fig. 38).
Individuals of the Jamaican population
were usually intermediate in size be-
tween Cuban and Hispaniolan bats. In
view of the relatively continuous clinal
variation between these populations, I
do not feel that recognition of subspe-
cies is justified and regard M. blainviUii
as monotypic. In any event, the Cuban
and Jamaican populations do not war-
rant recognition as two separate races,
as supposed in the past.

If at a later date, additional evidence
should dictate the necessity of recog-
nizing bats from Hispaniola and Puerto
Rico as subspecifically distinct from the
populations on Cuba and Jamaica, there
will be a nomenclatural problem in-
volving the name Aello cuvieri because
its type locality is unknown. Concern-
ing this matter. Hall and Kelson (1959:
95) stated: "type from unknown locali-
ty, here assumed to be Cuba, but pos-
sibly Jamaica." Both Dobson (1878:
456) and Rehn (1902:162) presumed
that the specimen on which the name
cuvieri was based originated from Ja-
maica, which I feel is reasonable, and I
recommend that the type locality be
restricted to that island.

Specimens examined (125). — CUBA. Ha-
hana: 4 mi San Jose de Las Lajas, 1 (AMNH).
Oriente: Baracoa, 6 (USNM); San Carlos
Estate, Guantanamo, 1 (USNM); Santiago de
Cuba, 1 (AMNH); Caneij, 4 (ROM); Cueva
de la Cantera, Caneij, 1 ( ROM ) ; Cueva de
Las Majaes, Siboneij, 3 (2 AMNH, 1 ROM).

112

MISCELLANEOUS PUBLICATION MUSEUM OF NATURAL HISTORY

TABLE 5. Selected external and cranial measurements of several populations of Mormoops megalo-
phi/Ua. Each measurement is represented by the observed range, below which is the mean, plus or
minus two standard errors of the mean; superscript numbers indicate sample size if different from
that gi\'en in left-hand column. See text for the localities.

Locality

N

Length of
Forearm

Zygomatic
Breadth

Rostral
Breadth

Length of

Maxillary Condylobasal

Too throw Length

Depth of
Cranium

Mormoops megalophylla megalophyUa

1

30

54.9±0.28^

9.6+0.05

7.4+0.03

8.0+0.04

14.3+0.07

9.7+0.06

(53.0-56.5)

(9.4-9.9)

(7.3-7.6)

(7.8-8.3)

(14.0-14.7)

(9.4-10.0)

2

12

54.6+0.58'^

9.5+0.06

7.3+0.07

7.9+0.19

14.1+0.09

9.6+0.10

(52.3-56.3)

(9.3-9.6)

(7.1-7.5)

(7.7-8.9)

(13.8-14.3)

(9.2-9.8)

3

5

54.9±0.77^

9.5+0.13

7.3+0.05

8.0+0.10

14.2+0.10

9.6+0.14

(54.2-56.0)

(9.3-9.7)

(7.3-7.4)

(7.8-8.1)

(14.0-14.3)

(9.5-9.8)

4

6

54.5+0.90

9.5+0.13

7.3+0.04

7.9+0.12

14.1+0.12=

9.6+0.16"

(52.9-56.2)

(9.3-9.7)

(7.3-7.4)

(7.7-8.0)

(14.0-14.3)

(9.3-9.8)

5

4

53.1 ±0.90

9.3+0.05

7.0+0.13

7.8+0.08

13.8+0.21

9.6+0.08

(52.4-54.4)

(9.3-9.4)

(6.9-7.2)

(7.7-7.9)

(13.6-14.0)

(9.5-9.7)

6

27

54.5±0.42

9.4+0.06

7.3+0.05

7.9±0.04

14.1+0.07

9.6+0.08

(51.0-56.8)

(9.1-9.7)

(7.0-7.5)

(7.8-8.1)

; 13.7-14.5)

(9.3-10.0)

7

18

54.4±0.40

9.5+0.09'^

7.3+0.06

7.8+0.06

14.0+0.09"

9.6±o.or-'

(52.6-55.4)

(9.2-9.7)

(7.1-7.5)

(7.6-8.0)

,13.7-14.2)

(9.5-9.9)

8

39

54.3+0.34

9.4+0.06^^*

7.3+0.04

7.9+0.04

14.0+0.07-^'^^

9.6+0.07=^

(51.6-56.4)

(9.0-9.7)

(7.1-7.7)

(7.6-8.2)

J3.6-14.4)

(9.2-10.0)

9

19

54.1 ±0.60^"

9.4+0.08

7.3+0.08

8.0+0.07

14.1+0.12'^

9.6+0.05'"

(51.8-57.2)

(9.1-9.9)

(7.0-7.6)

(7.7-8.2)

13.6-14.6)

(9.4-9.7)

10

40

55.5±0.25^^

9.5+0.04

7.4+0.04

7.9+0.04

14.1+0.06="'

9.7+0.04*

( 53.7-57.2 )

(9.3-9.9)

(7.1-7.8)

(7.5-8.1)

13.8-14.5)

(9.4-9.9)

11

24

55.0+0.36-^'

9.6+0.08"

7.4+0.05

8.0+0.04

14.2+0.10

9.8±0.09^'

(52.5-56.8)

(9.3-10.1)

(7.2-7.7)

(7.8-8.2)

13.6-14.6)

(9.1-10.1)

12

4

53.6±0.48"

9.5+0.08

7.5+0.13

7.9+0.13

13.8+0.21

9.6±0.10

(52.3-54.9)

(9.4-9.6)

(7.3-7.6)

(7.8-8.1) (

13.6-14.1)

(9.5-9.7)

13

32

54.1+0.37

9.5+0.06

7.5+0.04

8.0+0.04

14.0+0.11

9.7+0.05

(52.0-57.3)

(9.1-9.8)

(7.3-7.8)

(7.7-8.3) (

13.6-14.8)

(9.3-10.0)

14

26

55.8+0.59

9.7+0.08^

7.6+0.06

8.1+0.05

14.4+0.15

9.8+0.09

(53.0-58.8)

(9.2-10.1)

Mormoops

(7.2-7.8)
megalophylla

(7.9-8.3) (
intermedia

13.6-14.9)

(9.2-10.1)

17

3

54.1+0.10'

9.4+0.07

7.7+0.18

7.9+0.22*

13.8+0.24

10.0+0.20

(54.0-54.2)

(9.4-9.5)

(7.5-7.8)

(7.6-8.1) (

13.6-14.0)

(9.8-10.1)

18

16

52.0+0.20'"

9.3+0.10''

7.4+0.05

7.7+0.07

13.6+0.14'^'

9.7+0.03

(49.8-53.8)

(9.0-9.7)

(7.2-7.6)

(7.5-7.9) (

13.3-14.4)

(9.4-10.4)

19

2

50.3, 53.4

9.2, 9.6

7.5, 7.6

7.5,7.6

13.1, 13.5

9.3, 9.5

SMITH: SYSTEMATICS OF MORMOOPID BATS

113

TABLE 5. (Continued)

Locality

N

Length of
Forearm

Zygomatic
Breadtli

Rostral
Breadth

Length of
Maxillary
Toothrow

Condylobasal
Length

Depth of
Cranium

Mormoops megalopliyUa t

umidiceps

15

5

55.9+0.29'^'

9.5±0.24

7.5±0.17

7.9+0.13

13.9+0.27

9.8+0.21

(52.6-58.8)

(9.3-9.8)

(7.2-7.7)

(7.7-8.1)

(13.6-14.3)

(9.5-10.1)

16

4

56.0+1.13

9.7±0.21

7.6+0.25

8.0+0.10

14.2+0.28

10.2+0.13

(54.6-57.3)

(9.5-10.0)

(7.3-7.9)

(7.9-8.1)

(13.9-14.5)

(10.1-10.4)

20

23

56.0+0.45

9.7±0.0ff«

7.7+0.07

7.9+0.07

14.1+0.09^

10.1+0.07

( 53.7-57.3 )

(9.5-10.0)

(7.3-8.0)

(7.7-8.2)

(13.8-14.5)

(9.8-10.4)

21

13

56.4±1.01

9.6±0.10

7.7+0.08'-

7.9+0.08

14.1+0.13

io.i+o.ir=^

(53.0-59.8)

(9.3-9.9)

(7.5-7.9)

(7.8-8.2)

(13.7_14.5)

(9.7-10.4)

Mormoop

s megalophylla carteri

22

21

58.8±0.48'^

10.1±0.08

7.9+0.12

8.3+0.06

14.7+0.10

10.3+0.06

(56.6-60.6)

(9.8-10.3)

(7.0-8.3)

(8.1-8.7)

(14.4-15.1)

(10.0-10.6)

Pillar del Rio: Cueva del Indio, San Vicente,
15 (AMNH); Ctieva del Rio, San Vicente, 1
(ROM); San Vicente, 2 (ASC); Cueva de
William Palmer, Guanajay, 3 (TCWC). Las
Villas: Cueva Grande, Yaguajay, 4 (FMNH);
Cueva de los Masones, Trinidad, 2 ( ASC ) .
JAMAICA. Hanover: Lucea, 1 (AMNH).
Portland: Port Antonio, 1 (USNM). St.
Andrew: Kingston, 1 ( USNM ). St. Ann: Dry
Harbor, 2 (AMNH); Moneague, 1 (USNM).
St. Catherine: St. Clair Cave, 9 (2 LAGM, 5
ROM, 2 USNM). St. Elizabeth: Oxford Cave,
9 (AMNH). HAITI. Diquini, 1 (MCZ).
DOMINICAN REPUBLIC. Altagracia: Cueva
Las Lagunas de Nisibon, 1 (FMNH). San
Juan: Rancho La Guardia, 4 (AMNH).
PUERTO RICO. Mona Island, 8 (USNM);
Mayaguez, 1 (USNM); Los Chorros Gave,
near Arecibo, 37 ( ROM ) ; Cueva Tnijillo Alto,
5 (AMNH).

Additional marginal record. — Exuma Is-
land, Koopman (1951:229).

Mormoops megalophylla Peters
Peters' Ghost-faced Bat

( Synonymy under subspecies )

Distribution . — From southwestern
Texas and southern Arizona, southward
through Mexico (including southern
Baja California, the Mexican Plateau,
and Yucatan Peninsula) into Central
America as far as eastern Honduras and
El Salvador. This species has not yet
been recorded from Nicaragua, Costa

Rica, or Panama. In South America, M.
megalophylla is known from several
widely separated localities along the
Caribbean coast of Colombia, Vene-
zuela, Trinidad, and the Dutch West
Indies, and also is known from the arid
upper Patia Valley in northern Ecuador
(Fig. 39).

Description. — Largest species of the
genus Mormoops, forearm usually more
than 50. Rostrum extremely upturned
and broad; braincase, viewed in profile,
squared and somewhat flattened dor-
sally or pyramidal lacking noticeable
flattening ( Fig. 37 ) ; forehead arises
abruptly from rostrum; basioccipital
and basisphenoid broad and trough-like
between auditory bullae, with a promi-
nent longitudinal, medial septum; ears
rounded and strongly connected by two
high bands that fuse on top of rostrum;
tragus complex, with an extremely prom-
inent secondary fold (Fig. lOG); labio-
nasal plate also complex, each nostril
surrounded by a separate pad; margin
above and between nostrils with sev-
eral wart-like tubercules; long and
prominent ridge separates each nostril
(Fig. IIF); dorsal pelage long and lax,
with four zones of coloration on each
hair.

ii4

MISCELLANEOUS PUBLICATION MUSEUM OF NATURAL HISTORY

50.0 52.0 54 0 56 0 58 0 60.0
I I I I I I I I I I \ 1 —

9.0 5.2 9.4 9.6 9.8 10.0 10.2
I 1 I I I I — I — I r

r^

A

10

Figure 40. Dice-squares illustrating geographic variation of Mormoops megalophijUa from selected

mainland localities (see text for key to localities). The measurements illustrated (in mm) are: A,

length of forearm; B, zygomatic breadth. See figure 26 for explanation of symbols.

Relatiomhips. — See account of M.
hiainvillii.

Geographic variation. — My analysis
indicates that there is a relatively small
amount of geographic variation in North
American populations ( Table 5 and Fig.
40, localities 1-14). There appears to
be a considerable amount of variation
within samples and only a gradual in-
crease in overall size in the extreme

southern (Honduras) part of the range.
Although a rather large number of
specimens of M. megalophyUa were
examined from South America, these
were from relatively few and widely
separated localities. South American
representatives of this species generally
average larger than those from Mexico
and northern Middle America (Table 5
and Fig. 40, localities 17-22). In addi-

SMITH: SYSTEMATICS OF MORMOOPID BATS

115

tion to somewhat larger size, there are
, several other features that seem to char-
acterize all South American populations.
One of these concerns the shape of the
l^raincase and particularly the outline of
the forehead, viewed in profile. In tJie
Mexican representatives of the species,
the forehead rises abruptl)% but flattens
gradually on top of the braincase. This
flattening is not so apparent in the South
American bats (Fig. 37) and, in fact,
the forehead seems to include both
frontal and parietal regions. There are
no other striking cranial differences.

Externally several noteworthy char-
acteristics further distinguish South
American bats from North American
representatives. The color of fresh
pelage in these bats is a deep cinnamon
red with an iridescent, purplish frost
over the back and rump. Ventrally, the
pelage is pinkish or reddish in color.
These colors vary somewhat depending
on geographic location, age of the
pelage, or degree of bleaching. The
most striking feature of the pelage of
the South American representatives of
M. megalophyUa is the development of
a white or pinkish, fan-shaped "cape"
over the shoulders, formed by relatively
long stiff hairs. This cape and the rich
cinnamon coloration are absent in North
American bats of this species. I do not
believe these features represent specific
differences between North and South
American populations, although the ob-
served geographic variation indicates
somewhat greater subspecific differences
than in other polytypic species of the
family and possibly represents incipient
speciation. It seems likely that these
distinct geographic units were fornied
by spatial separation of a formerly wide-
spread population in Pleistocene times,
resulting in the subsequent isolation of
North and South American segments of
the species. In view of the relatively
concentrated collecting efforts in the
past decade or so in southern Middle
America and the failure to obtain M.
77iegalophylla there, the present distri-

butional gap in this region is real. I
assume that any reproductive isolation
that may have developed during spatial
isolation has not yet been tested, and I
therefore elect to regard the South
American populations as a distinctive
subspecies group rather than as a dis-
tinct species.

The general pattern of geographic
variation along the Caribbean coast of
South America is one of gradual in-
crease in overall size to the eastward.
However, in the Dutch West Indies
(localities 17-19), bats are smaller and
paler than those on the adjacent main-
land and are subspecifically distinct;
the extreme of this trend being seen in
specimens from the islands of Curagao
and Bonaire. A large series from the
upper Patia Valley of Ecuador also con-
stitutes a morphologically and geo-
graphically distinct population, which is
herein given subspecific recognition.
Bats from this region seem to be iso-
lated on the arid Pacific versant of the
Andes and are the largest representa-
tives of the species.

Key to localities used in statistical analysis
of geographic variation. — 1, Concon, Uvalde
Co., Texas; 2, northern Coahuila; 3, Tamau-
lipas; 4, Sonora; 5, Mulege, Baja California;
6, southern Sinaloa; 7, vicinity of Amatlan,
Nayarit; 8, vicinity of Ocotlan and Jamay,
Jalisco; 9, Colima; 10, 1 mi E Raboso, Puebla;
11, southwestern Guerrero; 12, Campeche; 13,
Lanquin Cave, Guatemala; 14, vicinity of
Divisadero, El Salvador; 15, Cartegena, Co-
lombia; 16, San Gil, Santander, Colombia; 17,
Aruba; 18, Curasao; 19, Bonaire; 20, \icinity
of Guiria, Venezuela; 21, Trinidad; 22, La Paz,
Ecuador.

Mormoops megalophylla megalophyUa

Peters

Mormops megalophylla Peters, 1864, Monatsb.
preuss. Akad. Wiss., Berlin, p. 381.

Mormoops megalophyUa senicula Rehn, 1902,
Proc. Acad. Nat. Sci. Philadelphia, 54:169,
11 June. Type locality, Fort Clark, Kinney
Co., Texas. Holotype, adult female, skin
and skull, USNM 84801; obtained by Dr.
E. A. Mearns; original no. 4273.

Mormoops megalophylla rufescens Davis and
Carter, 1962, Southwestern Nat., 7:65, 1
June. Type locality, 5 mi W Alamos, So-

116

MISCELLANEOUS PUBLICATION MUSEUM OF NATURAL HISTORY

nora, Mexico. Holotype, adult male, skin
and skull, LACM 9978; obtained by Ken-
neth E. Stager; original no. 1652.

Holotype. — Adult female, preserved
in alcohol, skull removed, Zoologisches
Museum Berlin no. 2826; type locality,
Parras, Coahuila, Mexico (see remarks
below); obtained by Schneider.

Distr i but ion .■ — From southwestern
Texas and southern Arizona, southward
throughout Mexico into Central Amer-
ica as far as eastern Honduras and El
Salvador (Fig. 39).

Description. — Size small, both crani-
ally and externally, for a mainland race
of the species. Standard external meas-
urements (average, extremes in paren-
theses) for 18 males and 13 females, re-
spectively, from 3.5 mi S and 2.5 mi E
Ocotlan, 500 ft, Jalisco, are as follows:
total length, 88.9 (85.0-97.0), 90.9 (87.0-
95.0); length of tail vertebrae, 24.9
(22.0-28.0), 25.1 (23.0-27.0); length of
hind foot, 13.0 (12.0-14.0), 13.0 (13.0);
length of ear from notch, 14.3 (13.0-
15.0), 14.5 (14.0-15.0). See table 5
(localities 1-14), for additional selected
external and cranial measurements.

In general, fresh pelage of this sub-
species is brownish to pale brown
(Prout's Brown to Snuff Brown) dor-
sally. A large series taken on 13 August
near Raboso, Puebla, was composed of
individuals that were in various stages
of molt. Generally, old pelage of these
bats was reddish brown, whereas fresh
pelage was dark brown ( Prout's Brown ) .
The four zones of color of individual
hairs from the back are, from base to
tip, dark brown, pale brown to tannish,
dark brown, and blackish brown. In-
dividuals from Texas and northern Mex-
ico, and from northern Middle America,
tend to be slightly darker than speci-
mens from the intervening areas; how-
ever, a great deal of variation was ob-
served in the samples examined. The
specimens that Davis and Carter ( 1962 )
assigned to M. m. rufescens are reddish
brown. The majority of these were col-
lected at a time of year ( December to

August) when the pelage is usually red-
dish in color. Five adult males, caught
on 28 July near Santa Lucia, Sinaloa,
were actively molting from a reddish
brown pelage to a darker, brownish
gray pelage, which seems to indicate
that the reddish color heretofore thought
to be distinctive of M. ni. rufescem
actually represents seasonal variation.

C o m p a r i s o n s- . — Mormoops megaJo-
phyUa niegalopJu/Ua is easily distin-
guished from South American subspe-
cies by the absence of a fan-shaped
"cape" over the shoulder region that is
formed by relatively long, stiff hairs,
and the somewhat flattened forehead as
viewed from the side (Fig. 37). Except
for the insular race M. m. intermedia,
M. m. meg,aJophyUa is smaller in size,
both cranially and externally than other
subspecies (Table 5).

Remarks. — In the original descrip-
tion, Peters (1864) did not mention a
precise type locality in Mexico. In an
earlier paper, he (1856) supposed these
bats were from Cuba and referred to
them as "M. hlainviUii." Rehn (1902:
167) compared three specimens of M.
megalophylla (two from Merida, Yuca-
tan, and one from San Juan Bautista,
Tabasco) with Peters' (1856: pi. 1, figs.
1-5) illustrations and concluded that the
type locality should be regarded as
southern Mexico. Dilford C. Carter re-
cently examined Peters' holotype of M.
megalopJjylla in Berlin and noted that
the locality "Parras, Mexico," in Peters'
own handwriting, was associated with
this specimen. In view of these circum-
stances, it seems appropriate to regard
the type locality of megalophylla as the
well-known collecting site, Parras, Coa-
huila.

Three subspecific names have been
applied to North American representa-
tives of the species megalophylla. The
name senicula was based primarily on
size and shape of the second upper pre-
molar, a characteristic found to be
variable by Villa and Jimenez ( 1960:
503). This same variation was detected

SMITH: SYSTEMATICS OF MORMOOPID BATS

117

I

by Davis and Carter (1962:65) who fol-
lowed Villa and Jimenez in regarding
senicula as a junior synonym of M. m.
megalophijUa. I agree with these au-
thors in this matter.

Davis and Carter (1962:65-67) re-
viewed the North American representa-
tives of this species and concluded that,
on the basis of color, there were two
geographic groups; the one from north-
western Mexico they named M. m. ru-
fescens. Aside from the alleged reddish
or cinnamon color of the pelage, these
authors further diagnosed rufescens as
having a short forearm (about 54) and
a short, narrow skull. Bats from north-
western Mexico do average slightly
smaller than those from the northeastern
and southern parts of the country, as
noted by Davis and Carter, but the dif-
ferences are so slight that, in the ab-
sence of an effective barrier, subspecific
recognition of this population does not
seem to be warranted. Specimens in a
sample from Mulege, Baja California,
are somewhat smaller than other Mexi-
can specimens, even those from the
adjacent mainland, and additional ma-
terial from this little-studied region may
reveal that subspecific recognition is
warranted for these peninsular bats.

The few specimens examined from
the northern part of the Yucatan Penin-
sula average smaller than those from
Campeche and other nearby regions.
Also, as mentioned previously, bats from
Honduras and El Salvador are slightly
larger than individuals from further
northward. An accurate appraisal of
these trends will not be possible until
the relationships between North Amer-
ican and South American races are bet-
ter understood.

Specimens examined (787). — TEXAS.
Brewster Countij: Oak Creek, Chisos Mts.,
5000 ft, 2 (TCWC). Hidalgo County: Edin-
burg, 1 (TCWC). Kinnetj County: Fort
Clark, 1 (USNM). Medina County: Valdina
Farms sink hole, 3 (LACM); Nye Cave, 1
(LACM). Uvalde County: Concon, 11 (3
AMNH, 2 CAS, 6 USNM); Fria Cave, Con-
con, 36 (LACM); 9 mi W Sabinal, 2 (USNM).
MEXICO. Baja California: 0.25 mi S Mulege,

100 ft, 14 (MVZ); 3 km S San Antonio, 1
(KU). Campeche: 1 km N, 13 km W Escar-
cega, 65 m, 2 (KU); 12 km W Escdrcega, 4
(KU); 7 km N, 51 km E Escdrcega, 1 (KU);
IH km S, 65 km E Escdrcega, 6 (KU); La
Tuxpena, 1 (USNM); 105 km E Escarcega, 1
(KU). Chiapas: 5 km S Palenque, 1 (ENCB);
3 mi E Cintalapa, 1700 ft, 1 (TCWC); 4.5
mi W San Cristobal, 4 (UMMZ); La Soledad,
approx. 16 mi NE Las Margaritas, 3600 ft, 1
(LACM); 5 mi N Arriaga, 800 ft, 2 (TCWC);
Zapaliita, 5700 ft, 1 (TCWC); 1 mi S Zapa-
luta, 5700 ft, 1 (TCWC); 2 mi S Zapaluta,
5700 ft, 1 (TCWC); 18 mi S La Trinitaria,
2800 ft, 3 (TCWC); Cueva Leon, 4 km ENE
San Lucas, 1 (TCWC). Chihuahua: Cari-
mechi, Rio Mayo, 3 (TCWC). Coahuila:
Fortin, Ranclio Las Margaritas, 3300 ft, 1
(KU); Las Margaritas, 5 (KU); Puerta San-
tana, 6 mi S, 3 mi W Hacienda La Mariposa,
2900 ft, 10 (KU); 6 mi E Hermanas, 1200 ft,
3 (KU); Cueva del Buen Abrigo, Noria no. 5
E Coyote, Torreon, 12 (IB). Colima: 7 mi
W, 0.5 mi S Santiago, sea level, 20 (KU);
Playa de Oro, 8 km W, 2 km S Santiago, 1
(KU); 8 mi SE Colima, 20 (LACM). Dn-
rango: La Cueva de Espaiia, 2 km S, 11 km
W Nazareno, 400 ft, 5 (TCWC); 2 mi N
Pueblo Nuevo 6000 ft, 1 (MSU). Guanaguato:
No .specific locality, 1 (USNM). Guerrero:
Grutas de Cacahuamilpa, 27 (24 ENCB, 3
KU); Agua de Obispo, 37 km S Chilpancingo,
980 m, 2 (KU); 5 mi E Omilteme, 6200 ft, 1
(USNM); Grutas de Juxlahuaca, 4 km NE
Colotlipa, 1100 m, 12 (11 ENCB, 1 IB); 2 mi
NW Acapulco, 50 ft, 15 (KU). Jahsco: 15
km W Ameca, 4200 ft, 2 ( KU ) ; Los Masos, 8
(6 AMNH, 2 FMNH); 18 km NW Purifica-
cion, 8 (KU); 16 mi NE Tamazula, 5000 ft,
2 (TCWC); 5 mi W Chapala, 5000 ft, 1
(KU); 3.5 mi S, 2.5 mi E Ocotlan, 5000 ft, 45
(41 KU, 4 MSU); 0.5 km N, 3 km W Jamay,
1650 m, 9 (IB). Morelos: Cueva del Salitre,
10 km NNW Tecuman, 36 (IB). Nayarit:
San Bias, 7 (MVZ); 0.5 mi E San Bias, 10 ft,
27 (KU); Rancho Palo, near Amatlan, 36 (35
AMNH, 1 FMNH). Nuevo Leon: Mina Jesus
Maria, 10 km SSE Vallecito, 12 (IB). Oaxaca:
5 mi W Cliiltepec, 2 (AMNH); 8 km NW
Salina Cruz, .3 ( ENCB ) ; Tapanatepec, 2
(AMNH). Puehla: 1 mi E Raboso, 4350 ft,
54 (KU). Sinaloa: Mazatlan, 2 (USNM);
Panuco, 22 km NE Concordia, 25 (KU); 1 mi
E Santa Lucia, 3650 ft, 1 (KU); 1.5 km S
Santa Lucia, 3200 ft, 5 (KU). Sonora: Cueva
del Tigre, 14.9 mi SE Carbo, 7 (LACM); 20
mi SE Carbo, 2700 ft, 1 (LACM); Rancho
San Jose, 15 mi NW Guaymas, 2 (KU); La
Aduana, 1 (LACM); 4 mi N Alamos, 1500 ft,
2 (KU); 5 mi W Alamo, 6 (5 LACM, 1
AMNH). Tabasco: 5 mi SW Teapa, 1 (KU);
Arroyo del Solpho, Tapiajulnpa, 2 (AMNH);
San Juan Bautista, 1 (USNM). Tamaulipas:

118

MISCELLANEOUS PUBLICATION MUSEUM OF NATURAL HISTORY

Rancho Santa Rosa, 25 km N, 13 km W Ciu-
dad Victoria, 260 m, 1 (KU); Sierra de Ta-
maulipas, 16 mi W, 16 mi S Piedras, 1400 ft,
4 (KU); Cueva del Pachon, 3 (USNM);
Tampico, 7 (USNM). Veracruz: Mirador, 27
(USNM); 4 km WNW Fortin, 3200 ft, 1
(KU); Orizaba, 5 (USNM); 6 km WSW Boca
del Rio, 10 ft, 1 (KU); 3 km E San Andres
Tuxtla, 1000 ft, 4 (KU). Yucatan: 10 mi W
Progresso, 1 (LACM); Merida, 2 (USNM);
Griita Balancanche, 5 km E Chiclien Itza, 1
(KU). GUATEMALA. Alta Verapaz: 3 km
ENE San Cristobal Verapaz, 1370 m, 3
(TCWC); Lanquin Cave, 1022 ft, 60 (KU);
Piedra de Jalali, 4 km NE Cahabon, 1
(AMNH). Chiquimula: 20 km SSE Chiqui-
mula, 550 m, 1 (TCWC); Jocotdn, near Chi-
quimula, 1350 ft, 1 (KU). Huehuetenango:
San Pedro Soloma, 2270 m, 1 (TCWC). Pe-
ten: 2 mi S Flores, 32 (AMNH); Laguna de
Zotz, 1 (USNM). HONDURAS. Cowayagua:
Comayagua, 580 m, 1 (TCWC). Olancho:
San Jose, Rio Tinto, 340 m, 5 (TCWC).
Santa Barbara: 2 km S San Nicolas, 660 m,
13 (TCWC). EL SALVADOR. Cuscatldn:
2 km W Snchitoto, 390 m, 1 (TCWC). La
Union: Tabanco, 650 ft, 4 (MVZ). Morazdn:
Encuentros Mine, 3 mi W Divisadero, 700 ft,
51 (MVZ); San ]ose Mine, approx. 1 mi W
Divisadero, 1 (MVZ).

Mormoops megalophylla intermedia
Miller

Mormoops intermedia Miller, 1900, Proc. Biol.

Soc. Washington, 13:160, 31 October.
Mormoops megalophylla intermedia Rehn,

1902, Proc. Acad. Nat. Sci. Philadelphia,

54:170, 11 June.

Holotype. — Adult female, preserved
in alcohol, skull not removed, USNM
102174; from cave at Hatto, north coast
of Cura9ao, Dutch West Indies, ob-
tained on 29 April 1900; collector un-
known.

Distribution. — Dutch West Indies
(Fig. 39).

Description. — Smallest of the South
American subspecies of the species, both
cranially and externally (Table 5).
Standard external measurements (aver-
age, with extremes in parentheses) for
four males from Curagao are as follows:
total length, 78.0 (74.0-83.0); length of
tail vertebrae, 19.5 (18.0-21.0); length
of hind foot, 10.5 (9.0-12.0). Additional
selected external and cranial measure-
ments are given in table 5 (localities

17-19). In addition to overall small
cranial size, the rostrum is short, the
upper teeth are noticeably crowded, and
the second lower premolar is extremely
small as compared to the larger first and
third lower premolars.

Nearly all of the specimens of this
subspecies examined were preserved in
alcohol and most were badly bleached.
In general, tlie color of the pelage was
pale brown. The hairs that form the
"cape" over the shoulders are white.
The cape is not particularly prominent
on the holotype, but is prominent on
other individuals that I examined.

Comparisons. — From M. in. megalo-
phyUa of the North American mainland,
M. m. intermedia is distinguished by the
presence of the fan-shaped shoulder
cape and more steeply rising braincase
(Fig. 37). From M. m. ttimidiceps of
the adjacent mainland, M. m. intermedia
may be recognized by its smaller ex-
ternal and cranial size, as well as shorter
rostrum, more crowded upper teeth, and
relatively small second lower premolar.
In both North American and South
American mainland populations, the
second lower premolars are not reduced,
compared to the first and third, as it is
in M. m. intermedia.

Remarks. — There is a noticeable
trend to smaller overall size from Aruba
to Bonaire in this insular subspecies. In
general, bats from the continental island
of Aruba are intermediate in size be-
tween those of the adjacent mainland
and the islands of Curasao and Bonaire
(Table 5). As has been mentioned pre-
viously, suboptimal ecological factors as
well as the small landmass of these
islands probably have been important
features in the differentiation of this
insular race.

Specimens examined (110). — ARLTBA.
Quadirikiri Caves, 4 (AMNH); Lago Oil
Transport Co., 5 (AMNH); 5 mi E Seroe
Colorado, 14 (AMNH); Ex-cave, 4 (USNM).
CURACAO. Round Cliff, 4 (AMNH); North-
west end of island, 7 (USNM); no specific
locality, 70 (2 AMNH, 50 FMNH, 18 USNM).
BONAIRE. Oeroesjan Blanco, Grot Colombia,
near Barcadera, 2 (RMNH).

SMITH: SYSTEMATICS OF MORMOOPID BATS

119

Mormoops megalophylla tumidiceps

Miller

Mormoops tumidiceps Miller, 1902, Proc. Acad.
Nat. Sci. Philadelphia, 54:403, 12 Septem-
ber.

Mormoops megalopliylla tumidiceps — Goodwin
and Greenhall, 1961, Bull. Amer. Mus.
Nat. Hist., 122:225, 26 June.

Holotype. — Adult male, preserved in
alcohol, skull removed, USNM 186443;
from Point Gourde Caves, Trinidad, ob-
tained in 1889 by Sir William Robinson.

Distribution. — Northern South Amer-
ica, primarily along Caribbean coast,
but known also from several inland
localities (Fig. 39).

Description. — Size moderately large,
externally and cranially, among South
American subspecies. Standard external
measurements (average, with extremes
in parentheses) for 18 males and three
females, respectively, from Rio Salado,
9 km N and 4 km E Guiria, Sucre, Vene-
zuela, are as follows: total length, 95.8
(88.0-104.0), 97.0 (92.0-100.0); length
of tail vertebrae, 27.7 (24.0-31.0), 26.0
(24.0-29.0); length of hind foot, 11.9
(10.0-13.0), 11.7 (11.0-12.0); length of
ear from notch, 14.8 (13.0-16.0), 14.7
(14.0-15.0). See table 5 (localities 15-
21) for additional selected external and
cranial dimensions.

Specimens in fresh pelage are a deep
cinnamon red dorsally, with an irides-
cent purplish frost. Hairs that form the
cape are pale pink to rose pink in color.
Ventrally, the pelage is pale pinkish to
reddish.

Comparisons. — For comparisons with
other South American subspecies, see
the account of M. in. intermedia and the
following account of a new subspecies
from Peru.

Remarks. — Generally, overall varia-
tion in this subspecies is relatively small
and similar in degree to that observed in
M. m. megalophijUa. Bats from Trini-
dad average slightly smaller in size than
specimens from the vicinity of Guiria,
Venezuela, on the Peninsula of Paria.

Specimens examined (106). — COLOMBIA.
Bolivar: Cartagena, 12 (2 ROM, 10 USNM).

Guajira: Nazaret, 2 (ROM). Magdalena:
Valledupar, Cesar, 2 (ROM). Santander: San
Gil, 6 (4 FMNH, 2 ROM); Barichara, Maca-
regua, 1700 m, 11 (FMNH). Sucre: Tolu
Viejo, 1 (ROM). VENEZUELA. Bolivar:
Hato San Jose, 146 km S, 7 km E Ciudad
Bolivar, 306 m, 1 (USNM). Nueva Esparta:
Cueva Honda del Piache, SE of El Valle, 2
(RMNH). Sucre: Rio Salado, 9 km N, 4 km
E Guiria, 90 m, 25 (USNM); Ensenada Cau-
ranta, 7 km N, 5 km E Guiria, 1 m, 10
(USNM). TRINIDAD. Port of Spain, 2
(AMNH); Point Gourde Caves, 1 (USNM);
St. Joseph, 2 (AMNH); Tunapuna, 1
(AMNH); Mt. Tamana Caves, 24 (4 AMNH,
15 FMNH, 2 KU, 1 TCWC, 2 ROM); Las
Cuevas, 4 (LACM).

Mormoops megalophylla carteri new

subspecies

Holoti/pe. — Adult male, skin and
skull, TCWC 11643; from Gruta Rumi-
chaca, 2 mi E La Paz, 8700 ft, Carchi
Province, Ecuador; obtained by Dilford
C. Carter on 5 July 1964, original no.
5142.

Distribution. — Known only from the
type locality, but probably occurs along
the arid Pacific coast to northern Peru
and possibly in the dry intermontane
Cauca Valley of the Colombian Andes
(Fig. 36).

Description. — Size largest, externally
and cranially, of the geographic races of
M. megalophylla (Table 5, locality 22).
Standard external measurements of the
holotype followed by the average, with
extremes in parentheses, of 16 males
and 18 females, respectively, from the
type locality are as folows: total length,
94.0, 98.1 (93.0-103.0), 98.6 (95.0-
102.0); length of tail vertebrae, 25.0,
28.3 (23.0-31.0), 27.5 (25.0-30.0); length
of hind foot, 11.0, 10.8 (10.0-13.0), 9.9
(9.0-11.0); length of ear from notch,
16.0, 15.2 (14.0-16.0), 15.4 (15.0-16.0).
See table 5 for additional selected ex-
ternal and cranial measurements of type
series. Selected external and cranial
measurements of the holotype are as
follows: length of forearm, 58.7; zygo-
matic breadth, 10.0; breadth of brain-
case, 9.3; mastoid breadth, 8.9; zygo-
rostral length, 12.0; interorbital breadth,

120

MISCELLANEOUS PUBLICATION MUSEUM OF NATURAL HISTORY

5.5; rostral breadth, 8.1; length of maxil-
lary toothrow, 8.2; length of mandibular
toothrow, 9.2; condylobasal length, 14.8;
depth of cranium, 10.2.

The pelage of the holotype is dark
cinnamon brown dorsally, paler ventral-
ly. Five other males from the type
locality are olive-brown in color.

Comparisoivs. — From M. m. ttimidi-
ceps, this new subspecies may be distin-
guished by its larger cranial and ex-
ternal size — forearm averaging 58.0 or
more and condylobasal length measur-
ing 14.5 or more.

Remarks. — This population appears
to represent another case of geographic
isolation along the arid Pacific coast of
South America. These bats probably
reached the region of northwestern
Ecuador through southward dispersal
along the arid intermontane Cauca Val-
ley of Colombia and northern Ecuador.
In the upper part of this valley there is
a relatively low pass over which bats
could have gained access to the upper
Patia Valley and thence the arid Pacific

versant of Ecuador. Tliis route seems to
be the most reasonable, although it is
conceivable that they also may have
arrived in northwestern Peru by follow-
ing the narrow corridor formed by the
Pacific versant of the Andes in western
Colombia and Ecuador. According to
Haffer (1967a) and Chapman (1926),
this region may have been dominated
by xeric conditions during pre-Pleisto-
cene times. However, the subsequent
development of tropical forest in this
region in the Pleistocene may have
blocked this route of dispersal for non-
forest species.

It is with great pleasure that I name
this subspecies in honor of Dilford C.
Carter, who not only made these and
many other specimens available to me,
but also has been most helpful in pro-
viding valuable information concerning
liolotypes housed in various European
museums.

Specimens exainined (47). — ECUADOR.
Carchi: Gnita Rumichaca, 2 mi E La Paz,
8700 ft, 47 (46 TCWC, 1 KU).

ZOOGEOGRAPHY AND SPECIATION OF MORMOOFIDS

Primarily because of their ability to
fly, bats have been excluded from many
discussions concerning mammalian zoo-
geography, a paradox inasmuch as birds
have been widely used in zoogeographic
studies. Some features that relate to the
evolution and speciation of the Mormo-
opidae are discussed below.

Greater Antilles

This region has been important in
the evolution and differentiation of sev-
eral mormoopid species. Of these, three
(Pteronotus fuliginosus, P. macleayii and
Mormoops blainvillii) are endemic, pre-
sumably having evolved as a result of
isolation in the Antilles. Insular popu-
lations of another species (Pteronotus
parnellii) have diverged from their
mainland relatives to a greater or lesser
degree at the subspecific level. All the
species excepting P. macleayii have been

recorded from all four islands of the
Greater Antillean complex. None is
known from any of the Lesser Antillean
islands, although a mandible of M.
hlainviUii was reported from a pre-
Colombian cave deposit on Exuma Is-
land in the Bahamas (Koopman, 1951:
229).

The concept that bats could traverse
wide sea barriers and thereby populate
insular areas was popular in the early
part of the Twentieth Century, as was
the concept of land bridges to account
for the West Indian dispersal of some
land mammals (Simpson, 1956). G. M.
Allen (1911:176) commented on the
supposed flying capabilities of bats:
"Indeed, the very fact that where bats
are found in islands they have usually
become more or less differentiated from
their nearest neighbors, and this in a
uniform and constant manner, is proof

SMITH: SYSTEMATICS OF MORMOOPID BATS

121

that such fortuitous methods of distribu-
tion as have been claimed for these ani-
mals are largely inoperative." It is now
conceded that bats ordinarily are capa-
ble of crossing only narrow water gaps
without a great deal of difficulty and
that they generally are restricted in the
same way as other mammals by wide
water barriers (Koopman, 1958, 1970).

Climatic and geologic events of late
Tertiary and Quaternary times probably
have been of paramount importance in
the dispersal and evolution of bats, both
in the Antilles and on the Neotropical
mainland. Appreciable fluctuations in
sea level caused a restriction or expan-
sion of the landmass in these areas. In
the case of the Greater Antilles, a drop
in the water level would have resulted
in a considerably narrower water gap
than now exists between Jamaica and
the Honduras-Nicaragua region of Mid-
dle America. In view of the absence of
representatives of Greater Antillean
mormoopids in the Lesser Antilles and
the lack of evidence to support a route
of dispersal for mormoopids from Flor-
ida to Cuba, I presume that mormo-
opids as well as some other groups of
bats arrived in Jamaica by crossing from
the Honduras-Nicaragua region, or pos-
sibly to Cuba from Yucatan.

Pteronotiis fuliginosiis and iruicleoijii
are closely related and both appear to
be related to the inainland species per-
sonatus. In both Antillean species, Cu-
ban individuals are markedly smaller
than those from Jamaica and are herein
referred to distinct subspecies. The de-
velopment of racial differences between
Cuban and Jamaican populations points
to the relative effectiveness of the water
barrier that separates these islands. On
the other hand, representatives of F.
fuliginosus from Jamaica, Hispaniola,
and Puerto Rico exhibit only minor dif-
ferences in what appears to be a con-
tinuum of geographic variation, evident-
ly reflecting a close geographic relation-
ship in the past among the three islands
in question. Both fuliginosus and niac-

leayii could have evolved in isolation on
the Antilles from a single invasion of
personatusAike bats, with subsequent
sympatry, or they could be the result of
two invasions, with speciation of the
first immigrant stock completed prior to
arrival of the second. In view of the
greater similarities between these two
species than between either of them and
persotmtus, it is most likely that they
resulted from a single invasion.

In Mormoops hlainvillii, a somewhat
larger bat (and presumably a better
flier) than either fuliginosus or inaclea-
ijii, there is relatively little geographic
variation. There is, however, a slight
indication of a pattern in this species
that suggests separation of bats on
Cuba from those on Hispaniola and
Puerto Rico, with bats from Jamaica
providing the intermediate link.

The degree to which the three en-
demic Antillean species have diverged
morphologically and evolutionarily from
their mainland relatives is considerably
greater than that observed in Pteronotus
parnellii, evidently a more recent (prob-
ably Pleistocene) immigrant to the
Greater Antilles. The association of a
dated time sequence with the invasion
of the Antilles by mormoopids (and
other chiropterans ) is impossible in the
absence of a fossil record. By general
inference, time required for morpho-
logical (and genetic) differentiation de-
creases with taxonomic rank. Presum-
ably, therefore, the time required for
the evolution of the endemic Antillean
phyllostomatid subfamily Phyllonycteri-
nae was considerably longer than that
required to evolve most or all of the
endemic Antillean stenodermine genera,
the differentiation of which may well
have preceded that at the specific level
seen in mormoopids. Bats that have
only recently invaded the West Indies
would be expected to exhibit relatively
few morphological differences from pa-
rental mainland populations. This ap-
pears to be the case in Pteronotus davyi,
the only mormoopid that has success-

122

MISCELLANEOUS PUBLICATION MUSEUM OF NATURAL HISTORY

fully invaded the Lesser Antilles (Fig.
31). However, such linear inferences
should be clearly regarded as specula-
tive owing to our present lack of in-
formation concerning modes of chirop-
teran evolution. At this stage, a reason-
able estimate of the time of arrival of
the three mormoopid species that are
endemic in the Antilles might be middle
to late Pliocene.

Pteronotus parnelUi evidently arrived
later in the Greater Antilles than did
the three endemic species. Pteronotus
p. parnellii (from Cuba and Jamaica)
and P. p. portoricensis (Puerto Rico)
clearly are only subspecifically distinct
from mainland representatives of the
species, and the differences between
these two insular races are small. Had
they been located on adjacent islands, I
would not have retained portoricensis
as a distinct subspecies. However, in
view of the interposed distribution of
the markedly distinct subspecies P. p.
pusiUus and P. p. gotmvensis (from His-
paniola and La Gonave Island, respec-
tively), I regard Puerto Rican bats as a
distinct subspecies.

There are several possible explana-
tions for the unique distribution of An-
tillean subspecies of parnelUi. Perhaps
they represent independent insular evo-
lution from a single immigrant stock.
Another possibility would be to assume
two invasions of the region by parnelUi,
the first represented by pusiUus (and
gonavensis) and the second by the
larger (and evidently related) parneUii
and portoricensis. It is uncertain wheth-
er there is gene flow between Hispan-
iolan pusiUus and the populations on
Cuba, Jamaica, and Puerto Rico. If not,
then serious consideration should be
given to elevating pusiUus (with gona-
vensis as a distinct subspecies) to full
specific rank, and placing portoricensis
in synonymy under P. p. parneUii.

Neotropical Mainland

The geographic range of mainland
mormoopids extends from the northern

limits of the Neotropics in North Amer-
ica (Mormoops megalophyUa even ex-
tends slightly into the temperate regions
of that continent) to near the southern
limits of the Neotropics ( Mato Grosso,
Brasil) in South America. Generally the
ranges of mainland species overlap
broadly, with the exception of Pterono-
tus suapurensis, which has a somewhat
restricted geographic range from south-
ern Mexico to as far south as Tingo
Maria, Peru, and possibly the Mato
Grosso region of Brasil. The center of
origin from which the family dispersed
is obscured by the absence of a fossil
record. Origin of the group, however,
could have been in southern Middle
America or northern South America,
areas that fall approximately in the mid-
dle of the present distribution of the
mainland species of the family.

From the analyses of geographic
variation, there is an indication that
populations of Pteronotus parneUii, per-
sonatus, clavyi, and Mormoops megalo-
phyUa have been split into northern and
southern segments sometime in the past,
leading to subspeciation. In Mormoops,
there still remains a sizable gap be-
tween the two segments, which may in
fact signal incipient species; without
conclusive evidence, however, I treat
the two as subspecies groups. The few
specimens of P. personatus and davyi
from Middle America suggest that the
ranges of both are continuous, or nearly
so, from North America into South
America; certainly the range of P. par-
nelUi is continuous.

In North America, the Mexican Pla-
teau seems to be an important physical
and ecological barrier to dispersal of
members of the genus Pteronotus. The
ranges of P. parnellii, personatus, and
(lavyi extend northward along either
side of the plateau into the lowlands of
eastern and western Mexico to the ex-
treme northern limits of the Neotropics
in the southern parts of Sonora and
Tamaulipas. The tropical environment
in western Mexico (northern Sinaloa

SMITH: SYSTEMATICS OF MORMOOPID BATS

123

and Sonora ) is more arid than in eastern
Mexico. Specimens of P. daviji ftilvus
from Sonora are noticeably larger and
exhibit more secondary sexual variation
than do bats of most other Mexican
populations of this subspecies. By com-
parison, specimens of P. parnellii from
western Mexico average smaller than
bats of other mainland populations and
exhibit little secondary sexual variation.

The Mexican Plateau has not been a
barrier to the dispersal of Mormoops
megalophylla, which is known from a
number of localities on the plateau ( Fig.
39). This species has been reported
from as far north as southern Texas (in-
cluding parts of the Edwards Plateau)
and southern Arizona, and it is the only
mormoopid presently known from Baja
California. The exact nature of limiting
factors to the distribution of this species
are not yet well understood. Generally
speaking, individuals of M. megalophyl-
la are relatively small in Mexico and
northern Middle America, although there
is considerable variation between popu-
lations. In Honduras and El Salvador,
members of this species are slightly
larger than bats from more northern
localities.

The Isthmus of Tehuantepec, a low-
land area in southern Mexico over
which a water barrier prevailed during
periods of high sea level in the past,
apparently has been of little conse-
quence to the dispersal and speciation
of mormoopid species excepting P. par-
nellii. Pteronotus p. rnexicamis, a small
and pale subspecies, apparently differ-
entiated to the north of the isthmus,
whereas the comparatively larger and
darker P. p. mesoamericanus evidently
differentiated in nuclear Middle Amer-
ica. Apparently gene flow between
these two geographic units is less pro-
nounced along the dry southern part of
the isthmus than along the humid north-
ern side. It is also interesting to note
that the northernmost known occurrence
of P. suapurensis, perhaps the most
mesophyllic of the mormoopid species,

is from San Andres Tuxtla, Veracruz,
which is situated about half way across
the humid northern part of the isthmus.

Davis (1968, 1969) suggested that
the ecological and physiographic condi-
tions of the Caribbean and Pacific ver-
sants of Middle America have been im-
portant factors in the differentiation of
subspecies in Urodernui hilobatum and
Artiheiis phaeofis. Unfortunately, there
is a scarcity of material of most mormo-
opid species from critical areas in Mid-
dle America and an accurate appraisal
of the effectiveness of these versants
with reference to this group is not yet
feasible. They do seem to have affected
subspeciation in at least one species, P.
parnellii. Populations assigned herein to
P. p. ruhiginosus and P. p. mesoameri-
canus are segregated along the humid
Caribbean versant and semi-arid Pacific
versant, respectively. These two sub-
species are most distinct from each other
in northern Nicaragua, probably owing
to the rather broad ecotonal belt be-
tween the two versants in that area.
Farther to the south, the versants come
into closer proximity and the ecotonal
belt is correspondingly narrower. In this
region, ruhiginosus-Vike bats dominate
the zone of intergradation.

During the Pleistocene, the region of
eastern Panama and northwestern Co-
lombia may have been of considerable
importance to dispersal, both southward
and northward, of P. parnellii as well as
other mormoopids. Climatic fluctua-
tions, which no doubt greatly influenced
the extent of humid tropical forest
throughout the Neotropics, probably re-
sulted in changes in the character of the
habitat in this narrow corridor. There
likely was continuous arid habitat be-
tween the Caribbean coast of South
America and Pacific versant of Middle
America in dry epochs, whereas wet
periods would have been conducive to
continuous humid forest from the Ama-
zon Basin through the Caribbean ver-
sant of Middle America.

In South America, considerably less

124

MISCELLANEOUS PUBLICATION MUSEUM OF NATURAL HISTORY

is known about the distribution of mor-
moopids and the features that influ-
enced these patterns appear to be much
more complex than in North America.
As noted above, all five mainland spe-
cies occur in South America. With the
exception of P. suapurensis, all are rep-
resented there by populations that are
subspecifically distinct from those of
North America (although some, such as
Pteronotiis parnelUi ruhiginosus, Ptero-
notus persoiuitiis personatus, and Ptero-
notus daviji davyi extend northward into
southern Central America). In three
species (P. parneUii, P. davyi, and M.
megalophyJlo) there are at least two rec-
ognizable South American subspecies.

Haffer (1967a, 1967b, 1969, 1970)
examined the "nonforest" and "forest"
bird faunas of northwestern South
America and the Amazon Basin and has
presented some interesting ideas con-
cerning Pleistocene and post-Pleistocene
zoogeography of that region. His argu-
ments primarily revolve around forest
expansion and contraction as a result of
wet glacial periods and dry interglacial
periods, and the corridors of dispersal
that opened or closed with these fluc-
tuations. Haffer's concepts can be ap-
plied easily to the bat fauna of this
region.

During forest expansion in the Ama-
zonian region, northward and south-
ward dispersal of semi-xerophyllic spe-
cies along the eastern slope of the Andes
would have been effectively blocked,
whereas in drier periods an arid corri-
dor would have extended from northern
Colombia and Venezuela along the east-
ern slope of the Andes to southern
Brasil. Perhaps such a corridor per-
mitted Pteronotus davyi to enter the Rio
Maraiion Valley through which these
bats spread and eventually crossed the
Andes in the upper Huancabamba Val-
ley to northwestern Peru (where the
subspecies P. d. incae later differen-
tiated).

Perhaps the most interesting area in
South America is the region in north-

western Colombia through which north
and south immigrants passed enroute to
(or from) South America. This region
is complicated by the position of the
Colombian Andes and the dry inter-
montane Cauca and Magdalena valleys.
Based on the presently known distribu-
tional patterns, the general pattern of
mormoopid dispersal through this region
apparently has followed along the Ca-
ribbean coast of Colombia and Vene-
zuela. Pteronotus parneUii, suapurensis,
and possibly davyi likely made their
way to the Amazon Basin either over a
depression in the Andes in western
Venezuela or conceivably through the
Delta Amacuro in eastern Venezuela.

Some populations of Mornioops
rnegalophylla evidently migrated up the
dry intermontane Cauca Valley and
eventually into the upper Patia Valley
on the dry Pacific versant of Ecuador,
after having crossed the relatively low
Andean pass that separates the two val-
leys. Subsequent isolation apparently
resulted in the establishment of the dis-
tinctive subspecies M. m. carteri in this
region. Both M. m. carteri and P. d.
incae (discussed above) conceivably
could have arrived in the arid Pacific
region of Ecuador and Peru by dispers-
ing along the narrow Pacific lowland
corridor of western Colombia and Ecua-
dor. Much of this region, however, is
covered by humid tropical forest that
probably developed in the early part of
the Pleistocene according to Haffer
(1967a) and Chapman (1926). This
forest presumably would have blocked
these two species because neither seems
to be a deep forest inhabitant and it is
doubtful that either arrived in western
Ecuador and Peru before the Pleisto-
cene, judging by the degree of morpho-
logical differentiation from other South
American populations.

Mormoops megalophyUa apparently
is the only mormoopid that has success-
fully colonized the Dutch West Indies.
Analysis of geographic variation indi-

SMITH: SYSTEMATICS OF MORMOOPID BATS

125

cates that these insular populations have
indeed differentiated subspecifically
from those on the mainland, as was pre-
viously supposed. A eertain amount of
intergradation appears to be maintained
through the continental island of Aruba,
which is situated between the mainland
and the islands of Cura9ao and Bonaire.

It is interesting to note that P. daviji (a
species known from the adjacent main-
land and reported from as far north in
the Lesser Antilles as Marie Galante) is
apparently absent from the Dutch West
Indian fauna, emphasizing the impor-
tance of niche availability as well as the
fortuitous nature of insular invasion.

SUMMARY

The taxonomic history of the family
Mormoopidae Saussure, 1860 is re-
viewed beginning with the description
of Monnoops and Aello by Leach
( 1821 ) . Most of the remaining taxa
were discovered and described during
the last half of the nineteenth century
and have been allied with the fami-
lies Noctilionidae and Phyllostomatidae.
The correct derivation and authorship
of the family-group name is docu-
mented.

Prior to this study, the mormoopid
genera were arranged together in the
subfamily Chilonycterinae (family Phyl-
lostomatidae). This arrangement was
tested utilizing evidence derived from
a comparative examination of the oste-
logical structure of distal and proximal
ends of the humerus, proximal end of
the femur, and structure of the tragus.
These data are correlated with pub-
lished results of other comparative
studies involving acoustic orientation,
hair structure, karyotypes, histological
aspects of facial glands, ectoparasites,
brain size, and immunologic studies. It
is concluded that mormoopids comprise
a family distinct from the Phyllosto-
matidae, both morphologically and be-
havioral ly.

Owing to the absence of a substan-
tial fossil record for bats, the phyloge-
netic relationships of the Mormoopidae
are not readily apparent. Features of
the external and internal anatomy seem
to indicate that mormoopids likely are
most closely related to the Noctilionidae
and Phyllostomatidae, perhaps more
closely allied with the former than the

latter. It is postulated that all three
of these families may have arisen as
autochthonous lineages from a paleo-
chiropteran stock present in the New
World as early as middle to late Paleo-
cene times.

Two genera, Pteronotus and Mor-
moops, are recognized; the former is
divided into three subgenera, Phyllodia,
Chilonycteris, and Pteronotus. Members
of the genus Mormoops are more spe-
cialized than those of the genus Ptero-
notus; the dichotomy is marked by a
number of unique features. Within
Pteronotus, the subgenus PhijUodia is
more distinct from Chilonycteris and
Pteronotus than are the latter subgenera
from each other.

Patterns of non-geographic variation
(secondary sexual variation, individual
variation, and seasonal variation with
reference to pelage and molt) are dis-
cussed, and a key to the genera and
subgenera is presented.

Patterns of geographic variation of
each mormoopid species was statistically
analyzed and correlated with the eco-
logical and physiographic features of
the regions inhabited by these species.
Two species were found to be mono-
typic, whereas six are polytypic involv-
ing 21 subspecies, three of which (Ptero-
notus parnellii mesoamericanus, Pterono-
tus davyi incae, and Mormoops megalo-
phyUa carteri) are described as new.

The zoogeography of the family is
obscured by the absence of a fossil rec-
ord and extensive, broadly sympatric
geographic distributions in the Neo-
tropics. The center of origin of mormo-

126

MISCELLANEOUS PUBLICATION MUSEUM OF NATURAL HISTORY

opids could have been in southern Mid-
dle America or northwestern South
America. Representatives of both gen-
era have successfully invaded the
Greater Antilles, presumably from the
Middle American region, and have dif-

ferentiated as full species or subspecies
from their mainland relatives. Zoogeo-
graphic features that may have governed
the dispersal and speciation, both on
the Greater Antilles and the Neotropical
mainland, are discussed.

RESUMEN

Se traza la historia taxonomica de la
familia Mormoopidae Saussure, 1860,
partiendo de la descripcion de Mormo-
ops y Aello dada por Leach ( 1821 ) . La
mayoria de las especies de esta familia
fue descubierta y descrita durante la
segunda mitad del siglo 19 y ha sido
intercolada entre las familias Noctilioni-
dae y Phyllostomatidae. La derivacion
y la paternidad de autor del nombre de
la familia son documentadas.

Anterior a esta investigacion fueron
colocados los generos mormoopidos
juntos en la subfamilia Ghilonycterinae
(Familia: Phyllostomatidae). Esta co-
locacion fue comprobada utilizando la
evidencia obtenida de una examinacion
comparada de la estructura osteologica
de las partes proximas y distales del
humero, de la parte proxima del femur
y la estructura del trago. Estos datos se
correlacionen con resultados publicados
en otros estudios comparados de ecolo-
cacion, la estructura del pelo, cromo-
somas, aspectos histologicos de gland-
ulas faciales, ectoparasitos, tamaiio del
cerebro y estudios inmunologicos. De
ahi que los mormoopidos forman una
familia distinta y separada de la Phyllo-
stomatidae, tanto en cuanto a la mor-
fologia como en cuanto a la comporta-
miento.

Las relaciones filogeneticas de la
Mormoopidae son borrosas debido a la
ausencia de una buena historia fosilifera
de los murcielagos. Segiin varias carac-
teristicas de la morfologia, parece que
los mormoopidos tienen parentesco con
las familias Noctilionidae y Phyllosto-
matidae, quiza mas con la primera que
con la segunda. Se sugiere que las tres
familias, arriba mencionadas, se han evo-

lucionado como lineas autoctonas de
paleoquiropteros presentes en el Nuevo
Mundo desde un momento tan temprano
como el medio del Paleoceno.

Se reconocen dos generos, Pterono-
tus y Mormoops. Se divide el genero
Pteronotus en tres subgeneros, Phijllo-
dia, Chilomjcteris y Pteronotus. La dico-
tomia entre los dos generos es bien
marcado por varios rasgos unicos. Den-
tro del genero Pteronotus, la semejanza
entre los dos subgeneros Chilomjcteris
y Pteronotus es mayor de lo que es la
de cualquiera de los dos referente al
PhijUodia.

Se comentan los patrones de varia-
cion sexual secundaria, individual y
estacional con referencia al pelage y la
muda. Tambien, se presenta una clave
referente a los generos y subgeneros.

La variacion geografica de cada espe-
cie mormoopida fue analizada estadis-
ticamente ye se comenta junta con los
factores ecologicos y fisiograficos de las
regiones en las que viven estas especies.
Dos especies son monotipicas. Seis son
politipicas y se traten de 21 subespecies.
Se llaman a la atencion tres de estas
subespecies (Pteronotus parnellii meso-
americanus, Pteronotus davyi incae y
Mormoops megalophijlla carteri), las
cuales son enteramente nuevas.

La zoogeografia de la familia tam-
bien resulta borrosa no solo por la au-
sencia de una historia fosilifera sino tam-
bien por las distribuciones geograficas
muy simpatricas en la Neotropica. El
centro de origen se situada probable-
mente en el sur de la America Gentral
o en las partes del noroeste de la Amer-
ica del Sur. Representantes de ambos

SMITH: SYSTEMATICS OF MORMOOPID BATS

127

generos florecen en las Antillas Mayores.
Parece que estos murcielagos emigraron
de la America Central y se han diferen-
ciado en especies o subespecies de los

parentes del continente. Se comentan
los factores zoogeograficos que han in-
fluido en la dispersion y la evolucion de
la familia en las islas y el continente.

LITERATURE CITED

Allen, G. M.

1911. Mammals of the West Indies. Bull.
Mus. Comp. Zool., Harvard, 54:175-
263.

1916. A third species of Cliilonijcteris from
Cuba. Proc. New England Zool. Club,
6:1-7.

1917. Two undescribed West Indian Bats.
Proc. Biol. Soc. Washington, 30:156-
170.

1935. Bats from the Panama region. Jour.
Mamm., 16:226-228.
Allen, J. A.

1894. Remarks on specimens of Chilonyc-
teris rubiginosus from western Mexico,
and on the color phases of Pteronotus
davyi Gray. Bull. Amer. Mus. Nat.
Hist, 6:247-248.

1904. New bats from tropical America, with
note on Otopterus. Bull. Amer. Mus.
Nat. Hist., 20:227-237.

1911. Mammals from Venezuela collected
by Mr. M. A. Carriker, Jr., 1909-1911.
Bull. Amer. Mus. Nat. Hist., 30:239-
273.
Alvarez, T.

1963. The Recent mammals of Tamaulipas,
Mexico. Univ. Kansas Publ., Mus.
Nat. Hist., 14:363-473.
Baer, J.

1952. Ecology of animal parasites. The Uni-
versity of Illinois Press, Urbana, 223
pp.
Baker, R. J.

1967. Karyotypes of bats of the family Phyl-
lostomidae and their taxonomic impli-
cations. Southwestern Nat., 12:407-
428.

1970. Karyotypic trends in bats. Pp. 65-69,
in Biology of bats. (W. A. Wimsatt,
ed. ) . Academic Press, New York and
London, 1 :xii -f 406 pp.
Baker, R. J., and L. Christian son

1966. Notes on bats from Sonora, Mexico.
Southwestern Nat., 11:310-311.

Baker, R. J., and T. C. Hsu

1970. Further studies in the sex chromosome
systems of American leaf-nosed bats
(Chiroptera: Phyllostomatidae ) . Cy-
togenetics, 9:131-138.

Baker, R. J., and J. L. Patton

1967. Karyotypes and karyotypic variation
in vespertilionid bats. Jour. Mamm.,
48:270-286.

Benedict, F. A.

1957. Hair structure as a generic character
in bats. Univ. California Publ. Zool.,
59:285-548.
Bonaparte, C. L.

1831. Saggio di una distribuzione metodica
degli animali vertebrati. Rome, 144
pp.
Burt, W. H., and R. A. Stirton

1961. The mammals of El Salvador. Misc.
Publ. Mus. Zool., Univ. Michigan,
117:1-69.
Cavalli-Sforza, L. L., a.xd A. W. F. Edwards
1967. Phylogenetic analysis: models and
estimation procedures. Evolution, 21:
550-570.
Chapman, F. M.

1926. The distribution of bird-life in Ecua-
dor: a contribution to a study of the
origin of Andean bird-life. Bull.
Amer. Mus. Nat. Hist., 55:1-784.
Cockrum, E. L., and G. R. Bradshaw

1963. Notes on mammals from Sonora, Mex-
ico. Amer. Mus. Novit., 2138:1-9.

Constantine, D. G.

1958a. Color variation and molt in Mormoops

megalopJiJ/Uo. Jour. Mamm., 39:344-

347.
1958b. Bleaching of hair pigment in bats by

the atmosphere in caves. Jour.

Mamm., 39:513-520.
Da Cunha Vieira, C. O.

1942. Ensaio monografico sobre os qui-

ropteros do Brasil. Arquiv. d. Zool. d.

Estado d. Sao Paulo, 3:219-471.
1955. Lista remissiva dos mamiferos do

Brasil. Arq. d. Zool. d. Estado d. Sao

Paulo, 8:341-474.
Dalquest, W. W., and H. J. Werner

1954. Histological aspects of the faces of

North American bats. Jour. Mamm.,

35:147-160.

Dalquest, W. W., H. J. Werner,

and J. H. Roberts

1952. The facial glands of a fruit-eating bat,
Artibeus jamaicensis Leach. Join-.
Mamm., 33:102-103.

Davis, W. B.

1965. Review of the Eptesicus brasiliensis
complex in Middle America with the
description of a new subspecies from
Costa Rica. Jour. Mamm., 46:229-
240.

128

MISCELLANEOUS PUBLICATION MUSEUM OF NATURAL HISTORY

1968. Review of the genus Uroderma (Chi-
roptera). Jour. Mamm., 49:676-698.

1969. A re\'iew of the small fruit hats (genus
Artibeus) of Middle America. South-
western Nat., 14:15-29.

1970. A review of the small fruit bats (genus
Artibeus) of Middle America. Part II.
Southwestern Nat., 14:389-402.

Davis, W. B., and D. C. Carter

1962. Notes on Central American l^ats with

description of a new subspecies of

Mormoops. Southwestern Nat., 7:64-

74.
Davis, W. B., D. C. Carter, and R. H. Pine
1964. Noteworthy records of Mexican and

Central American bats. Jour. Mamm.,

45:375-387.
DE LA Torre, L.

1955. Bats from Guerrero, Jalisco and Oa-

xaca, Mexico. Fieldiana:Zool., 37:695-

701.
1961. The evolution, variation, and system-

atics of the Neotropical liats of the

genus Sitirnira. Ph.D. dissertation,

Univ. Illinois, 146 pp.
Dice, L. R., and H. J. Leraas

1936. A graphic method for comparing sev-
eral sets of measurements. Contrib.

Lab. Vert. Genetics, Univ. Michigan,

3:1-3.
DoBsoN, G. E.

1875. Conspectus of the suborders, families,

and genera of Chiroptera arranged

according to their natural affinities.

Ann. Mag. Nat. Hist., 4th ser., 95:

345-357.
1878. Catalogue of the Chiroptera in the

collection of the British Museum.

British Museum, London, xlii + 567

pp.
Duncan, D. B.

1955. Multiple range and multiple F tests.
Biometrics, 11:1-42.

Dunn, E. R.

1940. Some aspects of herpetology in lower
Central America. Trans. New York
Acad. Sci., ser. 2, 2:156-158.

Felten, H.

1956. Fledermjiuse (Mammalia, Chiroptera)
aus El Salvador. Teil 2. Sencken-
bergiana Biol., 37:69-86.

FiNDLEY, J. S.

1969. Brain size in bats. Jour. Mamm.,
50:340-344.

FiTZINGER, J.J.

1870. Kritishe Durchsicht der Ordnung
der P'latterthiere oder Handfliigler
(Chiroptera). Sitzb. k. Akad. Wiss.,
Wien, 61:1-114.
Flower, W. H., and R. Lydekker

1891. An introduction to the study of mam-
mals living and extinct. Adam and
Charles Black London, xvi -f 763 pp.

Forman, G. L., R. J. Baker, and J. D. Gerber

1968. Comments on the systematic status of
vampire l^ats (family Desmodontidae).
Syst. Zool., 17:417-425.

Furman, D. p.

1966. Spinturnicid mites of Panama. Pp.
125-166, ()i Ectoparasites of Panama.
(R. L. Wenzel and V. J. Tipton, eds.),
Field Mus. Nat. Hist., Chicago, xii +
861 pp.

Gabriel, K. R.

1964. A procedure for testing the homoge-
neity of all sets of means in analysis
of variance. Biometrics, 20:4.59-477.

1967. Simultaneous test procedures in multi-
variate analysis of variance. Univ.
North Carolina Dept. Biostat., mim-
eo sen, 539:1-31.

Gabriel, K. R., and R. R. Sokal

1969. A new statistical approach to geo-
graphic variation analysis. Syst. Zool.,
18:259-278.

Gerber, J. D., and C. A. Leone

1971. Immunologic comparisons of the sera
of certain phyllostomatid bats. Syst.
Zool., 20:160-166.
Gill, T.

1872. Arrangement of the families of mam-
mals with analytical tables. Smith-
sonian Misc. Coll., ll:vi + 98 pp.

1901. The bat genus Pteronotus renamed
Dermonotus. Proc. Biol. Soc. Wash-
ington, 14:177.

Goodwin, G. G.

1969. Mammals from the state of Oaxaca,
Mexico, in the American Museum of
Natural History. Bull. Amer. Mus.
Nat. Hist., 141:1-270.
Goodwin, G. G., and A. M. Greenhall

1961. A review of the bats of Trinidad and
Tobago. Bull. Amer. Mus. Nat. Hist.,
122:187-302.
Gosse, P. H.

1851. A naturalist's sojourn in Jamaica.
Mem. Council Royal Soc. Agric,
xxiv + 508 pp.

Gray, J. E.

1821. On the natural arrangement of verte-
brose animals. London Med. Repos.,
15:296-310.

1825. An outline of an attempt at the dis-
position of Mammalia into tribes and
families, with a list of the genera ap-
parently appertaining to each tribe.
Ann. Philos., 26:3.37-344.

1838. A revision of the genera of bats (Ves-
pertilionidae), and the description of
some new genera and species. Mag.
Zool. Bot., 2:483-505.

1839. Descriptions of some Mammalia dis-
coxered in Cuba by W. S. MacLeay,
Esq. Ann. Nat. Hist., 4:1-7.

SMITH: SYSTEMATICS OF MORMOOPID BATS

129

1843a. [Letter addressed to Mr. Waterhouse].

Proc. Zool. Soc. London, pp. 20-22.
1843b. [Letter addressed to the Curator].

Proc. Zool. Soc. London, p. 50.
1866a. Synopsis of the genera of Vesper-

tilionidae and Noctilionidae. Ann.

Mag. Nat. Hist., ser. 3, 17:89-93.
1866b. Notice of the examination of the type

specimen of Aello cuvieii of Leach.

Proc. Zool. Soc. London, pp. 147-148.
Gregory, W. K.

1949. The humerus from fish to man. Amer.

Mus. Novit., 1400:1-54.
Griffin, D. R.

1958. Listening in the dark. Yale Univ.

Press, New Haven, Connecticut, xviii

+ 413 pp.
Griffin, D. R., and A. D. Grinnell

1958. Ability of bats to discriminate echoes

from louder noise. Science, 128:145-

147.
Griffin, D. R., and A. Novick

1955. Acoustic orientation of neotropical

bats. Jour. Exp. Zool., 1.30:251-300.
Grinnell, A. D., and D. R. Griffin

1958. The sensitixity of echolocation in bats.
Biol. Bull., 114:10-22.

Grummon, R. a., and a. Novick

1963. Obstacle avoidance in the bat, Macro-
tis mexicanus. Physiol. Zool., 36:361-
369.

Gundlach, J-

1840. Beschreibung von vier auf Cuba
gefangnen Fledermausen. Arch.
Naturgesch., 6:356-358.

1861. [Beobachteten Flederthiere auf Cuba].
Monatsber. k. Preuss. Akad. Wiss.,
Berhn, pp. 149-156.
Haffer, J.

1967a. Zoogeographical notes on the "non-
forest" lowland bird favmas of north-
western South America. El Hornero,
10:315-3.33.

1967b. Some allopatric species pairs of birds
in northwestern Colombia. Auk, 84:
343-365.

1969. Speciation in Amazonian forest birds.
Science, 165:131-137.

1970. Geologic-climatic history and zooge-
ographic significance of the Uraba
region in northwestern Colombia. Cal-
dasia [Bogota], 10 (50): 603-636.

Hall, E. R., and K. R. Kelson

1959. The mammals of North America.
Ronald Press, New York, l:xxx + 546
+ 79.

Handley, C. O., Jr.

1966. Checklist of mammals of Panama. Pp.
753-795 in Ectoparasites of Panama.
(R. L. Wenzel and V. J. Tipton, eds.).
Field Mus. Nat. Hist., Chicago, xii -f-
861 pp.

Hershkovitz, p.

1951. Mammals from British Honduras,
Mexico, Jamaica, and Haiti. Fieldi-
ana:Zool., 31:547-569.

HussoN, A. M.

1962. The bats of Suriname. E. J. Brill,
Leiden, 282 pp.

Jepsen, G. L.

1966. Early Eocene bat from Wyoming.
Science, 154:1333-1339.

1970. Bat origins and evolution. Pp. 1-64
in Biology of bats. (W. A. Wimsatt,
ed. ), Academic Press, New York and
London, 1 :xii + 406 pp.

Jones, J. K., Jr.

1966. Bats from Guatemala. Univ. Kansas
Publ., Mus. Nat. Hist., 16:439-472.

Jones, J. K. Jr., and C. J. Phillips

1969. Comments on systematics and zooge-
ography of bats in the Lesser Antilles.
Studies on the Fauna of Curacao and
other Caribbean Islands, 32:131-145.

Jones, J. K., J. D. Smith, and R. W. Turner

1971. Noteworthy records of bats from Nic-
aragua, with a checklist of the chi-
ropteran fauna of the country. Occas.
Papers Mus. Nat. Hist., Univ. Kansas,
2:1-35.

Koch, C.

1862-63. Das Wesentliche der Chiropteren
mit besonderer Beschreibung der in
deni Herzogthum Nassau und den
angriinzenden Landestheilen vorkom-
menden Fledermause. Jahrg. Ver.
Naturk. Herzogthum Nassau, 17-18:
261-593.

Koopman, K. F.

1951. Fossil bats from the Bahamas. Jour.

Mamm., 32:229.
1955. A new subspecies of Chilonycteris

from the West Indies and a discussion

of the mammals of La Gonave. Jour.

Mamm., 36:109-113.

1958. Land bridges and ecology in bat dis-
tribution on islands off the northern
coast of South America. Evolution,
12:429-439.

1970. Zoogeography of bats. Pp. 29-50, in
About bats (B. H. Slaughter and D.
W. Walton, eds.). Southern Metho-
dist Univ. Press, vii + 339 pp.

Koopman, K. F., and P. S. Martin

1959. Subfossil mammals from the Gomez
Farias region and the tropical gradient
of eastern Mexico. Jour. Mamm., 40:
1-12.

Koopman, K. F., and E. L. Cockrum

1967. Bats. Pp. 109-150, in Recent mam-
mals of the World. ... (S. Anderson
and J. K. Jones, Jr., eds.), Ronald
Press, New York, vii -j- 453 pp.

130

MISCELLANEOUS PUBLICATION MUSEUM OF NATURAL HISTORY

Kruskal, J. B.

1956. On the shortest spanning subtree of a
graph and the tra\'eling salesman
problem. Proc. Amer. Math. Soc,
7:48-50.

KULZER, E.

1968. Der Aug des afrikanischen Flughundes
Eidolon helvnm. Natur und Museum,
98:181-194.
Lawrence, B.

1943. Miocene bat remains from Florida,
with notes on the generic characters
of the humerus of bats. Jour. Mamm.,
24:356-369.
Leach, W. E.

1821a. The characters of three new genera of
bats without foliaceous appendages to
the nose. Trans. Linn. Soc, 13:69-72.

1821b. The characters of seven genera of bats
with foliaceous appendages to the nose.
Trans. Linn. Soc, 13:73-82.
Lessox, R. p.

1842. Nouveau tableau du regne animal.
Paris, vi -f- 204 pp.
Machado-Allisox, C. E.

1965a. Notas sobre Mesostigmata neotrop-
icales III. Cameronieta thomasi:
nuevo genero y nueva especie parasita
de Chiroptera (Acarina, Spintmni-
cidae). Acta Biol. Venezuelica, 4:244-
253.

1965b. Las especies Venezolanas del genero
Periglischriis Kolenati, 1857, (Acarina,
Mesostigmata, Spinturnicidae). Acta
Biol. Venezuelica, 4:259-348.

1967. The systematic position of the bats
Desmodus and Chilonycteris, based
on host-parasite relationships ( Mam-
malia; Chiroptera). Proc. Biol. Soc.
Washington, 80:223-226.
McKenxa, M. C.

1967. Classification, range, and deployment
of the Prosimian Primates. Colloques
Internationaux du Centre National de
la Recherche Scientifique, 163:603-
610.
Miller, G. S., Jr.

1902. Twenty new American bats. Proc.
Acad. Nat. Sci. Philadelphia, 54:389-
412.

1904. Notes on the bats collected by Wil-
liam Palmer in Cuba. Proc. U.S. Nat.
Mus., 27:337-348.

1907. The families and genera of bats. Bull.
U.S. Nat. Mus., 57:xvii + 282 pp.
Mitchell, H. A.

1964. Investigation of the cave atmosphere
of a Mexican bat colony. Jour.
Mamm., 45:568-577.
NoviCK, A.

1958. Orientation in Paleotropical bats. I.
Microchiroptera. Jour. Exp. Zool.,
138:81-154.

1962. Orientation in Neotropical bats. I.

Natalidae and Eml^allonuridae. Jour.

Mamm., 43:449-455.
1963a. Orientation in Neotropical bats. II.

Phyllostomatidae and Desmodontidae.

Jour. Mamm., 44:44-56.
19631:). Pulse duration in the echolocation of

insects by the bat, Fteronotus. Er-

gebnisse Biologic, Berlin, 26:21-26.
1965. Echolocation of flying insects ])y the

bat, Chilonycteris psilotis. Biol. Bull.,

128:297-314.
NoviCK, A., axd J. R. Vaisxys

1964. Echolocation of flying insects by the

bat, Chilonycteris parnelhi. Biol.

Bull., 127:478-488.
Palmer, T. S.

1904. Index generum mammalium. . . . N.

Amer. Fauna, 23: 1-984.
Peters, W.

1856. Ueber die systematische Stellung der

Gattimg Mormops Leach und iiber

die Classification der PhtjUostomata so

wie iiber eine neue Art der Gattung

Vampynis. Monatsber. k. Preuss.

Akad.'Wiss., Berlin, 1856:409-415.

1864. Uber einige neue Saugethiere (Mor-
moops, Macwtus, Vespems, MoJossns,
Caproniijsj, Arnphibien (Platydactylus,
Otocryptis, Euprepes, Ungalia, Dro-
micus, Tropidonotus, Xenodon, Hy-
lodesj und Fishe (Sillago, Sebastes,
Channa, Myctophum, Carassixis, Bar-
bus, Capocta, Poecdia, Saiirenchelys,
Leptocephalus). Monatsber. k. Preuss.
Akad. Wiss., Berfin, 1865:381-399.

1865. Vorlage von Abbildungen zu einer
Monographic der Chiropteren, und
Ubersicht der von ihm befolgten sys-
tematischen Ordnung der hieher ge-
horigen Gattrmgen. Monatsber. k.
Preuss. Akad. Wiss., Berlin, 1865:256-
258.

1866. Fernere Mittheilungen zur Kenntnifs
der Flederthiere, namentlich iiber
Arten des Leidener und Britischen
Museums. Monatsber. k. Preuss. Akad.
Wiss., Berhn, 1866:672-681.

1872. Uber die zu der Gruppe der Mor-
mopes gehorigen Fledertliiere. Mo-
natslier. k. Preuss. Akad. Wiss., Ber-
lin, 1872:358-361.
Petersox, R. L.

1968. A new bat of the genus Vampyressa
from Guyana, South America with a
brief systematic review of the genus.
Life Sci. Contrib., Royal Ontario
Mus., 73:1-17.

Phillips, C. J., and J. K. Jones, Jr.

1969. Dental abnomialities in North Ameri-
can bats. I. Emballonuridae, Nocti-
lionidae, and Chilonycteridae. Trans.
Kansas Acad. Sci., 71:509-520.

SMITH: SYSTEMATICS OF MORMOOPID BATS

l3l

PiRLOT, P.

1965. Chiropteres de Test de Venezuela. (II.
Delta de rOrenoque). Mammalia,
29:375-389.

Proctor, J. R.

1966. Some processes of numerical taxonomy
in terms of distance. Syst. Zool., 15:
131-140.

Rehx, J. AC.

1901. A study of the genus Centurio. Proc.
Acad. Nat. Sci. Philadelphia, 53:295-
302.

1902. A revision of the genus Mormoops.
Proc. Acad. Nat. Sci., Philadelphia,
54:160-172.

1904a. A study of the Mammalian genus
Chilonycteris. Proc. Acad. Nat. Sci.
Philadelphia, 56:181-207.

1904b. A study of the bats of the genus
Dermonotus ( Pteronotus Auct. ) . Proc.
Acad. Nat. Sci. Philadelphia, 56:250-
256.
Revilliod, p.

1917a. Fledermause aus der Braunkohle von
Messel bei Darmstadt. Abhandl. der
Grossherzoglich Hessischen Geol.
Landesanstalt zu Darmstadt, 3(2):
160-201.

1917b. Contribution a I'etude des Cliiropteres
des terrains tertiaires. Premiere
partie. Adhandl. Schweizerischen der
Paliiontol. Cesell. (Mem. Soc.
Paleontol. Suisse), 43:1-57 -f 8 un-
numbered pages.

1920. Contribution a I'etude des chiropteres
des terrains tertiaires. Deuxieme
partie. Abhandl. Schweizerischen der
Paliiontol. Cresell. ( Mem. Soc.
Paleontol. Suisse), 44:61-128 4-4 im-
numbered pages.

1922. Contribution a I'etude des chiropteres
des terrains tertiaires. Troisie, e
partie. Abhandl. Schweizerschen der
Paljiontol. Cesell. ( Mem. Soc.
Paleontol. Suisse), 45:131-195 -f 4
unnumbered pages.

RiDGWAY, R.

1912. Color standards and color nomen-
clature. Washington, D.C., privately
printed, iv -j- 44 pp.

ROHLF, F. J.

1967. Correlated characters in numerical
taxonomy. Syst. Zool., 16:109-126.

1968. Stereograms in numerical taxonomy.
Syst. Zool., 17:246-255.

RUDNICK, A.

1960. A revision of the mites of the family
Spintiunicidae (Acarina). Univ. Cali-
fornia Publ. Entomol., 17:157-284.

Ryan, R. M.

1963. The biotic provinces of Central Amer-
ica. Acta Zool. Mexicana, 6:1-55.

Sanborn, C. C.

1938. Notes on Neotropical bats. Occas.
Papers Mus. Zool., Univ. Michigan,
373:1-5.
Saussure, M. H.

1860. Note sur quelques mammiferes de
Mexique. [part 5] Revue et Mag.
Zool., 2:281-295.
Savage, D. E.

1951. A Miocene phyllostomatid bat from
Colombia, South America. Univ. Cali-
fornia Publ. Geol. Sci., 28:357-366.

SCHLOSSER, M.

1911. Beitriige zur kenntniss der oligozanen
landsiiugetiere aus dem Fayum.
Beitriige zur Paliiontogie und geologic
Oesterreich-ungarns und des Orients,
Wien, 24:51-167.
Silva-Taboada, C, and R. H. Pine

1969. Moiphological and behavioral evi-
dence for the relationship between the
bat genus Brachyphylla and the Phyl-
lonycterinae. Biotropica, 1:10-19.
Simpson, G. G.

1945. The principles of classification and a
classification of mammals. Bull. Amer.
Mus. Nat. Hist., 85:xvi -f 350 pp.
1956. Zoogeography of West Indian land
mammals. Amer. Mus. Novit., 1759:1-
28.
Simpson, G. G., A. Roe, and R. C. Lewontin

1960. Quantitative Zoology. Harcourt, Brace
and Company, New York and Bur-
lingame, vii -f- 440 pp.

SOKAL, R. R.

1961. Distance as a measure of taxonomic
similarity. Syst. Zool., 10:70-79.

SoKAL, R. R., AND R. C. Rinkel

1963. Geographic variation of Alate Pem-
pJiigus popiilitransversus in eastern
North America. Univ. Kansas Sci.
Bull., 44:467-507.

SOKAL, R. R., and F. J. ROHLF

1969. Biometry: the principles and practice
of statistics in biological research. W.
H. Freeman and Co., xxii -f 776 pp.

SoKAL, R. R., AND P. H. A. Sneath

1963. Principles of numerical taxonomy.
W. F. Freeman and Co., xviii -f- 359
pp.

SoKAL, R. R., AND P. A. Thomas

1965. Geographic variation of Pemphigus
populi-transversus in eastern North
America: stem mothers and new data
on alates. Univ. Kansas Sci. Bull.,
46:201-252.

Spix, J. Von B.

1823. Simiarum et vespertilionum Brasilien-
sium species novae ou histoire natu-
relle des especies nouvelles de singes
et de chauves-souris observees et
recueillies pendant le voyage dans

132

MISCELLANEOUS PUBLICATION MUSEUM OF NATURAL HISTORY

rinterieur de Bresil execute par ordre
de S. M. le Roi de Baviere dans les
annees 1817, 1818, 1819, 1820. Mo-
nachii, 4:viii + 72 pp.

Starrett, a., and R. S. Casebeer

1968. Records of l:)ats from Costa Rica.
Contrib. Sci., Los Angeles Co. Mus.,
148:1-21.

Steel, R. G. D., and J. H. Torrie

1960. Principles and procedures of statistics.
McGraw-Hill, New York, xvi -|- 481
pp.

Thomas, O.

1892. A preliminary list of the mammals of
Trinidad, jour. Field Nat. Club, 1:
158-168.

Trouessart, E. L.

1897. Catalogue mammalium tarn viventum
quam fossilium. Nova edito (prima
completa). Berlin, R. Friedliinder und
Solm, 1:1-218.

Tomes, R. F.

1863. On a new genus and species of leaf-
nosed bats in the museum at Fort Pitt.
Proc. Zool. Soc, London, pp. 81-85.

TUTTLE, M. D.

1970. Distribution and zoogeography of
Peruvian bats, with comments on nat-
ural history. Univ. Kansas Sci. Bull.,
49:45-86.

Van Gelder, R. G.

1959. A taxonomic revision of the spotted
skunks ( genus Spilogale). Bull. Amer.
Mus. Nat. Hist, 117:229-392.
Van Valen, L.

1966. Delta theridia, a new order of mam-
mals. Bull. Amer. Mus. Nat. Hist.,
132:1-126.

Vaughan, T. a.

1959. Functional moiphology of three bats:
Eumops, Myotis, Macrotus. Univ.
Kansas Publ., Mus. Nat. Hist., 12:1-
153.

Vaughan, T. A., and G. C. Bateman

1970. Fimctional morphology of the fore-
limb of mormoopid bats. Jour.
Mamm., 51:217-235.
Villa-R., B., and a. Jimenez

1960. Acerca de la posicion taxonomica de
Mormoops megalupJujUa senicula
Rehn, y la presencia de virus rabico

en estos murcielagos insectivores. An.

Inst. Biol., 31:501-509.
Wagner, J. A.

1840. Die Affen und Flederthiere. Supple-

mentband, erste Abtheil., in J. Ch. D.

\<m Schreder, Die Sjiugthiere in Ab-

bildungen nach der Natur mit

Beschreibungen. xiv -)- 558 pp.
1843. Diagnosen neuer Arten brasilischer

Handflungler. Arch. Naturgesch., 9:

365-368.
1847. Beitriige zur Kenntniss der Siiugthiere

Amerika's. Abhandl. Math.-Physik. k.

Bayerischen Akad. Wiss., Miinchen,

5:121-208.
Walker, E. P., and others

1964. Mammals of the World. The Johns

Hopkins Press, Baltimore, l:xliii-t-

646 pp.
Walton, D. W., and G. M. Walton

1968. Comparative osteology of the pelvic

and pectoral girdles of the phyllosto-

matidae ( Chiroptera : Mannnalia ) .

Jour. Grad. Research Center, 37:1-35.
Wenzel, R. L., and V. J. Tipton

1966. Some relationships between mammal

hosts and their ectoparasites. Pp.

677-723, in Ectoparasites of Panama.

(R. L. Wenzel and V. J. Tipton, eds.),

Field Mus. Nat. Hist., Chicago, xii +

861 pp.
Wenzel, R. L., V. J. Tipton, and A. Kiewlicz
1966. The Streblid batflies of Panama ( Dip-

tera Calypterae:Streblidae). Pp. 405-

575, in Ectoparasites of Panama. ( R.

L. Wenzel and V. J. Tipton, eds.),

Field Mus. Nat. Hist., Chicago, xii -|-

861 pp.
Werner, H. J., and W. W. Dalquest

1952. Facial glands of the tree bats, Lasiii-

rus and Dasyptertis. jour. Mamm.,

33:77-80.
Werner, H. J., W. W. Dalquest,
AND J. H. Roberts

1950. Histological aspects of the glands of

bats, Tadarida cijnocephala (Le

Conte). Jour. Mamm., 31:395-399.
Winge, H.

1892. Jordfundne og nulevende Flagermus

(Chiroptera) fra Logoa Santa, Minas

Geraes, Brasilien, Copenhagen, Museo

Lundii, 2:1-65.

I

3 2044 093 361 616

Date Due

BOUND

1972